100% found this document useful (1 vote)
339 views211 pages

Manual WP 34s 3 1

Manual for the WP 34S Calculator

Uploaded by

richrichri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
100% found this document useful (1 vote)
339 views211 pages

Manual WP 34s 3 1

Manual for the WP 34S Calculator

Uploaded by

richrichri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 211

WP 34S OWNER’S MANUAL

This manual documents WP 34S. WP 34S is free software: you can


redistribute it and / or modify it under the terms of the GNU General
Public License as published by the Free Software Foundation, either
version 3 of the License, or (at your option) any later version.

WP 34S is distributed in the hope that it will be useful, but without any
warranty; without even the implied warranty of merchantability or fit-
ness for a particular purpose. See the GNU General Public License
for more details.
You should have received a copy of the GNU General Public License
along with WP 34S. If not, please see http://www.gnu.org/licenses/ .

This manual contains valuable information. It was designed


and written in the hope it will be read thoroughly by you.
For really quick and impatient users, however, there is a stand-alone
two-page Troubleshooting Guide starting on page 180 of this manual.
For those who don’t even read this: Sorry, we can’t help you.

This manual may change without notice if we, the developers, modify WP 34S. We reserve the right to do so
at any time. The cover picture may stay unchanged – the document status can be seen from the last note on
last page. We recommend you watch http://sourceforge.net/projects/wp34s/develop to stay informed.

WP 34S Owner’s Manual Edition 3.1 Page 1 of 211


WP 34S would not have reached its present state without our love for Classics, Woodstocks, Spices, Nuts,
Voyagers, and Pioneers. Thus we want to quote what was printed in HP pocket calculator manuals until
1980, so it will not fade:

“The success and prosperity of our company will be assured only if we offer our customers
superior products that fill real needs and provide lasting value, and that are supported by a
wide variety of useful services, both before and after sales.”

Statement of Corporate Objectives.


Hewlett-Packard

Just in Case …

... you do not have your WP 34S calculator yet: WP 34S runs on an HP-20b Business Consultant or an
HP-30b Business Professional. Both are financial pocket calculators. So if one of those is sitting on your
desk unchanged as produced for HP, please turn to Appendix A for instructions how to convert it into a full-
fledged WP 34S yourself. On the other hand, if you do not want to bother with cables on your desk
connecting it to your computer, with flashing the calculator firmware and attaching a sticky overlay, you can
purchase an HP-30b-based WP 34S readily on the internet; see e.g. http://commerce.hpcalc.org/34s.php or
http://www.thecalculatorstore.com/epages/eb9376.sf/en_GB/?ObjectPath=/Shops/eb9376/Products/%22WP34s%20Pack%22 .

The first way (doing it yourself) will just cost you some time, the second will cost you some money. If you
choose buying your WP 34S at one of the sites mentioned, we (the developers) will get a very modest
fraction of the price to support our otherwise unpaid efforts on the WP 34S project. Either way will work – it is
your choice.

Furthermore, there are three optional hardware modifications requiring some fine soldering, cutting,
gluing, and drilling a little hole in the plastic case of your business calculator. It is no sorcery for a
young electronics engineer with a slow hand and good eyesight, but might come close to it for others
(including me, W.B.). So check Appendix H and decide whether you want to do that yourself. Else look
for somebody who is able to do it for you.

The WP 34S package includes a full size emulator, so you may test it on your computer before buying any
hardware. And later on, you may test your programs there as well before sending them to your WP 34S. The
function sets are identical.

For the following, we assume the conversion is done and you hold your WP 34S in your hands.

WP 34S Owner’s Manual Edition 3.1 Page 2 of 211


TABLE OF CONTENTS

Welcome! ................................................................................................................ 7
Print Conventions and Common Abbreviations ....................................................9

Getting Started ...................................................................................................... 10


Keyboard Basics ................................................................................................11
Entering Numbers ..............................................................................................15
Real Calculations ...............................................................................................16
Elementary Stack Mechanics .............................................................................17
Error Recovery ...................................................................................................21
Calling Commands Unseen on the Keyboard .....................................................22

Addressing Objects ............................................................................................... 24


The Virtual Keyboard for Addressing – Transient Alpha Mode (αT) .....................25
Addressing Real Numbers..................................................................................26
Real Matrices and Vectors .................................................................................28
Complex Calculations .........................................................................................29
Addressing Complex Numbers ...........................................................................33

Your WP 34S in Various Operating Modes ........................................................... 35


The Display ........................................................................................................35
Recognizing Calculator Modes ...........................................................................36
Common Commands Returning Specific Displays: STATUS, VERS, ERR, etc. .38
Floating Point Modes – 1: Introduction and Localisation ....................................39
Floating Point Modes – 2: Displaying Decimal Numbers, Fractions, Times, etc. 40
Floating Point Modes – 3: Real Statistical Calculations ......................................45
Integer Modes – 1: Introduction and Virtual Keyboard.....................................52
Integer Modes – 2: Displaying Integer Numbers .............................................53
Integer Modes – 3: Bitwise Operations ...........................................................56
Integer Modes – 4: Integer Arithmetic .............................................................58
Full Alpha Mode – 1: Introduction and Virtual Keyboard..................................61
Full Alpha Mode – 2: Displaying Text ..............................................................64

Programming Your WP 34S.................................................................................. 66


Labels ................................................................................................................67
Tests ..................................................................................................................67
Local Data ..........................................................................................................69
Programmed Input and Output, User Interaction and Dialogues .........................69
Keyboard Codes and Direct Keyboard Access ...................................................71
Flash Memory (FM) and XROM..........................................................................72

WP 34S Owner’s Manual Edition 3.1 Page 3 of 211


Index of Operations (IOP) ..................................................................................... 74
A ........................................................................................................................76
B ........................................................................................................................77
C ........................................................................................................................78
D ........................................................................................................................81
E ........................................................................................................................83
F.........................................................................................................................85
G ........................................................................................................................86
H ........................................................................................................................88
I ..........................................................................................................................88
J .........................................................................................................................89
K ........................................................................................................................89
L .........................................................................................................................89
M ........................................................................................................................92
N ........................................................................................................................94
O ........................................................................................................................95
P ........................................................................................................................95
R ........................................................................................................................97
S ...................................................................................................................... 100
T....................................................................................................................... 105
U ...................................................................................................................... 105
V ...................................................................................................................... 106
W ..................................................................................................................... 106
X ...................................................................................................................... 107
Y ...................................................................................................................... 108
Z....................................................................................................................... 108
α ....................................................................................................................... 109
β ....................................................................................................................... 111
γ ....................................................................................................................... 111
δ ....................................................................................................................... 111
ε ....................................................................................................................... 111
ζ ....................................................................................................................... 111
π ...................................................................................................................... 111
σ....................................................................................................................... 112
φ ...................................................................................................................... 113
χ ....................................................................................................................... 113
The Rest .......................................................................................................... 113
Nonprogrammable Control, Clearing and Information Keys .............................. 119
Alphanumeric Input .......................................................................................... 123

WP 34S Owner’s Manual Edition 3.1 Page 4 of 211


Catalogs and Browsers ....................................................................................... 125
Catalog Contents in Detail ................................................................................ 130
Accessing Catalog Items the Fast Way ............................................................ 133
Constants (CONST) ......................................................................................... 134
Unit Conversions (CONV) ................................................................................ 139
Predefined Global Alpha Labels (CAT) ............................................................. 143

The Stopwatch Application ................................................................................. 145

Appendix A: Setup and Communication ............................................................. 147


How to Flash Your HP-20b or HP-30b .............................................................. 147
Overlays – Where to Get Them and How to Make Them Yourself if Required.. 151
Updating Your WP 34S .................................................................................... 152
I/O Overview .................................................................................................... 153

Appendix B: Memory Management ..................................................................... 154


Status and Configuration Data ......................................................................... 154
Global Registers............................................................................................... 154
Summation Registers ....................................................................................... 156
Subroutine Return Stack (SRS) and Program Memory..................................... 156
Making Room for Your Needs .......................................................................... 157
Addressing and Accessing Local Data, Recursive Programming...................... 158
Switching between Standard Real (SP) and Integer Modes ............................. 159

Appendix C: Messages and Error Codes ............................................................ 162

Appendix D: The WP 34S Emulator on Your Computer...................................... 165


Data Transfer between Your WP 34S and Your Computer ............................... 166
Mapping of Memory Regions to Emulator State Files ....................................... 168
Simulating a Printer on Your Computer ............................................................ 168

Appendix E: Character Sets................................................................................ 169

Appendix F: Corresponding Operations to the HP-42S and -16C Function Sets 172
HP-42S ............................................................................................................ 172
HP-16C ............................................................................................................ 178

Appendix G: Troubleshooting Guide ................................................................... 180


Overcoming Troubles in Calculator Operation .................................................. 180
Overcoming Flashing Troubles ......................................................................... 182

WP 34S Owner’s Manual Edition 3.1 Page 5 of 211


Appendix H: Additional Information for Advanced Users .................................... 184
Changing Word Size in Integer Modes ............................................................. 184
Mode Storing and Recalling ............................................................................. 184
Commands for Advanced Users ....................................................................... 185
Double Precision (DP) Calculations and Mode Switching ................................. 186
Further Commands Used in Library and XROM Routines ................................ 189
Assembler Output ............................................................................................ 191
Basic Electrical Hardware Specifications of the HP-20b and HP-30b ............... 191
Tuning the Hardware of Your WP 34S.............................................................. 192

Appendix I: Advanced Mathematical Functions .................................................. 198


Numbers .......................................................................................................... 198
Statistical Distributions ..................................................................................... 199
More Statistical Formulas ................................................................................. 204
Orthogonal Polynomials ................................................................................... 207
Even More Mathematical Functions ................................................................. 208

Appendix J: Release Notes................................................................................. 210

WP 34S Owner’s Manual Edition 3.1 Page 6 of 211


WELCOME!

Dear user, now you have got it: your very own WP 34S. It uses the mechanics and
hardware of an HP-20b Business Consultant or an HP-30b Business Professional, re-
spectively, so you benefit from the excellent processor speed of these pocket calculators.
And with an HP-30b you also get the famous rotate-and-click keys, giving you the tactile
feedback that has been appreciated in vintage Hewlett-Packard calculators for decades.

On the other hand, the firmware and user interface of your WP 34S were thoroughly
thought through and discussed by us, newly designed and written from scratch, loaded
with functions, pressed into the little memory available, and tested over and over again to
give you a fast and compact scientific calculator like you have never had before –
fully keystroke programmable, comfortably fitting into your shirt pocket, and RPN1.

The function set of your WP 34S is based on the famous HP-42S RPN Scientific of 1988,
the most powerful programmable RPN calculator industrially built so far.2 We expanded
the set, incorporating the functions of the renowned computer scientist’s HP-16C, the
fraction mode of the HP-32SII, and probability distributions similar to those of the HP-21S.
We also included numerous additional useful functions for mathematics, statistics,
physics, engineering, programming, I/O, etc., such as
+ Euler’s Beta and Riemann’s Zeta functions, Bernoulli and Fibonacci numbers,
Lambert’s W, the error function, and the Chebyshev, Hermite, Laguerre, and
Legendre orthogonal polynomials (no more need to carry heavy printed tables),
+ many statistical distributions and their inverses: Poisson, Binomial, Geometric,
Cauchy-Lorentz, Exponential, Logistic, Weibull, Lognormal, and Gaussian,
+ programmable sums and products, first and second derivatives, solving quadratic
equations for real and complex roots,
+ testing for primality,
+ integer computing in fifteen bases from binary to hexadecimal,
+ extended date and time operations and a stopwatch3 based on a real-time clock,
+ financial operations such as mean rate of return and margin calculations,
+ 88 conversions, mainly from old Imperial to universal SI units and vice versa,
+ 50 fundamental physical constants as accurate as used today by national standards
institutes such as NIST or PTB, plus a selection of important constants from
mathematics, astronomy, and surveying,
+ bidirectional serial communication with your computer, as well as printing on an HP
82240A/B 4,
+ battery-fail-safe on-board backup memory,
+ Greek and extended Latin letters covering the languages of almost half of the world’s
population (upper and lower case in two font sizes), plus mathematical symbols.

1
RPN stands for reverse Polish notation, a very effective and coherent method of calculating (see p. 15).
2
The matrix menu of the HP-42S cannot be supported by WP 34S for hardware reasons. Your WP 34S
features a set of basic matrix commands and several library routines dealing with matrices instead.
3
The stopwatch requires adding a quartz crystal and two tiny capacitors.
4
Printing requires adding an IR diode and a resistor. Even a USB board is available for your WP 34S.
WP 34S Owner’s Manual Edition 3.1 Page 7 of 211
WP 34S is the first RPN calculator overcoming the limits of a four-level stack –
stop worrying about stack overflow in calculations. WP 34S features a choice of two stack
sizes expanded by a complex LASTx register: traditional four stack levels for HP
compatibility, eight levels for convenient calculations in complex domain, for advanced real
calculus, vector algebra in 4D, or whatever application you have in mind. You will find a full
set of commands for stack handling and navigation in either stack size.

Furthermore, your WP 34S features up to 107 global general purpose registers, 112 global
user flags, up to 928 program steps in RAM, up to 6014 program steps in flash memory, a
30 byte alpha register for message generation, 16 local flags as well as up to 144 local
registers allowing for recursive programming, and 4 user-programmable hotkeys for your
personal favorite functions. Most of the memory layout is conveniently user-settable.

WP 34S is the result of a collaboration of two individuals, an Australian and a German, that
started in 2008. We developed WP 34S in our free time – so you may call it our hobby
(although some people close to us found other names for it). From the very beginning, we
have discussed our project in the forum of the Museum of HP Calculators. We would like
to thank all its international members, who taught us a lot and contributed their ideas and
lent their support throughout several stages of our project. Special thanks go to Marcus
von Cube (Germany) who joined us in bringing WP 34S to life by designing an emulator for
v1.14, allowing for widespread use and convenient testing. Starting with v1.17, the
software began running on an HP-20b hardware. A very useful assembler/disassembler
has been provided by Neil Hamilton (Canada) since v1.18 – even a symbolic preprocessor
has been added since v2.1. For v3.0, Pascal Méheut (France) contributed a versatile
flashing tool for various operating systems. With v3.1, printing on an HP82240A/B printer
became possible thanks to the gracious support by Christoph Gießelink (Germany); a set
of micro USB boards was developed by Harald Pott (Germany); Ciaran Brady (UK) wrote a
Beginner’s Guide for our WP 34S, and Christian Tvergaard and Peter Murphy (both USA)
carefully proofread this manual. We greatly appreciate all your support!

We named our baby WP 34S in honor of the HP-34C from 1979, one of the most powerful
compact LED pocket calculators. WP 34S is our humble approach – within the constraints
of HP’s hardware – to a future 43S we can only dream of succeeding the HP-42S at one
point. May our project help to convince those that have access to more resources than we
do: Catering to the market for serious scientific instruments is well worthwhile!

We carefully checked all aspects of WP 34S to the best of our ability. Our hope is that
WP 34S is free of severe bugs. However, this cannot be guaranteed. We promise to
continue improving WP 34S whenever necessary. If you discover any strange results,
please report them to us. If they turn out to be caused by internal errors, we will provide
you with an update as soon as it is available. Just as before, we will continue striving to
maintain short response times.

Enjoy!

Paul Dale and Walter Bonin

WP 34S Owner’s Manual Edition 3.1 Page 8 of 211


Print Conventions and Common Abbreviations

Throughout this manual, standard font is Arial. Emphasis is added by underlining or bold
printing. Specific terms, titles, trademarks, names or abbreviations are printed in italics,
hyperlinks in blue underlined italics. Bold italic letters such as n are used for variables;
constant sample values (e.g. of labels or displayed characters) use bold normal letters.
Calculator COMMANDS are generally called by their names, printed in capitals in running
text.

This font is taken for explicit references to calculator keys. Alphanumeric and numeric
displays (such as Hello! and 12,34 ) are quoted using the respective calculator fonts
where applicable and beneficial.

Courier is employed for file names and numeric formats.

Register ADDRESSES are printed using bold Times New Roman capitals, while register
contents are expressed by lower case bold italics. So e.g. the value y lives in stack register
Y, r45 in general purpose register R45, and alpha in the alpha register. Overall stack
contents are generally quoted in the order [ x, y, z, …] . Lower case normal italics are for
units.

All this holds unless stated otherwise.

The following abbreviations are used at various locations in this manual:


DP = double precision (see p. 186).
FM = flash memory (a special kind of RAM, see p. 72).
IOP = index of operations (see p. 74).
IR = infrared.
LCD = liquid crystal display.
RAM = random access memory, allowing read and write operations.
ROM = read-only memory, allowing only read operations.
SP = single precision (see p. 159).
USB = universal serial bus, i.e. a specific interface.
XROM = extended ROM (see p. 72).
Some more abbreviations may be used locally.

Finally: WARNING indicates the risk of severe errors. There are only four warnings printed in this
manual. Locking up your calculator is the worst that can happen to it, as far as we know.

WP 34S Owner’s Manual Edition 3.1 Page 9 of 211


GETTING STARTED

If you know how to deal with a good old Hewlett-Packard RPN scientific calculator,
you can start using your WP 34S almost right away. Use this manual to get
information on some basic design concepts that put your WP 34S ahead of previous
RPN calculators. Continue using this manual for reference.

On the other hand, if this is your first RPN scientific calculator at all or the first you
use for a long time, we recommend you get an HP-42S Owner’s Manual. It is
available at low cost on a DVD distributed by the Museum of HP Calculators (see
here: http://www.hpmuseum.org/cd/cddesc.htm ).

Read part 1 of said manual as a starter. It includes an excellent introduction to RPN.


RPN is the reason why your WP 34S works without an key; nor does it need the
keys nor . Once you got used to RPN you may never choose a calculator
featuring again.
Part 2 of the HP-42S Owner’s Manual will help you programming your WP 34S for
quick and easy solving lengthy, repeated or iterative computations.

Further documentation, including complete information about the other vintage calcu-
lators and the famous PPC ROM mentioned in this manual, is readily accessible on
said DVD, too.

Alternatively to the HP-42S Owner’s Manual, you can download a dedicated WP 34S
Beginners Guide, recently written by one of our users:
http://sourceforge.net/projects/wp34s/files/doc/WP_34S_Beg_Guide.pdf .

Most traditional commands on your WP 34S will work as they did on the HP-42S. This little
manual here is meant as a supplement presenting all the new features. It contains the
necessary information about them, including equations and technical explanations; it is not
intended, however, to replace textbooks on mathematics, statistics, physics, engineering,
or programming, nor is it a hypothetical Beginner’s Guide to RPN Computing.

Your WP 34S is designed to help you in your calculations and computations.


But it is just a tool – although a very powerful one – it cannot think for you nor can it
check the sensibility of a problem you apply it to. Do not blame us nor your WP 34S
for errors you have made. Gather information, think before and while keying in, and
check your results: these will remain your responsibilities always.

The following text starts with presenting you the user interface, so you learn where you will
find what you are looking for. It continues by demonstrating some basic methods, the
WP 34S memory and addressing items therein, and the display and indicators that give
you feedback about what is going on. Then the major part of this manual consists of an
index of all available operations and how to access them, and of lists of all catalog
contents. This manual closes with appendices covering special topics, e.g. a list of error
messages your WP 34S will return if abnormal conditions prevent it from executing your
command as expected. There you will also find instructions for keeping your WP 34S up-
to-date when new firmware revisions appear.

WP 34S Owner’s Manual Edition 3.1 Page 10 of 211


Keyboard Basics

Start exploring your WP 34S: Press its bottom left key to turn it on – notice that .ON. is
printed below that key. If you turn on your WP 34S the very first time, you will get what you
see displayed below. To turn it off again, press the green key (notice a little ⓗ showing
up top left in display), then (which has OFF printed on its lower part). Since your
WP 34S has Continuous Memory, turning it off does not affect the information it contains.
To conserve battery energy, your WP 34S will shut down some five minutes after you stop
using it – when you turn it on again, you can resume your work right where you left off.

To adjust display contrast, hold down .ON. while you press or .

Look at the keyboard of an HP-42S for


comparison. The most striking differ-
ence to it is the colorful surface of your
WP 34S. You get five functions per
key on average. White print on key top
is for the primary function of the key.
For secondary functions, green labels
are put on the slanted lower faces of
34 keys, golden and blue labels are
printed below them on the key plate.
Grey letters are bottom left of 26 keys.

To access a white label, just press the


corresponding key (thus it is called
primary function). For a golden, blue,
or green label, press the prefix ,
, or , respectively, then the
corresponding key.
Take the key for example. Press-
ing …
will enter the digit 5 in display,
+ will calculate the arithmetic
mean values of data accumulated in
the statistic registers via ,
+ will compute the standard
deviations for the same data via ,
+ will open a catalog (i.e. a
set) of extra statistical functions via
. All labels printed underlined
on the keyboard point to catalogs.
The grey R will become relevant in
alpha mode, i.e. for input of text.

WP 34S Owner’s Manual Edition 3.1 Page 11 of 211


, , and allow for easily accessing a multiple of the 37 primary functions this
hardware can take. The active prefix is indicated by ⓕ, ⓖ, or ⓗ top left in the display for
visual feedback, if applicable. You may hold down , , or if you want to call
several functions in sequence showing the same color.

Time for a little problem solving example. Turn your WP 34S on again if necessary (it
may have shut down automatically in the meantime). Anyway it will still show its last
display

5
.

Now let us assume you want to fence a little rectangular patch of land, 40 yards long
and 30 yards wide. 6 You have already set the first corner post (A), and also the
second (B) in a distance of 40 yards from A. Where do you set the third and fourth
post (C and D) to be sure that the fence will form a proper rectangle? Simply key in

This key separates the two numbers in


input here.7

Note is reached by pressing


+ top right on the keyboard.

So, just get 80 yards of rope, nail its one end on post A and its other end on B, fetch
the loose loop and walk 30 yards away. When both sections of the rope are tightly
stretched, stop and place post C there. You may set post D the same way.

This method works for arbitrary rectangles: whatever other distances may apply in
your case. As soon as you press , your WP 34S does the necessary calculation
of the diagonal automatically for you. You just provide the land, posts, rope, hammer
and nails. And it will be up to you to set the posts!

5
If your WP 34S fails to show this display for any reason whatsoever (e.g. because you played around with it
a bit in the meantime), you will get it by sliding the battery cover open, locating the little hole below the label
RESET, and actuating the button behind it using a suitable pin. Then close the cover again and press .
6
This manual is written for an international readership, and we very well know the SI system of units agreed
on internationally and adopted by almost all countries on this planet. Despite this fact, we use (old British)
Imperial units here so our US-American readers can follow. But the example will work with meters as well.
7
Note the number 40 being adjusted to the right and a radix mark added. This indicates input being closed
for this number.
Generally, we shall quote only numeric displays in the following for space reasons, using the proper font.
And for better readability on paper, we refer to keyboard labels using dark print on white like e.g. or
, omitting the prefix for the latter since redundant by color print. Also we will use points as radix
marks, although significantly less visible than commas, unless specified otherwise explicitly. By experience,
“comma people“ seem to be more capable of reading radix points and interpreting them correctly than vice
versa.
WP 34S Owner’s Manual Edition 3.1 Page 12 of 211
As you have found on the keyboard next to two other labels showing arrows as well,
the labels on your WP 34S are generally grouped according to their purposes. Generally,
functions and their inverses – if existent – are placed next to each other. If a function is
called using , its inverse is called using . Besides
the keys for numeric input ( , , , …, , , , , , and ),
the three prefixes explained so far ( , , and ), and
the four elementary arithmetic operations ( , , , and ),
there are five larger groups of labels as outlined below:

Stack and
register
Mathematical operations
functions typ-
ically working
with real Integer
numbers and logic
operations

Probability
and statistics

Program-
ming

Getting used to the groups, you will very soon find your functions on your WP 34S easily.

Most of the 168 labels printed on it point to operations carrying simply the same name.
, for instance, calls the function ALL and calls FIX. In the example above,
however, pressing called the function POL – there are a few special labels like this.
Let us introduce them to you, starting top left on the keyboard:
WP 34S Owner’s Manual Edition 3.1 Page 13 of 211
1. , , , and are named hotkeys – calling directly the user programs carrying
these labels. Take your chance to make up to four of your favorite functions directly
accessible. Unless the respective labels are defined, these keys act as
, , , or , respectively, as is printed above them.

2. (i.e. + ) is the
prefix for hyperbolic functions
SINH, COSH, and TANH;
(i.e. + ) is the
prefix for their inverses ASINH,
ACOSH, and ATANH. Similar-
ly, stands for ASIN, etc.
3. is a prefix for direct con-
version of the value displayed
(i.e. x). It may be trailed by
, , , , or
(the respective function
names then read e.g. H.MS).
trailed by , , or
will display x converted to an
integer of the respective base.
is also used for indirect
register access (see below).
converts polar to rect-
angular coordinates in a plane,
converts vice versa. So
this pair covers the two classic
coordinate transformations
(see REC and POL).

4. is for calling complex operations (see p. 29). and enter fraction


mode for proper or improper fractions (see PROFRC and IMPFRC).
5. and represent the classic two time modes, where the latter stands for
decimal hours and also for decimal floating point numbers in general (see DECM).
6. sets alpha mode (see p. 61)., while , , , or set integer modes for
calculating with binary, octal, decimal, or hexadecimal numbers (see p. 52).
7. returns the logarithm for base 10, does the same for base 2 (see LOG…).
8. calls x! in all numeric modes. and (calling COMB and PERM here)
work like on an HP-15C. calls CORR, calls ABS, and calls ROUND.
9. There are three toggles: for radix marks, for programming (like on an
HP-15C), and for upper and lower case in alpha mode.

These are all the special labels featured. A complete list of each and every command
provided on your WP 34S, the keystrokes calling it, and the necessary individual
explanation is in the Index of Operations (IOP) below for your reference.
WP 34S Owner’s Manual Edition 3.1 Page 14 of 211
Let us return to our introductory example for two remarks8:

1. There is no need to enter units in your calculations. Just stay with a consistent set of
units and you will get meaningful results within this set.9 If you want to convert results
from one unit to another, however, see the catalog CONV described further below.

2. Although we entered integer numbers only for both sides of our little ground, your
WP 34S calculated the diagonal in default floating point mode. This allows for
decimal fractions of e.g. yards in input and output as well. Another mode lets you
enter proper fractions such as e.g. 6 ¼ where you need them. Your WP 34S features
more modes – we will introduce them to you from p. 35 on). Before, let us show you
some ways to enter numbers in your WP 34S and to further deal with them in typical
calculations .

Entering Numbers

… is as easy as typing. For 12.34, for


example, press . Any digit
mistyped may be deleted immediately by
and entered correctly thereafter.

For negative numbers such as –5.6 , press


or or –
the changes the sign of the number put
in.

For entering really big stuff such as the


age of the universe as we know it, do it this
way: resulting in 13.8
with nine digits trailing the point, i.e.
13 800 000 000 years. Really tiny numbers
such as the diameter of an atom
(0.000 000 000 1m – ten zeroes heading
the 1) are entered in full analogy:
correspond to 1∙10-10 m.

Any numeric input will just fill the display and is interpreted when completed, not earlier.

8
Generally, we assume you have graduated from US High School at minimum, passed Abitur, Matura, or an
equivalent graduation. So we will not explain basic mathematical rules and concepts here.
And in four decades of scientific pocket calculators, a wealth of funny to sophisticated sample applications
has been created and described by different authors – more and better than we can ever invent ourselves.
It is not our intention to copy them. Instead, we recommend the DVD mentioned above once again:
it contains nearly all the user guides, handbooks, and manuals published for vintage Hewlett-Packard
calculators beginning with their very first, the HP-9100A of 1968. Be assured that almost every calculation
described there for any scientific calculator can be done significantly faster on your WP 34S – and often
even in a more elegant way.
9
The big advantage of SI is that it is the largest consistent set available.
WP 34S Owner’s Manual Edition 3.1 Page 15 of 211
Real Calculations

Most of the commands your WP 34S features are mathematical operations or functions
taking and returning real numbers such as 1 or -2.34 or or 5.6E-7. Note that integer
numbers such as 3, 10, or -1 are just a subset of real numbers.

Many real number functions provided operate on one number only. For example,

enter 0.49
2
and press . You will get 0.7 since 0.7 = 0.49 .

That is easy, isn’t it? Generally, such functions replace x (i.e. the value displayed) by its
result f(x) . In this example, f(x) .

The vast majority of calculators works this way, so this is no real surprise.

Some of the most popular mathematical functions, however, operate on two numbers
instead. Think of + and –, for example.

Example: Assume having an account of 1,234 US$ and taking 56.7 US$ away from
it. What will remain? One easy way to solve such a task works as follows:

On a piece of paper On your WP 34S

Write down the 1st number: 1234 Key in the 1st number: 1234
st nd
Start a new line. Separate 1 from 2 :
nd
Write down the 2 number: 56.7 Key in the 2nd number:
56.7

Subtract: 1177.3 Subtract: 1177.3

This is the essence of RPN:


Provide the necessary operands, then execute the requested operation.

And a major advantage of RPN compared to other entry systems for calculators is that it
sticks to this basic rule. Always.10

As the paper holds your operands before you calculate manually, a place holding your
operands on your WP 34S is required. The stack does that. It will also take care of
intermediate results, if applicable, as your paper may do. Turn overleaf to see how this is
done.

10
Some people claim this being true for RPL only. RPL is a language developed from RPN in the 1980’s.
Maybe they are even right. In my opinion, however, RPL strains the underlying postfix principle beyond the
pain barrier, exceeding the limit where it becomes annoying for the human brain. Not for everybody, of
course, but also for many scientists and engineers. Thus we decided to stick to RPN on the WP 34S.
WP 34S Owner’s Manual Edition 3.1 Page 16 of 211
Elementary Stack Mechanics

Think of the stack like a pile of registers 11 : bottom up, Stack register content
they are traditionally named X, Y, Z, T, optionally D d
followed by A, B, C, and D on your WP 34S. New input is C c
always loaded in X, and only its content x is displayed.
B b
separates two input numbers by closing the A a
number x and copying it into Y 12, so X can take another T t
numeric input then without losing information.
Z z
After having completed the second numeric input in the Y y
little account example above, subtracts x from y and Display X x
puts the result f(x, y) = y – x into X for display. This
method applies for most two-number real functions.

A large part of mathematics is covered by two-number functions of this kind, and


combinations of such operations. Let us look at a chain calculation, for example

This is as a combination of six two-number functions: two subtractions, an addition, a


product, an exponentiation, and a division. And this is how it is solved on your
WP 34S, starting top left, and what happens on the stack during the solution: 13

T
Z
Y 12.3 12.3
X 12.3 12.3 45.6 -33.3
Input

You will have recognized that the first parenthesis was solved exactly as shown in
our little account example above. Now proceed to the second parenthesis:

T
Z -33.3 -33.3
Y A -33.3 78.9 78.9 -33.3
X 78.9 78.9 1.2 80.1 -2,667.33
Input

11
Learn more about the registers provided by your WP 34S in next chapter.
12
This is the classic way ENTER worked from the HP-35 of 1972 until the HP-42S ceased in 1995. It is often
said ENTER ‘pushes x on the stack’. In doing so, the higher stack contents are lifted out of the way before.
So z goes into T and y into Z before x goes into Y. See page 20 for detailed stack pictures.
The HP-30b employs a different ENTER – the WP 34S sticks to classic RPN, however.
13
There may be data loaded in the higher stack levels already. They are from previous operations and are
not relevant for this calculation, so leave them aside here.
WP 34S Owner’s Manual Edition 3.1 Page 17 of 211
Note the result of the first parenthesis was lifted automatically (A) to Y to avoid
overwriting it when the next number was keyed in step 1 of this row. This is called
automatic stack lift and is standard in RPN calculators.14
And after having solved the second parenthesis in step 4 of row 2, we had the results
of both upper parentheses on the stack – so everything was ready for multiplication
to complete the numerator and we did it.
Now we will simply continue and start calculating the denominator:

T
Z -2667.33 -2667.33 A -2667.33
Y A -2667.33 3.4 3.4 -2667.33 A -2.2 -2667.33
X 3.4 3.4 5.6 -2.2 7 -249.43…
Input

Last job remaining is the final division of numerator by denominator. Both are on the
stack in the right order. Just press and see the result: 10.6934534648 .

As you have observed several times now, the contents of the stack registers drop when a
two-number function is executed. Like the automatic stack lift mentioned above, this stack
drop affects all levels: x and y are combined giving the result f(x, y) loaded into X, then z
drops to Y, and t to Z. Since there is nothing available above for dropping, the top stack
level content will be repeated here. You may employ this top level repetition for some nice
tricks. See the following compound interest calculation, for example:

Assume the bank pays you 3.25% p.a. on an amount of 15,000 US$; what would be
your account status after 2, 3, 5, and 8 years? You are interested in currency values
only, so set the display format to 2 for this. It causes the output being shown
rounded to next cent (internally, the numbers are kept with far higher precision).15

T 1.0325 1.0325 1.0325 1.0325 1.0325 1.0325


Z 1.0325 1.0325 1.0325 1.0325 1.0325 1.0325
Y 10.69… 1.0325 1.0325 1.0325 1.0325 1.0325 1.0325
003
X 1.0325 1.03 15 15990.84 16510.55 17601.17 19373.66
Input 1.0325 15 3
after 2 …3 …5 …8
years years years years

Here, each multiplication consumes x and y for the new product put in X, followed by
z dropping to Y, and t to Z. Due to top level repetition the interest rate is kept as a
constant on the stack, so the accumulated capital value computation becomes a
simple series of strokes.

14
For a four level stack, the full automatic stack lift procedure is moving z to T, then y to Z, and finally x to Y.
The old content of the top level T is overwritten. So automatic stack lift affects all stack levels. It will not be
indicated after this example anymore. In fact it is worth mentioning when automatic stack lift is disabled
since this is under fixed conditions only and occurs far rarer.
15
In the following, we will use plain text for numeric input for space reasons, unless mentioned otherwise.
WP 34S Owner’s Manual Edition 3.1 Page 18 of 211
Debt calculations are significantly more complicated – so avoid debts whenever
possible! In the long run, it is better for you and the economy. Nevertheless, you can
cope with those calculations using your WP 34S as well (see further below).

There are also a few three-number real functions featured by your WP 34S (e.g.
DATE and %MRR) replacing x by the result f(x, y, z) . Then t drops into Y and so on,
and the content of the top stack level is repeated twice.

Some real functions (e.g. DECOMP or DATE ) operate on one number but return two or
three. Other operations (such as RCL or SUM) do not consume any stack input at all but
just return one or two numbers. Then these extra numbers will be pushed on the stack,
taking one level per real number.

In addition to the traditional stack and register control


operations , , , , , LASTx,
CLSTK, , , and – known for decades
and mostly found within the blue frame on the
keyboard – your WP 34S also features , DROP,
RCLS, STOS, SSIZE4, SSIZE8, SSIZE?, Y⇆, Z⇆, T⇆,
and ⇆. Turn overleaf to learn what , ,
DROP, , , , and LASTx do with the stack
if SSIZE4 or SSIZE8 is set. See the IOP for information
about the other commands.

For the first time ever in a calculator, your WP 34S offers you a choice of four or eight
stack levels (see SSIZE4 and SSIZE8). Thus, the fate of stack contents depends not only
on the particular operation executed and its domain but also on the stack size chosen.
Real functions in a four-level stack work as known for decades. In the larger stack of your
WP 34S everything works alike – just with more levels for intermediate results. Turn
overleaf for details.

WP 34S Owner’s Manual Edition 3.1 Page 19 of 211


Assumed Stack contents after executing … … functions of
stack contents … the stack register operations … one … two

Level
at the number numbers
beginning: ENTER FILL DROP x y R R LASTx such as x2 such as /
T t = 44.4 33.3 11.1 44.4 44.4 11.1 33.3 33.3 44.4 44.4
With 4
stack Z z = 33.3 22.2 11.1 44.4 33.3 44.4 22.2 22.2 33.3 44.4
levels Y y = 22.2 11.1 11.1 33.3 11.1 33.3 11.1 11.1 22.2 33.3
X x = 11.1 11.1 11.1 22.2 22.2 22.2 44.4 last x 123.21 2

D d = 88.8 77.7 11.1 88.8 88.8 11.1 77.7 77.7 88.8 88.8
With 8
stack C c = 77.7 66.6 11.1 88.8 77.7 88.8 66.6 66.6 77.7 88.8
levels B b = 66.6 55.5 11.1 77.7 66.6 77.7 55.5 55.5 66.6 77.7
A a = 55.5 44.4 11.1 66.6 55.5 66.6 44.4 44.4 55.5 66.6
T t = 44.4 33.3 11.1 55.5 44.4 55.5 33.3 33.3 44.4 55.5
Z z = 33.3 22.2 11.1 44.4 33.3 44.4 22.2 22.2 33.3 44.4
Y y = 22.2 11.1 11.1 33.3 11.1 33.3 11.1 11.1 22.2 33.3
X x = 11.1 11.1 11.1 22.2 22.2 22.2 88.8 last x 123.21 2

Using the stack, RPN makes all parentheses such as , , , , , or completely unnecessary in calculations. There is no
operator precedence. Here is another example showing a slightly more complicated formula and the keystrokes used for solving it:

7 4 5 6
6.5 5.9 3 7 1.7
3.5 complete denominator
7.6 .8 30 7 4
6 5 5.1
.3 1 complete numerator
complete result (0.37)

WP 34S Owner’s Manual Edition 3.1 Page 20 of 211


Even solving this formula requires only four stack levels as indicated by the colors above.
Note there are no pending operations – each operation is executed individually, one at a
time, allowing perfect control of each and every intermediate result. That is another
characteristic advantage of RPN.

Calculating such a formula from inside out stays a wise strategy. If you had started with
the numerator of that sample formula straight ahead, you would have needed five levels
for its complete solution. With eight levels as provided here, you will be on the safe side
even with the most advanced equations you compute in your life as a scientist or engineer.
Let your WP 34S do the arithmetic while you do the mathematics!

Error Recovery

Nobody is perfect – errors will happen. With RPN, however, error recovery is easy even in
long calculations since your WP 34S loads x into the special register L (for Last x) auto-
matically just before a function is executed.16 How does this help in real life?

1. If you got an error message in response to your function call, press to erase that
message and you will return to the state before that error happened. Now do it right!

2. If you have erroneously called a wrong one-number function (or ), just press
17
and to undo it. These two steps restore the stack exactly as it was
before the error happened – then resume calculating where you were interrupted.

3. Recovering from wrong two-number functions calls requires three steps.

Example: Assume – while you were watching an attractive fellow student or


collaborator – you pressed inadvertently instead of in the second last step
of the example on previous page. Murphy’s law! Do you have to start the
calculation all over now? No, that error is easily undone as follows:

T older stuff older stuff older stuff older stuff older stuff older stuff older stuff
Z old stuff older stuff old stuff older stuff old stuff older stuff older stuff
Y numerator old stuff num * den old stuff num old stuff old stuff
X denominator num * den den num den num/den 0.37
Input
Fine so far. Oops! 1 2 3 Resume

In step 1, recalls the complete denominator, the last x before the error.
Step 2 undoes the erroneous operation by executing its inverse.
Finally, step 3 regains the stack exactly as it was before the mistake. Now you can
simply resume your calculation and you will get the correct complete result.

16
There are only a few commands changing x but not loading L. They are mentioned explicitly in the IOP.
17
Note there is a grey L printed bottom left of the key . Nevertheless, we will quote these keystrokes as
instead of for reasons becoming clear in the next section.
WP 34S Owner’s Manual Edition 3.1 Page 21 of 211
An erroneous may be undone the same way:

T older stuff older stuff older stuff older stuff older stuff
Z old stuff older stuff old stuff older stuff old stuff
den
Y num old stuff num old stuff num
den
X den num den num den
Input
Oops! 1 2 3

… but where do we find ? This problem will be solved overleaf.

More advanced two-number functions may require more effort for error recovery.
Having called x inadvertently can be undone, too, but needs an additional
auxiliary register:

T older stuff older stuff older stuff older stuff older stuff older stuff older stuff
Z old stuff older stuff old stuff old stuff old stuff older stuff old stuff
Y num old stuff LOGdnnm LOGdnnm den old stuff num
X den LOGdnnm den den LOGdnnm num den
Input x 00 00

Oops! 1 2 3 4 5

Generally, this kind of recovery procedure requires that the inverse of the erroneous
operation is provided.

4. Inadvertent pushes on the stack can be undone by executing DROP once (for
, , , , and alike) or twice (e.g. for or ). You will lose the
contents of the top (two) stack level(s) by this mistake, however, so set your stack to
eight levels to minimize the effects.

Calling Commands Unseen on the Keyboard

Your WP 34S features more than 600 different commands. 168 labels are printed on the
keyboard. So how do you learn about the other commands? And when you know their
names, how can you call them?

The answer to the first question is most easy: read! The IOP contains everything.

The answer to the second question is less obvious, but easy as well: the ‘hidden
commands’ are stored in catalogs. Remember labels underlined point to such collections
of commands. You will find such labels on your WP 34S on the slanted fronts of ten keys.
For example, + points to , the largest catalog provided.

WP 34S Owner’s Manual Edition 3.1 Page 22 of 211


Example: Assume you want to learn about DECOMP – it was mentioned on p. 19.
You return to the Table of Contents above, look for the Index of Operations, jump
directly to the letter D therein, and look up DECOMP to get the necessary information
where it lives and what it does. OK, enter 0.375 and check it – calling such a
command is even easier than looking it up:

You have read DECOMP is stored in .


So just key in … and your WP 34S displays …
↕ ³√x being the first command stored in .
↕ DATE for obvious reasons.
↕ DATE→
↕ DAY
↕ DAYS+
↕ DECOMP voilà! Now execute it …

since 0.375 = 3 / 8 . Expected that.


But what is this:
0.46875 0.46875 ?
↕ DECOMP oh, good, that function is memorized !

meaning 0.46875 = 15 / 32 .

As browses forwards in a catalog, does backwards. There is a more elegant


method for calling cataloged commands described further below, but this here suffices and
will do for the time being.

And answering the question of last section, is stored in as well. Just key in
and you get ↕ ᵡ√y .

Now you know sufficient about calling commands and real calculations on the stack.

There are, however, far more places in your WP 34S where numbers may be stored and
saved, and there are also more objects than just real numbers your WP 34S can handle
for you. Let us show them to you.

WP 34S Owner’s Manual Edition 3.1 Page 23 of 211


ADDRESSING OBJECTS Special registers / General purpose Program User flags
stack registers steps
D * K *** R255 000 00
C * J *** R254 001 01
This picture shows the
B * R253 002 02
complete address space
Mode A * … … …
of your WP 34S. Depend-
T
ing on the way you con-
figure its memory, a sub- Alpha (30 B) Z … … …
set of all these addresses Y I ** R128 925 99
will be accessible. Display X L = LASTx R.15 = R127 926 X = 100
… 927 Y = 101
Z = 102
Offering two stack sizes, either T or D will be the top stack level. A - D will be … T Tracing
allocated for the stack if required. I will be loaded in addition to L in complex R.01 = R113 K = R111 A ‘=’
calculations (see p. 29). Registers J and K may carry parameters of statistical R.00 = R112 J = R110 B Overflow
distributions. Unless required for the purposes mentioned, A, B, C, D, I, J, and I = R109 C Carry
K are available as additional general purpose registers.
L = R108 D Danger
For numeric addressing of registers and flags, see the tables on pp. 26 and R99 D = R107 L = 108
33. Addresses ≥112 are used for local registers and flags (see p. 69). R98 C = R106 I = 109
… B = R105 J = 110
Some user flags have special use or effects. Flag A lights the big ‘=’ symbol in
display. The system sets flags B (‘big’) and C in integer modes (see p. 52ff) A = R104 K = 111
like the overflow and carry bits of the HP-16C – some integer operations also … T = R103 .00 = 112
read flag C. Flag D allows for special (i.e. infinite or non-numeric ‘NaN’) R04 Z = R102 …
results without getting an error, flag T sets the print line to ‘tracing mode’ – R03 Y = R101
both flags are user-settable, the system only reads those flags.
R02 X = R100 …
Note you will not get 928 program steps and 256 registers and 128 flags all R01 .14 = 126
together at the same time – see Appendix B for memory management. R00 .15 = 127

WP 34S Owner’s Manual Edition 3.1 Page 24 of 211


The Virtual Keyboard for Addressing – Transient Alpha Mode (αT)

During input processing in memory addressing, e.g. while entering parameters for
storing, recalling, swapping, copying, clearing, or comparing, you will need far less than
the 168 labels presented on the keyboard. Just 34 keyboard labels will do instead. The
calculator mode supporting exactly these 34 labels (and no more) is called transient
alpha mode. As shown in examples on the next two pages, it may be (temporarily) set in
cases as outlined above. Entering αT mode, the keyboard is automatically reassigned to
work as pictured here:

MODE
A B C D
This kind of picture is called a
virtual keyboard since it
I deviates from the physical (or
real) one of your WP 34S.
ENTER In such a picture, dark red
background is used to high-
J K L light changed key function-
7 8 9 / ality. White print denotes
primary functions also on
virtual keyboards, such as
▲ 4 5 6 × the top left key entering the
letter A in αT mode directly.
T* On the other hand, what is
▼ 1 2 3 – printed white on your physical
WP 34S is called a default
primary function.

EXIT 0 ● +
X* Y Z*

Note all keys are primary in αT mode – no shift keys needed. This allows for fast and
easy input of a limited character set. So you can reach all register addresses available
with a minimum of keystrokes.

Special rules may apply for T, X, and Z here – see two pages below.

αT mode will be terminated or left (returning to the mode set before) as soon as sufficient
characters are put in for the respective step. You may delete pending input (character by
character) using or just abort the pending command by – the latter will leave
αT mode immediately.

WP 34S Owner’s Manual Edition 3.1 Page 25 of 211


Addressing Real Numbers

1 User
, , x< ?, x≤ ?, x≈ ?, x≥ ?, or x> ?
input

Echo OP _ ? (with mode set),


e.g.
x>_?

2 User or Stack level or 18

input lettered register leaves mode opens indirect


, , ..., addressing

Echo OP n ? OP? x OP? _ OP? _


e.g. e.g.
x≤0? x≥? Y
3 User Compares x with the Compares x with the Register number See next page for
input real number 0. number on stack more about indirect
level Y.
… , addressing.

,
if the respective
register is allocated.

Echo OP? nn
e.g.
x≠? 23
Compares x with
the number stored in
R23.

18
You may skip this keystroke for register addresses >19 or local registers. The latter start their address
with a – see the section about programming and Appendix B below.

WP 34S Owner’s Manual Edition 3.1 Page 26 of 211


1 User , , RCLS, STOS, , , , , , y⇆, z⇆, t⇆,
input , , DSL, DSZ, ISE, ISZ,
, , , , DISP, BASE, bit or flag commands, etc.

Echo OP _ (with mode set), e.g. RCL _19

2 User Stack level or lettered Register or flag number


input register20 or number of bit(s) or opens indirect addressing
, , ,…, decimals (see below for
valid ranges)

Echo OP x OP nn OP _
e.g. SF K e.g. SCI 10

3 User Sets flag 111.


input
Stack level or lettered Register number
register … ,
, , , ... , … ,
if the respective register is
allocated.

Echo OP x OP nn
e.g. VIEW→L e.g. STO→45

Shows the content of the Stores x into the location


register where L is pointing to. where R45 is pointing to.

Type Number range21 (some more registers and flags carry letters)

}
Registers 0 … 99 for direct addressing of global numbered registers upper limits
.0 … .15 for direct addressing of local registers depend on
0 … 255 for indirect addressing (≤111 without local registers) allocation
Flags 0 … 99 for direct addressing of global numbered flags
.0 … .15 for direct addressing of local flags if allocated
0 … 127 for indirect addressing (≤111 without local flags)
Decimals 0 … 11
Integer bases 2 … 16
Bits 0 … 63, word sizes up to 64 bits

19
For and , any of , , , , , or may precede step 2, except in and
. Entering the operator twice will remove it, e.g. equals .
Note calls ENGOVR and calls SCIOVR. See the index of operations.
20
Exceptions: , , , , DISP, and BASE accept lettered registers in indirect addressing only.
Else, specifying register X as well as RCL T, STO T, RCL× T, STO× T, RCL Z, STO Z, RCL+ Z and
STO+ Z require an heading the letter, e.g. for the latter.
21
For short numbers, you may key in e.g. instead of .

WP 34S Owner’s Manual Edition 3.1 Page 27 of 211


Real Matrices and Vectors

Numbers arranged in table-like grids are called matrices by mathematicians.


If you do not know of matrices yet, feel free to set them aside – you can use your
WP 34S perfectly without them.

If you know of matrices, however, note your WP 34S features a set of operations for
adding, multiplying, inverting and transposing matrices, as well as for manipulating rows
in such matrices. In general, the respective commands are building blocks designed to
provide the low level support routines for creating more useful matrix functions in the
form of keystroke programs. I.e. they represent the basic linear algebra subprograms of
the WP 34S matrix support. On the other hand, your WP 34S also provides functions for
computing determinants or for solving systems of linear equations.

A matrix is represented within your WP 34S by its descriptor, formatted bb.rrcc with
rr being the number of rows and
cc the number of columns it features. Thus this matrix has rr × cc elements.
These elements are stored in consecutive registers starting at base address |bb| .

Example: A descriptor 7.0203 represents a 2×3 matrix – let us call it (M). As


you know, its six elements are arranged in two rows and three columns, and
they are numbered as follows:

The matrix descriptor 7.0203 tells you where the values of these elements are
stored:

Depending on the current contents of these registers, the actual matrix may
look like this:
, for example.

If cc is omitted in a descriptor it defaults to rr so a square matrix is assumed. For


example, a descriptor 13.0400 belongs to a 4×4 matrix with its elements stored in R13
through R28. The maximum number of matrix elements is 100 – it is the number of
general purpose registers available for such a task.
See the IOP and the catalog MATRIX for all commands featured.

ATTENTION: Your WP 34S cannot know whether a particular real number is a


matrix descriptor or a plain number. It is your task to take care of that.

A vector may be regarded as a special case of a matrix featuring either one row or one
column only. Thus, a vector descriptor looks like bb.01cc or bb.rr01 . Library
routines are readily provided for 3D vector calculus.

WP 34S Owner’s Manual Edition 3.1 Page 28 of 211


If you just want to do vector operations in 2D, there are simple alternatives (known for
long from earlier calculators) to full-fledged descriptor controlled computations: enter the
Cartesian components of each vector in X and Y (e.g. by converting its polar
components into Cartesian ones by , if necessary) and choose one of the following
alternative opportunities:
1. use for additions or for subtractions and recall the result via SUM, or
2. calculate in complex domain (see next paragraph). Therein, vector multiplication
is possible, too, using the commands CDOT or CCROSS.

Turn to a good textbook covering linear algebra for more information.

Complex Calculations

Mathematicians know of more complicated items than real numbers. There are also
complex numbers. If you do not know of them, leave them aside – you can use your
WP 34S perfectly without them.

If you know of complex numbers, however, note your WP 34S supports many operations
in complex domain as well. The key is employed as a prefix for calling complex
functions. E.g. calls the complex cosine, and it is displayed and listed as
C
COS (the elevated C is the signature for complex functions in your WP 34S).

ATTENTION 1: All functions operating on complex numbers require them in


Cartesian coordinates exclusively on your WP 34S. Each such number takes
two adjacent registers22: the lower one for its real part and the higher one for
its imaginary part.
You may use to convert xc = x + i ∙ y to its polar equivalent xc = r ∙ e i φ any
time you like – just remember reverting this by before starting a complex
calculation.

ATTENTION 2: All complex functions work with radians exclusively where an


angular input is required, and return radians where they output angular data.
As a reminder, RAD is lit when is pressed. The user is responsible for
providing suitable input values; there will be no automatic conversion of angular
data if the angular mode is switched.

Generally, if an arbitrary real function f operates on …


… one real number x only, then its complex sibling Cf will operate on the
complex number xc = x + i ∙ y . Note xc will consume two stack levels.
C
… one register, e.g. R12, then f will operate on two registers, e.g. R12 and R13.
C
… x and y, then f will operate on x, y, z and t .

22
The HP-42S supported a special data type for complex numbers. This is not viable using the hardware
of the HP-30b, however. See Appendix F for the reasons.

WP 34S Owner’s Manual Edition 3.1 Page 29 of 211


Where one-number real functions replace x by the result f(x) , one-argument complex
functions replace x by the real part and y by the imaginary part of the complex result
C
f(xc) . Higher stack levels remain unchanged. Such functions are e.g. C1/x, CABS, CFP,
C
IP, CROUND, Cx!, Cx2, C√x, C+/–, CΓ, the logarithmic and exponential functions with
bases 10, 2, and e, as well as hyperbolic and trigonometric functions and their inverses.

Two-number real functions replace x by the result f(x, y) as shown above. In complete
analogy, two-argument complex functions replace x by the real part and y by the
imaginary part of the complex result Cf(xc, yc) . The next stack levels are filled with the
complex contents of higher levels, and the complex number contained in the top two
stack levels is repeated as shown overleaf. Such complex functions are the basic
arithmetic operations in complex domain as well as CIDIV, CLOGX, Cyx, Cx√y, Cβ, and C||.
Turn to the stack diagrams overleaf for further details.

Where a complex operation (such as CRCL) does not consume any stack input at all but
just returns a complex number, this number will be pushed on the stack taking two
levels.

Whenever a complex result is displayed, a capital C is lit in the upper left corner of the
LCD reminding you to look also at stack level Y at least.

WP 34S Owner’s Manual Edition 3.1 Page 30 of 211


Calculating with complex numbers uses two registers or stack levels for each such number as explained above and shown here:

Assumed Stack contents after executing … … complex functions of


stack contents … the complex stack register operations … one
address
at the … two
number
Level
beginning: numbers
C
ENTER C
FILL C
DROP C
x y C
R C
R C
LASTx such as
C 2 such as C/
x

With 4 T Im(yc) = Im(tc) Im( xc ) Im( xc )


stack
yc = tc xc xc yc = tc yc = tc
Z Re(yc) = Re(tc) Re( xc ) Re( xc )
levels
Y Im( xc ) Im( xc ) Im( yc ) Im( (xc)2 ) Im( yc / xc )
yc yc last xc
X Re( xc ) Re( xc ) Re( yc ) Re( (xc)2 ) Re( yc / xc )

With 8 D Im( tc )
stack zc xc tc tc xc zc zc tc tc
C Re( tc )
levels
B Im( zc )
yc xc tc zc tc yc yc zc tc
A Re( zc )
T Im( yc )
xc xc zc xc zc xc xc yc zc
Z Re( yc )
Y Im( xc )
xc xc yc yc yc tc last xc (xc)2 yc / x
Re( xc ) c
X

So, an 8-level stack gives you the same flexibility in complex domain you are used to with a 4-level stack in real domain. See the IOP
for all commands supported in complex domain. Many of them are contained in the complex X.FCN catalog.
You can use complex domain for 2D vector algebra as well. The functions CABS, C+, C–, CCROSS, CDOT, and CSIGN wait for you.

WP 34S Owner’s Manual Edition 3.1 Page 31 of 211


1/x yX √x After pressing ,
your WP 34S allows for
CPX the complex operations
shown on this virtual
HYP –1 SIN –1 COS –1 TAN –1 keyboard (but STOPW
f g h represents an applica-
STO RCL R
R tion in real domain, see
p. 145).

ENTER x y +/–
CONST x
FILL
Note that a second
7 8 9 / , a , or
directly after will
e X LN 10 X LG 2X LB y X LOGX 1 /x || just cancel the prefix
4 5 6 , so you return to
× the default keyboard
!
Cy,xPy,x √x x 2 assignments.

1 2 3 –
X.FCN
x=≠ ? Constants in complex
domain and such cal-
EXIT i + culations occupy two
registers like all other
I x I RND IP FP STOPW complex numbers!

Complex constants are simply entered as imaginary_part real_part .

Pure real constants, identified by a zero imaginary part, are easily put in for complex
calculations as follows:
real_constant , or alternatively
n for integers 0 < n ≤ 9 only.
In programming, # n with 0 ≤ n ≤ 256 will save steps.

Example: The real number may be loaded via or


– both will result in y = 0 and x = .

Pure imaginary constants, on the other hand, having a zero real part, are entered thus:
imaginary_constant .

Example: The complex unit i may be loaded via or


– both will result in y = 1 and x = 0 .

Compare the stack mechanics shown on previous page.

WP 34S Owner’s Manual Edition 3.1 Page 32 of 211


Addressing Complex Numbers

1 User or
input

Echo OP _
(with mode set)
e.g.
©x=_?

2 User or Stack level or 23

input lettered register leaves mode opens indirect


, , , addressing.
, or

Echo OP n ? OP? x OP? _ OP? _


e.g. e.g.

3 User Compares x + i y Compares x + i y Register number See next page for


input with the real with z + i t . … , more about indirect
number 0. addressing.

,
if the respective
register is allocated

Echo OP? nn
e.g.
©x≠? 26
Compares x + i y
with r26 + i r27 .

23
You may skip this keystroke for register numbers >19 or local registers. The latter start with a – see
the section about programming and Appendix B below.
WP 34S Owner’s Manual Edition 3.1 Page 33 of 211
1 User
, , or
input

Echo OP _
(with mode set)
e.g.
©RCL _24

2 User Stack level or lettered Register number


input register … , opens indirect addressing,
25
, , , , or … , generally working as
if the respective register is in real domain.
allocated.

Echo OP x OP nn OP _
e.g. e.g.
©RCL L ©STO 18
C
3 User This is LASTx –
input the real part is recalled
from register L to X,
the imaginary part Stack level or lettered Register number
from I to Y. register … ,
, , ... , … ,
if the respective register is
allocated.

Echo OP x OP nn
e.g. e.g.
©x⇆→Z ©STO→45
Swaps x and the content of the register where z is pointing to, Stores x + i y into
and y with the content of the next one. 2 consecutive registers,
Assume e.g. x = 4, y = 3, z = 20, r20 = 5, r21 = 6, then you starting with the register
will find x = 5, y = 6, r20 = 4, r21 = 3 after x
c
Z. r45 is pointing to.

ATTENTION: A complex operation will always affect a pair of registers: the one
specified and the one following this. To avoid ambiguity we strongly recommend
storing complex numbers with their real parts at even register addresses always. This
is guaranteed if you specify an even register number in direct addressing.

24
For and , any of , , , or may precede step 2. See the index of operations.
25 C C C C
Exceptions: RCL Z, RCL+ Z, STO Z, and STO+ Z require an preceding , e.g.
for the latter.
WP 34S Owner’s Manual Edition 3.1 Page 34 of 211
YOUR WP 34S IN VARIOUS OPERATING MODES

The Display

The LCD is your window to your WP 34S – there you see what is going on and what the
results are. This display sports a total of 400 elements in three sections: numeric, dot
matrix and fixed symbols. The numeric section features a minus sign and 12 digits for the
mantissa, as well as a minus sign and 3 digits for the exponent. The dot matrix is 6 dots
high and 43 dots wide, allowing for some 7 to 12 characters, depending on their widths.
The fixed symbols on the top right side (except the big ‘=’) are called annunciators.

The ten annunciators are for indicating modes.

The numeric section in the lower part of the LCD is used for displaying numbers in
different formats, status data, or message parts. See the examples below for more.

The dot matrix section above is used for passing additional information to the user.
During command input, for example, the dot matrix displays the command chosen
until input is completed, i.e. until all required trailing parameters are entered. The
prefixes , , , and are shown until they are resolved ( goes
with by default). If you pressed any such prefix erroneously, recovery is as easy
as follows:
o = = = = = NOP
o = = = NOP
o = = NOP
o = =
= =
= =
In addressing, progress is recorded as explained in the tables above in detail. You
may edit or cancel such pending operations by or (see p. 119ff).

If two or more requests compete for display space, the priorities are:
1. error messages as described in Appendix C,
2. special information as explained below,
3. information about the modes your WP 34S is running in.

WP 34S Owner’s Manual Edition 3.1 Page 35 of 211


Recognizing Calculator Modes

The annunciators or specific characters either in the dot matrix or in the exponent section
indicate most modes and system states:

Indicator Set by Cleared by Explanation, remarks

à Lower case letters will be entered


in alpha mode (see p. 61).

INPUT , ON , OFF, Alpha mode (see p. 61)

Steady if debug mode is turned on


(see App. A and H) or while serial
I/O is in progress (see App. D).
= look at right look at right
Flashing when the stopwatch is
running (except in STOPW).
Else clear.

ã battery low battery voltage > 2.5 V


Your WP 34S will shut off when
voltage drops below 2.1V.

program pointer
BEG otherwise
at step 000
STO , Programming mode

flashes while a
RCL program is program stopped
running

RAD ,
Angular mode (see p. 42)
360 ,
almost every See p. 38 for handling of
RPN a temporary message
command temporary messages in general.

B__ Binary integer mode


3__ BASE 3 any other BASE setting
Respective integer mode (see p.
∙∙∙ ∙∙∙
53 for all of these modes)
7__ BASE 7 , , and
FRACT will set fraction
o__ Octal integer mode
mode (called FRC).
9__ BASE 9 Integer mode of base 9
, , , ,
d__ , , , Decimal integer mode
-1__ BASE 11 and TIME will set default
∙∙∙ ∙∙∙ floating point decimal Respective integer mode
-5__ BASE 15 mode (i.e. DECM).

h__ Hexadecimal integer mode

_c_ carry set, , carry clear Indicate the respective bits set in
integer modes (see p. 53).
__o overflow,

WP 34S Owner’s Manual Edition 3.1 Page 36 of 211


The indicators following below are all lit in the dot matrix exclusively:

Indicator Set by Cleared by Explanation, remarks


Transient signal of prefix pend-
© see p. 35
ing. RAD will be lit together with ©.

C complex result else See p. 29.

D DBLON DBLOFF See Appendix H.


see p. 35 Transient signal of prefix pending

Ⓖ , Angular mode (see p. 42)

ⓖ →
see p. 35 Transient signal of prefix pending

M.DY M.DY, SETUS
any other date or region
Y.MD, SETJPN, Date modes (see p. 39)
Y.MD setting
SETCHN

See p. 125ff for more about cata-


↕ entering a catalog leaving it
logs.

Defaults D.MY and DECM are not indicated. Radix marks and separators are seen in the
numeric output immediately, time modes (12h / 24h) in the time string. The numeric format
of fraction mode (FRC) is unambiguous as well. Check the examples shown on p. 40ff.

Some mode and display settings may be saved and recovered collectively by STOM and
RCLM. These are stack depth, display contrast, complete decimal display settings, trig
mode, choices for date and time display, parameters of integer and fraction mode, curve fit
model, rounding mode, and precision selected. STOM stores this information in the
register you specify. RCLM recalls such a register content and sets the WP 34S modes
accordingly.

ATTENTION: Ensure that you actually recall mode data – else your WP 34S may be
driven into very strange settings and it may cost you considerable effort to recover
from that unless you find your previous modes stored elsewhere! See Appendix H.

All keyboard input will be interpreted when the input is completed, according to the modes
set at input time.

WP 34S Owner’s Manual Edition 3.1 Page 37 of 211


Common Commands Returning Specific Displays: STATUS, VERS, ERR, etc.

Some common commands use the LCD in a special way. The respective operations are
listed below. Three of them (and more mentioned further below) present temporary
messages as defined here:

Whenever anything different from the actual contents of X in current mode is


displayed or any additional information is shown in the dot matrix, these extra data
are considered being a temporary message. This is further indicated by the
annunciator RPN turned off as mentioned on p. 36.
If such extra data are displayed outside of a catalog or browser, they will vanish with
the next keystroke. Pressing or will just clear the temporary message
returning to the normal display, any other key will be executed in addition.

Now here are the common commands delivering special information:

1. returns very useful information about current memory allocation and the
space available. It continues showing the user flags set. See p. 127 for a detailed
description of this browser.

2. VERS generates a temporary message showing you version and build of the firmware
running on your WP 34S.

If the basic firmware is installed, the build number


will follow the version number immediately.

If the timer firmware is installed, a ‘T’ will trail the


version number.

If the firmware for printing is installed, a printer


character will be displayed instead.

See in Appendix A how to install these different firmware packs.

3. ERR and MSG display the corresponding error message as a temporary message. See
Appendix C for more.

4. A few far-reaching commands (such as CLALL, for example) will ask you for
confirmation before executing. Answer (by pressing ) or (pressing ).
Also or will be read as , all other input will be ignored.

Further commands returning special displays and temporary messages are specific to
particular modes of your WP 34S and are thus covered in the sections following.

WP 34S Owner’s Manual Edition 3.1 Page 38 of 211


Floating Point Modes – 1: Introduction and Localisation

Floating point modes cover the ‘usual’ numbers you calculate with: decimal real or
complex numbers, fractions, measured values, times and dates. DECM is the startup
default mode of your WP 34S. Many commands apply exclusively to it and its sub-modes.

Set display preferences according to your region's practices at once using single
dedicated commands:

Command Radix Time Date27 JG28 Three digit Remarks


mark26 separators

SETCHN RDX. 24h Y.MD 1949 E3OFF Would require separators


every four digits.

SETEUR RDX, 24h D.MY 1582 E3ON SETEUR applies to South


America as well. It also
applies to Russia, Vietnam,
Indonesia, and South
Africa, but with deviating
JG’s.

SETIND RDX. 24h D.MY 1752 E3OFF Would require separators


every two digits over 105.
Applies also to Pakistan,
Bangladesh, and Sri
Lanka.

SETJPN RDX. 24h Y.MD 1873 E3ON

SETUK RDX. 12h D.MY 1752 E3ON Applies also to Australia


and New Zealand. 24h is
taking over in the UK.

SETUSA RDX. 12h M.DY 1752 E3ON

26
See http://upload.wikimedia.org/wikipedia/commons/a/a8/DecimalSeparator.svg for a world map of radix
mark use. Looks like an even score in this matter. Thus, the international standard ISO 31-0 allows either a
decimal point or a comma as radix mark, and requires a narrow blank as separator of digit groups to avoid
misunderstandings. The numeric display hardware of HP-20b or HP-30b, however, does not allow for
narrow blanks.
27
See http://upload.wikimedia.org/wikipedia/commons/0/05/Date_format_by_country.svg for a world map of
date formats used. The international standard ISO 8601:2004 states 24h for times, Y.MD for dates. This
combination is commonly used in East Asia.
28
This column states the year the Gregorian Calendar was introduced in the particular region, typically
replacing the Julian Calendar (in East Asia, national calendars were replaced in the respective years). Your
WP 34S supports both 1582 and 1752. See the index of operations for JG1582 and JG1752.
WP 34S Owner’s Manual Edition 3.1 Page 39 of 211
Floating Point Modes – 2: Displaying Decimal Numbers, Fractions, Times, etc.

You find nearly all


functions for floating
point format control
allocated to the top two
rows of keys on your
WP 34S.

1. For floating point decimal numbers, startup default allows displaying all digits as
long as they fit the display width; it will switch to SCIentific (i.e. mantissa plus exponent)
notation to avoid flowing over the display limits. This format is ALL 00, SCIOVR.
Besides ALL and SCI, there are two more numeric display formats, FIX and ENG. Their
effects can be most easily demonstrated and distinguished using an example:

Format ALL 00 FIX 4 SCI 4 ENG 4


Input
2 0
107.12345678 107.12345678 107.1235 1.0712 107.12
-2 -2 -3
2 1.86700472531 0.0187 1.8670 18.670

Within FIX, the radix mark will always stay at the FIXed position defined. The radix
2 2
mark floats in the other notations, where e.g. 1.0712 represents 1.0712∙10 , while
-3 -3
18.670 V would mean 18.67∙10 V = 18.67 mV. Within ENG, the exponent will
always be a multiple of three corresponding to the unit prefixes in SI – thus it is called
the ENGineer’s notation.

As soon as a number entry is completed, the mantissa will be displayed adjusted to the
right, the exponent to the left. Within the mantissa, either points or commas may be
selected as radix marks, and additional marks may be chosen to separate thousands.

Assume the display set to FIX 4 again. Key in 12345678 901 , and you get

or with E3ON 29.

Without these separators (i.e. E3OFF), the same number will look like this:

or

29
These separators may also be beneficial in fraction mode described below.
WP 34S Owner’s Manual Edition 3.1 Page 40 of 211
With ENG 3 and after , you will get

or

When the last operation executed has returned a complex result, C is lit in the top left
of the LCD pointing to the fact that you will find the result of this function in X and Y.

Floating point decimal numbers within may easily be entered


directly in default DECM. Within this range, your WP 34S calculates with 16 digits.

Values less than 10 398 are set to zero (see Appendix B for the reasons). For results
, error 4 or 5 will appear unless flag D is set (see Appendix C).

2. and display the full number x, i.e. all sixteen digits present internally in the
mantissa of standard floating point numbers (imagine a radix mark after the first digit),
and all digits of the exponent almost like in scientific display format. All this is shown in
one temporary message, working like the SHOW command in the HP-42S.

For example, 3 returns

reading 3.100 627 668 …∙ 101 .

3. Fraction mode works similarly to the one in the HP-35S or the HP-32SII. In particular,
DENMAX sets the maximum allowable denominator (see the IOP). Display will look like
in the examples below – fractions are adjusted to the left. If the fraction is exactly
equal, slightly less, or greater than the floating point number converted, =, Lt, or Gt is
indicated in the exponent, respectively. On your WP 34S, fraction mode can handle
numbers with absolute values less than 100,000 and greater than 0.0001. Maximum
denominator is 9999. Underflows and overflows will be displayed in the format set
before fraction mode was entered.

Example: Enter the following: and you will see:


↕ DENANY in catalog MODE.
initializes.
0 ↕ DENMAX
sets maximum precision.

3
since 1/3 > 0.333333333333.

since this is exact.

WP 34S Owner’s Manual Edition 3.1 Page 41 of 211


50 40625
This is exact as well.

Pressing results in

Now, enter for converting this into a proper fraction.30 You get

with a little hook left of the first digit shown. This indicates the leading integer
number is displayed incompletely – there are at least two digits preceding 47
but no more display space. Press (see the previous point) to unveil the
integer part of this proper fraction as 2247.

Numeric input in fraction mode is straightforward and logically coherent. If you enter
two points, the first is interpreted as a space, the second as a fraction mark:

Key in: and get in proper fraction mode:

12 3/4

1 1/5 ( decimal input )

1/2

3/25 ( decimal input )

1 0/1 ( = 1 0/2) 31

Fraction mode is entered via , , or by entering a second radix mark in


numeric input. It is left towards floating point decimal mode via , , , ,
, or .

4. There are three angular modes featured: DEG, RAD, and GRAD. 32 And degrees
(DEG) may be displayed in decimal numbers as well as in hours, minutes, seconds and
hundredth of seconds (H.MS). The usual angular conversions are provided:

30
This is a new feature compared to previous RPN calculators. By the way, ‘proper fractions’ cover “echte
Brüche” (like ¾) and “gemischte Brüche” (like 2 ½) in German.
31
This display of an integer number tells you unambiguously your WP 34S is in proper fraction mode. For
comparison, note the HP-32SII reads as ½ – but this is not consistent with its other input
interpretations in fraction mode.
WP 34S Owner’s Manual Edition 3.1 Page 42 of 211
Example: Enter the following: and you will see:
Choose radians as current
angular mode.
300 4 0.0105 So /300
0.6000 are exactly 0.6°
0.3600 or 0°36’0”.

The full spectrum of angular conversions provided is seen here:

From degrees decimal radians gon current


H.MS degrees (grad) angular
mode

… to degrees H.MS — — — —

… to decimal degrees —
rad→° G→°

… to radians — —
°→rad G→rad

… to gon (grad) — —
°→G rad→G

… to current angular mode — —


DEG→ RAD→ GRAD→

See the IOP for more.

5. For fitting measured and accumulated data points with a regression curve, four
mathematical models are provided as in the HP-42S. See the commands EXPF, LINF,
LOGF, and POWERF in the IOP. The command BESTF will set your WP 34S to select
the model resulting in the greatest absolute correlation coefficient (see CORR).

As shown in this example

,
the fit model applied is displayed temporarily after each command related to fitting (i.e.
after CORR, COV, L.R., sXY, x, ŷ). Like with all other auto-functionality, you should
know what you are doing here.

32
This traditional calculator notation is misleading in German at least: DEGrees on your WP 34S mean
“Grad”, while GRADs are called “Gon”.
WP 34S Owner’s Manual Edition 3.1 Page 43 of 211
6. In H.MS display mode, entered via , decimal numbers are converted and
displayed in a format hhhh°mm'ss.dd" with the number of hours or degrees limited to
9,000. This temporary message may look like

for RDX. or for RDX, .

For decimal times less than 5ms or 0.005 angular seconds but greater than zero,
an u for underflow will be lit in the exponent section:

Note there are no leading zeroes in the hours, minutes, and seconds sections.

For times or angles exceeding 9,000, an o is shown in the exponent section signaling
an overflow, and the value is displayed modulo 9,000.

For example:

will become after


until the next key is pressed.

7. WDAY returns a display looking like the following for an input of 13.01201 in default
D.MY mode (equivalent to inputs of 2010.0113 in Y.MD or 1.13201 in M.DY):

Expect similar displays after DAYS+.33

33
Dates before the year 8 may be indicated differently to what they really were due to the inconsistent
application of the leap year rule before this. We count on your understanding and hope this shortcoming
will not affect too many calculations.
WP 34S Owner’s Manual Edition 3.1 Page 44 of 211
Floating Point Modes – 3: Real Statistical Calculations

Besides the basic functions and , you will find a lot of statistical commands
embedded in your WP 34S, going far beyond the Gaussian distribution. They are all
concentrated in the light green frame shown in the picture. Many preprogrammed
operations are implemented in your WP 34S for the first time ever in an RPN calculator –
we packed-in everything we always had missed.

The labels below


and following to the
right cover sample
statistics. The five
labels at left deal with
general probability.

, , and
stand for x!, COMB,
and PERM as intro-
duced above.

The shifted functions of cover statistical distributions: standard normal Φ and its
inverse Φ , as well as the catalog covering nine more continuous distributions
and three discrete ones. All these functions have a few features in common:

Discrete statistical distributions (like Poisson and Binomial) are confined to integers.
Whenever your WP 34S sums up a probability mass function (pmf 34) to get a
cumulated distribution function (cdf) it starts at . Thus,
m
F (m) p ( n) Pm .
n 0

Whenever your WP 34S integrates a function, it starts at the left end of the integration
interval. Thus, integrating a continuous probability density function (pdf) to get a
cdf typically works as
x
F ( x) f d Px .

34
In a nutshell, discrete statistical distributions deal with “events” governed by a known mathematical model.
The pmf then tells the probability to observe a certain number of such events, e.g. 7. And the cdf gives the
probability to observe up to 7 such events, but not more.
For doing statistics with continuous statistical variables – e.g. the heights of three-year-old toddlers –
similar rules apply: Assume we know the applicable mathematical model. Then the respective cdf gives the
probability for their heights being less than an arbitrary limit value, for example less than 1 m. And the
corresponding pdf tells how these heights are distributed in a sample of let’s say 1000 children of this age.
BEWARE: This is a very rudimentary sketch of this topic only – turn to good textbooks about statistics
to learn dealing with it properly.
The terms pmf and pdf translate to German „Dichtefunktion“ or „Wahrscheinlichkeitsdichte“, cdf to
„Verteilungsfunktion“ or „Wahrscheinlichkeitsverteilung“.
WP 34S Owner’s Manual Edition 3.1 Page 45 of 211
Many frequently used pdfs and
pmfs look more or less like the
ones plotted in the upper dia-
gram here. The corresponding
cdfs are shown below, using
the same colors. Typically, any
cdf starts with a very shallow
slope, becomes steeper then,
and runs out with a decreasing
slope while slowly approaching
100%. This holds even if the
respective pdf does not look as
nice and symmetric as the
examples here.
Obviously you get the most
precise results on the left side
of the cdf using P. On its right
side, however, where P ap-
proaches 1, the error prob-
ability Q=1–P is more
precise. Thus, your WP 34S
also computes Q for each
distribution, independent of P .

On your WP 34S, with an arbitrary cdf named XYZ you will find the name
XYZu for its error probability Q (also known as upper tail probability), if applicable,
XYZ -1 for the inverse of the cdf (the so-called quantile function or qf ), and
XYZP for its pdf or pmf.
This naming convention holds for Binomial, Cauchy (a.k.a. Lorentz or Breit-Wigner),
Exponential, Fisher’s F, Geometrical, LogNormal, Logistic, Normal, Poisson,
Student’s t, and Weibull distributions. Chisquare and Standard Normal (Gaussian)
distributions are named differently for reasons of tradition. See the catalog PROB and
the respective entries in the IOP. For those interested, some mathematical background
and further links may be found in Appendix I.

There is also a wealth of commands for sample and population statistics featured,
applicable in one or two dimensions. All these are located to the right of key on the
keyboard. After clearing the summation registers by , use (or the primary
shortcut top left on the keyboard) to accumulate your experimental data – typically
counted or measured values – as on the HP-42S etc.; weighted data require the weight in
Y, pairs of data or coordinates of data points must be provided in X and Y as usual.
is provided for easy data correction.

WP 34S Owner’s Manual Edition 3.1 Page 46 of 211


Regarding the analysis functions featured, you find the arithmetic mean and standard
deviation handy on the keyboard, as well as a forecasting function and the
correlation coefficient (see CORR) for four different regression models (linear,
exponential, logarithmic, and power – see the commands LINF, EXPF, LOGF, and
POWERF). Many more functions (such as standard errors, covariances, means and
standard deviations for weighted data, standard deviations for populations, curve fitting
parameters, geometric means and scattering factors) are in STAT, and all the
accumulated data are in SUMS – just look them up in these catalogs and then check the
respective entries in the IOP.

To get an idea of the possibilities provided and of some constraints inherent to statistics,
see the two sample applications shown here:

Application 1: Assume you own a little tool shop and want to know the quality of the
parts you produce. You drew a representative sample of nominally equal parts and
precisely measured their real sizes. How can you know your batch will be ok? That is
easy, based upon analysis of that sample.

Example: 10 pins drawn from a production batch, diameters measured:


12.356, 12.362, 12.360, 12.364, 12.340, 12.345, 12.342, 12.344, 12.355, and
12.353. – From earlier investigations, it is known that diameters from this
production process are Gaussian.

Now you must know your objective:

Do you want to know what pin diameters you will get? Statistics cannot tell you
about all of them here but it will tell you where to find almost all (e.g. 99%) of
them.

Example (continued): Reset the summation registers and accumulate the ten
measured values:
03
12.356 1.000
12.362 2.000
12.360 3.000
12.364 4.000
12.340 5.000
12.345 6.000
12.342 7.000
12.344 8.000
12.355 9.000
12.353 10.000

WP 34S Owner’s Manual Edition 3.1 Page 47 of 211


By knowing these pin diameters are from a Gaussian process, you get the
best estimates for the mean and standard deviation of your batch by pressing
12.352 and
0.009 .

We stored these two values


in preparation for the next
step already.

99%
0.5% 0.5%
So based on the ten pin 12,321

12,352

12,383
sample analyzed, you may
expect 0.5% of all pins with
diameters smaller than
0.005 ↕ Norml⁸ 12.330
and another 0.5% with diameters greater than
0.995 ↕ Norml⁸ 12.375 .

Are you interested in the mean pin diameter of your batch? Do you want to know
the limit below which it will lie with a probability of e.g. 95%?35 Then determine
the applicable sample mean value and the size of its variation. Use them to find
said limit.

Example (continued): Since it is a Gaussian process, the arithmetic mean is


applicable, the standard error tells its variation, and Student’s t is needed.
Press …
↕ nΣ 10.000
1 9.000
↕ SERR 0.003
0.95 ↕ t⁸(p) 1.833
0.005
12.352
36
12.357
So the odds are 95% that the mean pin diameter of that batch will be less
than the confidence limit 12.357. But remember there is an inevitable chance
of 100% – 95% = 5% that it will be greater than that.
35
This value of 95% is called the confidence level of this calculation. 99% are frequently applied as well.
36
returns two numbers: x and y. Only x is interesting in this example. Pressing moves y
quickly out of the way. In a program, followed by DROP is a better alternative.
WP 34S Owner’s Manual Edition 3.1 Page 48 of 211
Do you want both upper and lower limits confining the mean pin diameter of that
batch? Take 0.025 and 0.975 as the arguments in two subsequent calculations
using the qf to get both limits below and above the sample result.

Example (continued, assuming J loaded in the step before):


↕ SERR 0.003
0.025 ↕ t⁸(p) -2.262
0.006
12.352
0.006
12.346 = lower
limit.
0.006 = last x
2 12.358 = upper
limit.37
Now you know you can expect the future mean diameter of that batch
between the confidence limits 12.346 and 12.358 – with a confidence of 95%.
This means here will be a chance of 2.5% that it will be smaller than the
lower limit and an equal chance that it will be greater than the upper limit.

These chances are an inevitable consequence of the fact that you know something
about a sample only (being a limited, usually small number of specimens drawn from
a population), but want or have to tell something about said total population.38 If you
cannot live with these chances or the widths of the confidence limits, do not blame
statistics but collect more (or more precise) data instead.

Application 2: Assume you have taken a sample out of a process at day 1. Then
you have changed the process parameters, waited for stabilization, and now have
taken another sample of same size at day 2. Being serious, you have thoroughly
measured and recorded the critical value (e.g. a characteristic dimension) for each
specimen investigated at both days. Now: do the results of both samples show any
significant difference? The following simple three-step test is well established. It may
easily save yourself some unwanted embarrassments in your next presentation or
after your next publication39 :

37 -1
The upper limit can be calculated this easy way since t (p) is symmetric.
38
Statisticians call these chances ‘probabilities of a type I error’ or ‘probabilities of an error of the first kind’.
By the way, ‘confidence limit’ translates to German “Vertrauensbereichsgrenze”, ‘confidence level’ to
„Vertrauensniveau“, and ‘type I error’ to “Fehler 1. Art”.
39
This test goes back to DGQ (Deutsche Gesellschaft für Qualität). It assumes your data are drawn from a
Gaussian process, which is frequently the case in real life (but needs to be checked). Note the term
‘significant’ is well defined in statistics – this definition may deviate from common language.
Generally, standard confidence limits and levels, also those defined for indicating significant differences,
may depend on the country or industry you are working in. Be sure to check the applicable valid standards.
WP 34S Owner’s Manual Edition 3.1 Page 49 of 211
1. Let your WP 34S compute and the standard error sE for both samples, then
x1 x2
their normalized distance d . Assume you are working with four
s E21 s E2 2
stack levels still, this calculation could look like the following:
↕ SERR
returns both standard errors in X and Y.
this is the complete denominator now.
and this is d.

Also provide the degrees of freedom for the next two steps:
↕ nΣ
recall the number of specimens measured.
calculate the degrees of freedom
and store them where they belong for the
next step.

2. Let your WP 34S calculate the critical limit tcr of Student’s t for f degrees of
freedom and a probability of 97.5% now:
.975 0.975
↕ t⁸(p) as mentioned above, the requested qf lives
in catalog PROB.
executes this function, taking the degrees of
freedom stored in J to get tcr.

If d t cr then the test indicates the difference between both samples being due to
random deviations only. Congratulations – you have got a robust process regarding
the parameters you changed.

3. Let your WP 34S compute a new critical limit tcs for f and 99.5%:
.995 0.995
↕ t⁸(p) note this function shows up immediately
when opening the catalog again.
get tcs.

If d t cs then the test indicates a significant difference between both samples.


Congratulations – your parameter change caused an effect.
For t cr d t cs , however, you cannot decide based on the information provided –
your samples may contain too little data or your measurements were not precise
enough or the process is scattering too far – so you better stay silent or mumble
something like “investigation in progress”.

WP 34S Owner’s Manual Edition 3.1 Page 50 of 211


We strongly recommend you turn to a good statistics textbook for more information about
statistical methods in general, the terminology used, and the mathematical models
provided.

WP 34S Owner’s Manual Edition 3.1 Page 51 of 211


Integer Modes – 1: Introduction and Virtual Keyboard

Integer modes are entered as described on p. 36. They are designed to deal exclusively
with integers – in input, output, and calculations. This is useful e.g. for computer logic and
system programming – typical applications of an HP-16C. Your WP 34S contains all the
functions of the HP-16C and more, and it allows for integer computing in fifteen bases from
binary to hexadecimal (see overleaf).

In integer modes, functions such as SIN


make no sense for obvious reasons.
Thus, for integer bases ≤ 10, the virtual
keyboard of your WP 34S will look as
shown left (where labels headed by a
red arrow will leave integer modes when
called, typically returning to floating
point decimal mode).

For base 16, on the other hand, primary


functions of the top six keys will be
reassigned automatically, becoming
direct numeric input. So this row will
then look virtually as shown below.
Wherever a default primary function is
not primary anymore after reassign-
ment, prefix will allow for accessing
it (e.g. will call here40). To
ease operation, pressing any key (or a
sequence of prefix(es) and a key) will
display its present assignment in the top
line for checking. If the last key is held
down for > 0.5 seconds, the display will
fall back to NULL and no operation will
be executed.

yX √x
A B C D
ALL FIX SCI ENG MODE
A B C D E F

Calculating in bases 11 … 15, those keys not needed for numeric input will work as shown
in the first picture above. In any integer base, attempts to enter an illegal digit from the
keyboard – such as e.g. 4 in binary – will be blocked.

40
In such cases, operations printed golden on the key plate cannot be called anymore. This means for the
key , for example, we cannot access in hexadecimal mode – no loss here, of course.
Reassignments are generally chosen this way. – Note may call a program if defined.
WP 34S Owner’s Manual Edition 3.1 Page 52 of 211
Integer Modes – 2: Displaying Integer Numbers

In integer modes, the mantissa section of numeric display shows the integer in X. The
exponent section is not needed for numeric display here but may transmit other
information. Sign and first digit of the exponent indicate the base set:

Base 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Exponent starts with b 3 4 5 6 7 o 9 d -1 -2 -3 -4 -5 h

Carry and overflow – if set – show up as a c in the second or an o in the third digit
of the exponent, respectively. So in base 12, the exponent section may well look
like -2co . Generally, carry and overflow behave like they do in the HP-16C. Note
that they are flags – if you want to set, clear, or check them individually, use the flag
commands of your WP 34S.

Word size and complement setting are indicated in the dot matrix using a format xx.ww,
with xx being 1c or 2c for 1’s or 2’s complement, respectively, un for unsigned, or sm for
sign-and-mantissa mode. Startup default is 2c. These modes control the handling of
negative numbers. They are understood most easily with a little example:

Set your WP 34S to WSIZE 12, LZON. This setting allows seeing all 12 bits in one
WP 34S display easily. Enter 147. Then turn to 1COMPL, BASE 2. You will see:

and – after – .
Forget the ’¦ top right for the moment – it will be explained later here. Note
the low byte of our number is displayed larger than its top four bits for easy
reading. Obviously in 1c inverts every bit, being equivalent to here.

Return to the original number via now, choose 2COMPL and you will get:

and – after – .
Note the negative number equals the inverse plus one in 2c.

Now return again to the original number via , choose SIGNMT and you will see:

and – after – .
Negating a number will just flip the top bit in sm (hence the name of it).

WP 34S Owner’s Manual Edition 3.1 Page 53 of 211


Finally, return to the original number via , choose UNSIGN and you will get:

and – after – .
Note the second number looks like in 2c, but in addition an overflow is set here.

This needs explanation, since changing signs has no meaning in un per definition41,
where the most significant bit adds magnitude, not sign, so the largest value
represented by a 12-bit word is 4095 instead of 2047. Thus, should be illegal here
or result in no operation at least. Nevertheless, in unsigned mode was allowed and
implemented this way in the HP-16C, so we follow this implementation for sake of
backward compatibility.

Thus, pressing will not suffice anymore for returning to the original number
here; you must also clear the overflow flag by explicitly (see p. 24).

As you have seen, positive numbers stay unchanged in all those four modes. Negative
numbers, on the other hand, are represented in different ways. Therefore, taking a
negative number in one mode and switching to another one will lead to different
interpretations. The fixed bit pattern representing e.g. -147 d in default 2c will be
displayed as -146 d in 1c, -1901 d in sm, and 3949 d in un.

Keeping the mode (e.g. 2c again) and changing the bases will produce different views of
the constant bit pattern as well. You will notice that the displays for bases 4, 8, and 16 will
look similar to those shown above, presenting all twelve bits to you, while in the other
bases a signed mantissa will be displayed instead.

Compare for example the outputs

for base 4 and

for base 5.42

Let us look to bigger words now: For example, turn to UNSIGN, LZOFF, WSIZE 64,
BASE 16, and enter 93A14B6. Then your WP 34S will display:

with or

41
This is clearly stated also in the HP-16C Computer Scientist Owner’s Handbook of April 1982 on page 30.
Unfortunately, however, they did not stick to this.
42
This takes into account that bases 2, 4, 8, and 16 are most convenient for bit and byte manipulations and
further close-to-hardware applications. On the other hand, the bases in between will probably gain most
interest in dealing with different number representations and calculating therein, where base 10 is the
common reference standard.
WP 34S Owner’s Manual Edition 3.1 Page 54 of 211
without separators selected (see SEPON and SEPOFF).

This number will need 28 digits in binary representation, being


1001.0011.1010.0001.0100.1011.0110. Now return to SEPON and choose BASE 2,
then the twelve least significant digits will be displayed initially:

They show up together with an indication '''| that there are four display windows in
total with the rightmost shown. The least significant byte is emphasized as you know
it from the example above. Press and you will get the more significant bytes (note
the constant 4-bit overlap with the previous display here):

The last display shows the four most significant bits of this binary number as the
indication |''' confirms.

If leading zeros were turned on (see LZON), there would be eight display windows
(corresponding to eight bytes) here, with the four ‘most significant’ bytes containing
only zeros.

Note numeric input is limited to 12 digits in all integer bases.

Browsing a large integer in steps of eight digits (i.e. byte-wise) is a specialty of binary
mode. In any other base the step size is the full display width, i.e. twelve digits without any
overlap. See, for example, the most and least significant parts of the same number in
base 3:

WP 34S Owner’s Manual Edition 3.1 Page 55 of 211


Integer Modes – 3: Bitwise Operations

Your WP 34S carries all the bitwise operations you may know from the vintage HP-16C,
plus some more. Generally, bits are counted from right to left, starting with number 1 for
the least significant bit. This is important for specifying bit numbers in the operations BC?,
BS?, CB, FB, and SB.

The following examples deal with 8-bit words showing leading zeros for easy reading. So
set WSIZE 8, LZON on your WP 34S. For seven functions, you will find the schematic
pictures in the table below as they are printed on the backplate of the HP-16C. The ‘C’ in a
box stands for the carry bit there.
Common start-up x
b
1011. 0 011
Operation Schematic picture Example Output
bc
SL 1 0110. 0 110
Shift Left b
SL 2 1100. 1 100
bc
SR 1 0101. 1 001
bc
Shift Right SR 2 0010. 1 100
b
SR 3 0001. 0 110
in 2COMP and 1COMP:
b
1111. 0 110
Arithmetic in UNSIGN:
ASR 3 b
Shift Right 0001. 0 110
in SIGNMT:
b
1000. 0 110
bc
RL 1 0110. 0 111
Rotate Left b
RL 2 1100. 1 110
bc
RR 1 1101. 1 001
bc
Rotate Right RR 2 1110. 1 100
b
RR 3 0111. 0 110
bc
Rotate Left RLC 1 0110. 0 110
through Carry b
RLC 2 1100. 1 101
bc
RRC 1 0101. 1 001
Rotate Right bc
RRC 2 1010. 1 100
through Carry
b
RRC 3 1101. 0 110
WP 34S Owner’s Manual Edition 3.1 Page 56 of 211
Note the picture for ASR correctly describes this operation for 1’s and 2’s complement
modes only. In all modes of the HP-16C, however, ASR 3 equals a signed division by 23,
hence the different results for the latter two modes shown above. The other bitwise
operations are insensitive to complement mode setting. For further details about those
operations, turn to the IOP.

Now let us show you the bitwise two-number functions provided as well. Of these, Boole’s
operators AND, OR, and XOR are found in the light blue frame drawn here:

Again, we will use 8-bit words for the following examples:

b
Y 0110. 1 011
Common input
b
X 1011. 1 001
Operation Output

AND b
0010. 1 001
NAND b
1101. 0 110
OR b
1111. 1 011
NOR b
0000. 0 100
XOR b
1101. 0 010
XNOR b
0010. 1 101

See the IOP for those and further commands working on bit level in integer modes (NOT,
LJ and RJ, MASKL and MASKR, MIRROR, RAN#, and nBITS). Unless on the keyboard,
the commands mentioned so far are found in the catalog X.FCN in integer modes. And
there are also BS? and BC? in TEST.

Finally, note that no such operation will set an overflow. Carry is only settable by shift or
rotate functions as demonstrated above. And ASR is the only bitwise operation being
sensitive to complement mode settings – ASR is the link to integer arithmetic operations.

WP 34S Owner’s Manual Edition 3.1 Page 57 of 211


Integer Modes – 4: Integer Arithmetic

Of the four basic arithmetic operations ( +, -, ×, and / ), the first three work in integer
modes as they do in DECM, but with up to 64 digits precision in binary mode. Look at +/-
like a multiplication times -1, and at yx like a repeated multiplication. Depending on the
input parameters and the mode settings, the overflow or carry flags will be set in such an
operation (see the end of this section).

Divisions, however, must be handled differently in integer modes since the result cannot
feature a fractional part here. Generally, the formula
applies, with the horizontal bar denoting real division, div representing integer division, and
rmdr standing for the remainder of the latter. While remainders for positive parameters are
simply found, remainders for negative dividends or divisors lead to confusion sometimes.
The formula above, however, is easily employed for calculating such remainders (also for
real numbers – see the first line of this example):

Examples: (and for a real case: )

In general, .

Unfortunately, there is a second function doing almost the same: it is called mod.
With the same pairs of numbers as above, it returns:
,
,
,
.

So mod returns the same as rmdr if both parameters have equal signs only.
The general formula for mod is a bit more sophisticated than the one above:

with e.g. and .

By the way, this formula applies to real numbers as well. So it may be used straight-
forwardly for calculating e.g. .

These four functions are called IDIV, RMDR, MOD and FLOOR in your WP 34S for
obvious reasons.

WP 34S Owner’s Manual Edition 3.1 Page 58 of 211


Furthermore, the exponential and logarithmic operations, x2 and , x3 and ,
COMB and PERM work in integer modes, too. Beyond these keyboard functions and those
operations mentioned further above, there are more in the catalog X.FCN such as ×MOD
and ^MOD. See the IOP for further information about them.

Finally, there are conditions where overflow or carry will be touched in integer arithmetic.
Note there is a maximum and a minimum integer for each word size and complement
setting – let us call them Imax and Imin. For example,
Imax = 15 and Imin = 0 for un 04, while
Imax = 7 and Imin = -8 for 2c 04,
Imax = 7 and Imin = -7 for 1c 04 and sm 04.

Now let us start from one incrementing by one and see what will happen:

UNSIGN 2COMP 1COMP SIGNMT


b d b d b d
0001 1 0001 1 0001 1 0001 b 1 d
0010 b 2 d
0010 b 2 d
0010 b 2 d
0010 b 2 d
… … … … … … … …
0111 b 7 d
0111 b 7 d
0111 b 7 d
0111 b 7 d
1000 b 8 d
1000 b o
-8 d o
1000 b o
-7 d o
1000 bco -0 dco
1001 b 9 d
1001 b -7 d
1001 b -6 d
0001 b 1 d
… … … … … …
b
1110 14 d 1110 b -2 d
1110 b -1 d

1111 b 15 d 1111 b -1 d
1111 b -0 d

0000 bco 0 dco 0000 bc 0 dc


0001 bc 1 dc

b d b d
0001 1 0001 1

From here we start another turn decrementing by one:

UNSIGN 2COMP 1COMP SIGNMT


b d b d b d
0001 1 0001 1 0001 1 0001 b 1 d

b
0000 0 d 0000 b 0 d
0000 b 0 d
0000 b 0 d

1111 bco 15 dco 1111 bc -1 dc


1110 bc -1 dc
1001 bc -1 dc

1110 b 14 d 1110 b -2 d
1101 b -2 d
1010 b -2 d

… … … … … … … …
b d b d b d
1001 -7 1001 -6 1110 -6
b d b d b d
1000 -8 1000 -7 1111 -7
b o d o b o d o b o d o
0111 7 0111 7 1000 -0
b d b d b d
0110 6 0110 6 1001 -1

Note the most significant bit is #3 in SIGNMT and #4 in all other modes here. And
subtractions are actually additions in SIGNMT.
WP 34S Owner’s Manual Edition 3.1 Page 59 of 211
With this, Imax, and Imin, the general rules for setting and clearing carry and overflow are as
follows:

Operation Effect on carry Effect on overflow

+, RCL+, STO+, Sets c if there is a carry out of the Sets o if the result would exceed
INC etc. most significant bit, else clears c. [Imax ; Imin], else clears o.

-, RCL-, STO-, Sets c in a subtraction m – s if Sets o if the result would exceed


DEC etc. in 1COMP / 2COMP the binary [Imax ; Imin], else clears o.
subtraction causes a borrow 43
into the most significant bit,
in UNSIGN m is smaller than s,
in SIGNMT m is smaller than s
and both have the same sign.
Else clears c.

×, RCL×, STO×, None. Sets o if the result would exceed


+/-, (-1)x, x2, x3, [Imax ; Imin], else clears o. Thus in
LCM, x! etc. UNSIGN, +/- always sets o and
(-1)x does so for odd x.

DBL× None. Clears o.

/ , RCL / , STO Sets c if the remainder is ≠ 0, else Clears o (except setting it for the
/ , DBL /, LB, clears c. division Imin / (-1) in 2COMP).
LG, LN, LOGx,

2x Sets c if x = - 1, or in UNSIGN if Sets o if the result would become


x = wsize or in the other modes if > Imax, else clears o.
x = wsize - 1. Else clears c.

yx, 10x Sets c for x < 0 (and for 00), else


clears c.

ex Sets c for x ≠ 0, else clears c.

ABS None. Clears o (except setting it for


x = Imin in 2COMP).

43
See e.g. the examples above. Both carry and borrow translate to “Übertrag” in German – the ‘borrow’ in
subtraction is a specialty of the USA. See the subtle methodic differences in subtracting shown in detail in
http://de.wikipedia.org/wiki/Subtraktion#Schriftliche_Subtraktion . The corresponding English article is far
less instructive.
WP 34S Owner’s Manual Edition 3.1 Page 60 of 211
Full Alpha Mode – 1: Introduction and Virtual Keyboard

Alpha mode is typically entered via (see p. 36). It is designed for text entry, e.g. for
entering prompts and answers. In this mode, the alpha register is displayed in the upper
part of the LCD. All direct input goes there, and the numeric line (kept from your last
calculation) is accessible by commands only. The display may look like this:

In alpha mode, most


mathematical operations
_ _ CPX are neither necessary
MODE nor applicable. So the
A B C D E F keyboard is reassigned
STO RCL f g h automatically when you
VW + R enter alpha mode.
G H I
All labels printed here on
red background append
ENTER corresponding charac-
CL ters to alpha directly or
J ( ) K L
via alpha catalogs. Note
XEQ 7 8 9 / four new catalogs be-
come active in this
M N O P mode. See the keys
and , and the labels
▲ 4 5 6 × and on your
! WP 34S.
Q R S T
A Those labels printed on
n
d
▼ 1 2 3 – darker red background
t
TEST X.FCN changed their functiona-
h U V W lity in other ways. See
e
the keys , ,
EXIT 0 ● R/S + and , as well as the
OFF PSE ./ ,
labels and .
X Y Z

Within alpha mode, primary function of most keys becomes appending the letter printed
bottom left of this key – grey on the key plate – to alpha. appends a space. When
alpha exceeds 30 characters, the leftmost character(s) are discarded. Alpha mode starts
with capital letters, and toggles upper and lower case. As in integer modes, will
44
access default primary functions wherever necessary.

44
The digits 0 and 1 may also be called using or , respectively.
WP 34S Owner’s Manual Edition 3.1 Page 61 of 211
Looking at the standard labels on the keyboard, we can safely offer you even more in this
mode: All labels printed on dark blue background in the virtual keyboard below append
characters to alpha as well. They are related to the labels printed on your WP 34S
keyboard at these locations, but deviate from them. Prefix leads to homophonic Greek
45
letters where applicable. And allows also accessing logic symbols via the Boolean
operations.

The alpha catalogs


called by ,
_ _ CPX , , ,
MODE and feature
A B C D E F even more charac-
STO RCL f g h ters (see p. 132).
R Check the IOP for
VW +
αSTO, αRCL, VWα+,
G H I
and more alpha
commands.
ENTER ±
⇆ ¬ CL
J ( ) K L
XEQ 7 8 9 /
& | ≠ \
M N O P
▲ 4 5 6 ×
!
Q R S T
A
n
d
▼ 1 2 3 –
th ? TEST X.FCN
e U V W
EXIT 0 ● R/S +
OFF PSE ./ ,
X Y Z
For a clearer picture of this virtual keyboard – with the extra colors and redundancies
removed – turn overleaf.

45
“Homophonic” according to ancient Greek pronunciation. And we assigned Gamma also to C due to the
alphabet, and Chi to H since this letter comes next in pronunciation. Three Greek letters require special
handling: Psi is accessed via (below ), Theta via (below and following T), and
Eta via . Omicron is not featured because it looks exactly like the Latin letter O in either
case.
Where we printed Greek capitals with lower contrast, they look like the respective Latin letters in our fonts.
Professors of Greek, we count on your understanding.
WP 34S Owner’s Manual Edition 3.1 Page 62 of 211
The labels underlined in
red call alpha catalogs.
Labels underlined in green
and labels shown between
heavier white key contours
do not append characters
to alpha directly – all other
labels do.

If you print this virtual


keyboard in the size as
given below, you can
attach it to the back of
your WP 34S for ready
reference. This will serve
you well also for learning
the Greek letters and their
relation to their Latin
homophonic siblings.
Thereafter, you may slide
it ‘under the hood’ where
the batteries live, as sort of
a Quick Reference Sheet.

WP 34S Owner’s Manual Edition 3.1 Page 63 of 211


Full Alpha Mode – 2: Displaying Text

Your WP 34S features a large and a small alphanumeric font for display. Both are based
on fonts by Luiz Viera (Brazil) as distributed in 2004. Some letters were added and some
modified for better legibility, also given that the dot matrix of your display is only six pixels
high.

See here all characters directly evocable through the virtual alpha keyboard above:

ABCDEFGHIJKLMNOPQRSTUVWXYZ
abcdefghijklmnopqrstuvwxyz
ABΓΔEZHΘIKΛMNΞOΠPΣTYΦXΨΩ
αβγδεζηθικλμνξoπρστυφχψω
0123456789 ()+-×/±.!?⇆¬\&|≠

As soon as a string exceeds the visible display using this large font, your WP 34S will take
the following small font automatically to show as much as possible:

ABCDEFGHIJKLMNOPQRSTUVWXYZ
abcdefghijklmnopqrstuvwxyz
ABΓΔEZHΘIKΛMNΞOΠPΣTYΦXΨΩ
αβγδεζηθικλμνξoπρστυφχψω
0123456789 ()+-×/±.!?⇆¬\&|≠
Many more characters of both fonts live in the alpha catalogs (see p. 132).

As soon as you enter alpha mode and as long as you stay therein, the contents of the
alpha register (abbreviated by alpha) are displayed in the dot matrix, showing its right end
(i.e. the last characters it is containing), while the numeric section keeps the result of the
last numeric operation. The display may look like:

Note that you may use the big ‘=’ controlled by flag A in addition to the dot matrix for
message display. Different information may be appended to alpha. See the commands
starting with ‘α’ in the IOP.

For example, TIME αTIME allows printing timestamps on messages or creating texts
like It's 11:54:32 PM or Um 23:54:32 Uhr depending on time mode setting and
the actual time at execution.
And DATE αDATE will append – depending on date format setting – either
2015-04-16 or 16.04.2015 or 04/16/2015 to alpha if called on the 7th of April in
2015.

WP 34S Owner’s Manual Edition 3.1 Page 64 of 211


Note alpha can take up to 30 characters. And your WP 34S features a rich set of special
letters and further characters, supporting at least 37 languages. So you may easily store
an actual Greek message like this, for example:

and will browse such long messages in steps of 6 characters. will stop with the
very first characters shown, stops showing the right end completely, i.e.

in this very special case.

Having left alpha mode, you can still display alpha: use VIEWα or VWα+ for this – it will
show you the left end (i.e. the first characters the alpha register contains).

Nevertheless, do not forget that your WP 34S is mainly designed as a programmable


calculator.

WP 34S Owner’s Manual Edition 3.1 Page 65 of 211


PROGRAMMING YOUR WP 34S

Your WP 34S is a keystroke-programmable calculator. If this statement makes you smile


with delight, this section is for you. Else please turn to the HP-42S Owner’s Manual first for
an introduction into keystroke-programming for solving repetitive tasks.

The basic building blocks within program memory are routines (also known as programs).
Such routines may contain subroutines, but that is not mandatory. Typically, a routine or
subroutine starts with a LBL statement and ends with RTN or END. In between, you may
store any sequence of instructions (commands, operations, statements) for repeated use.
Choose any operation featured – only a few commands are not programmable.
The statements in your routine may use each and every global register provided – there
are (almost) no limits. You are the sole and undisputed master of the memory!

This freedom, however, has a price: take care that each routine does not interfere with any
other in its quest for data storage space. It is good practice to record the global registers a
particular routine uses, and to document their purposes and contents for later reference.

In programming mode (i.e. while editing routines), the numeric display will indicate the
current program step (000 – 927) in the mantissa and the number of free steps in the
exponent, while the dot matrix will show the command contained in the current step, e.g.:

There is no routine-specific step counting like in the HP-42S.

The commands partic-


ularly required for pro-
gramming are located
within the violet frame
on this keyboard
picture. Here you find,
starting top left, com-
mands for executing
(XEQ) or going to
(GTO) a specific pro-
gram, keys for browsing
programs ( and ),
flag commands (SF,
CF, and STATUS),
catalogs of TESTs, pro-
gramming functions
(P.FCN), and extra
functions (X.FCN), com-
mands for SoLVing,
integrating, calculating
products and sums, keys for EXITing programming mode and for pausing program
execution (PSE), some commands for extracting parts (|x|, IP, and FP), keys for

WP 34S Owner’s Manual Edition 3.1 Page 66 of 211


Running / Stopping programs (R/S) and switching between Programming and Run mode
(P/R), commands for labeling a routine (LBL) or returning from a routine (RTN), and for
loop control (DSE and ISG). See the IOP for more about these commands.

Labels

Structuring program memory and jumping around in it is eased by labels you may tag to
any program steps – as known from previous programmable pocket calculators. Your
WP 34S features a full set of alphanumeric program labels as described overleaf.
Furthermore, different programs may be separated by END statements. Think of the
beginning and the end of program memory containing implicit END statements.

See the next page for addressing labels.

Searching labels, however, obeys the rule below. When a command such as XEQ lbl is
encountered, with lbl representing a label of one, two or three characters (such as A, BC,
12, Tst, Pg3, x1µ, etc.), your WP 34S will look for this label using the following method:

1. If lbl is purely numeric or a hotkey, it will be searched forward from the current position
of the program pointer. When an END statement is reached without finding lbl, the
quest will continue right after previous END (so the search will stay in the current
routine). This is the search procedure for local labels. It is as known from the HP-41C.

2. If, however, lbl is an alpha label of up to three characters of arbitrary case (auto-
matically enclosed in ‘ like 'Ab1'), searching will start at program step 000 and cover
the entire memory in the order RAM, FM, and XROM (see p. 72 for the latter two),
independent of the position of the program pointer. This is the search procedure for
global labels.

Tests

Like keystroke-programmable calculators before, your WP 34S features a set of tests.


Their command names have a trailing ‘?’. Generally, tests will work as in the HP-42S: they
will return true or false in the dot matrix if called from the keyboard; if called in a
program, they will execute the next program step only if the test is true, else skip that step.
So the general rule reads ‘skip if false’ (exception: KEY?).

As mentioned above, programs typically end with RTN or END. In running programs, both
statements work very similar and show only subtle differences: a RTN statement
immediately after a test returning false will be skipped – an END will not.

See the IOP for more information. All tests are contained in the catalog TEST.

Note that there are also commands featuring a trailing ‘?’ but returning numbers (e.g.
BASE?) or codes (e.g. KTP?) instead of true or false only – you will find these commands
in the catalog P.FCN.

WP 34S Owner’s Manual Edition 3.1 Page 67 of 211


Addressing Labels

1 User
, , , or , , , LBL?, , , , , αGTO, or αXEQ
input

Echo XEQ label OP _


e.g. XEQ C e.g. GTO _

2 User Calls the function , , , or 46 2-digit numeric


input labeled C. (local) label
sets alpha mode. opens indirect addressing and sets mode.

Echo OP label OP ‘_ OP _ OP nn
e.g. Σ B e.g. LBL 07

3 User Sums up the function Alphanumeric Stack level Register number


input given in a routine (global) label or lettered register … ,
labeled B. (1 to 3 characters47) , , , ... , … ,
if the respective
register is allocated48
Echo OP ‘label’ OP x OP nn
e.g. SLV'F1μ' e.g. ∫→T e.g. XEQ→44

Solves the function given Integrates the Executes the routine


in the routine labeled F1µ function whose label whose label is in R44.
(keyed in as explained in footer). is on stack level T.

Look up GTO in the IOP for special cases applying to this command exclusively.

46
Works with all these operations except .
47 rd
The 3 character terminates entry and closes alpha mode – shorter labels need a closing . For the example given here, press
and you are done. Statements including alpha labels exceeding one character decrement the number of free program steps by 2.
ATTENTION: LBL A and LBL’A’ are different animals! The latter is entered in alpha mode, the first via the hotkey directly.
48
Some registers may be allocated to special applications. Check the memory table on p. 23.

WP 34S Owner’s Manual Edition 3.1 Page 68 of 211


Local Data

After some time with your WP 34S you will have a number of routines stored, so keeping
track of their resource requirements may become a challenge. Most modern programming
languages take care of this by declaring local variables, i.e. memory space allocated from
general data memory and accessible for the current routine only – when the routine is
finished, the respective memory is released. On your WP 34S, registers are for data
storage – so we offer you local registers allocated to your routines exclusively.

Example: Let us assume you write a routine labeled P1 and need five registers for
your computations therein.
Then all you have to do is just enter the command LOCR 5 in P1 specifying you
want five local registers. Thereafter, you can access these registers by using local
numbers .00 … .04 throughout P1.

Now, if you call another routine P2 from P1, also P2 may contain a step LOCR
requesting local registers. These will also carry local register numbers .00 etc., but
the local register .00 of P2 will be different from the local register .00 of P1, so no
interference will occur. As soon as the return statement is executed, the local
registers of the corresponding routine are released and the space they took is
returned to free memory.

In addition, you get sixteen local flags as soon as you request at least one local register.

Local data holding allows for recursive programs, since every time such a routine is called
again it will allocate a new set of local registers and flags being different from the ones it
got before.

See the commands LOCR, LOCR?, MEM?, and POPLR in the IOP and in Appendix B for
more information, also about the limitations applying to local data.

Programmed Input and Output, User Interaction and Dialogues

A number of commands may be employed for controlling I/O of programs. In the IOP, their
behavior is described if they are entered from the keyboard. Executed by a program,
however, this will differ in a characteristic way.

When a program is started, the prior display contents are replaced by the Running
PrograMm message and will be updated at certain events only – not after each operation.
So where in manual mode each command may change the display immediately with its
execution, in automatic mode only PROMPT, PSE, STOP, VIEW, VIEWα, or VWα+ will
update the display, and this display will hold until the next command of this kind is
encountered.

ERR 0 or MSG 0 are the only ways to get Running PrograMm back once this message
has been replaced by a programmed display.

See the following examples – parameters are omitted here:

WP 34S Owner’s Manual Edition 3.1 Page 69 of 211


Take VIEW, VIEWα, or VWα+ for plain display updates. X is a valid parameter for
VIEW and VWα+. Note that frequent updates will slow down program execution,
since the anti-flicker logic waits for a complete display refresh cycle before allowing
the next update.

Use one of the following four code segments for displaying messages or other
information for a defined minimum time interval, specified by PSE (see the IOP for
details):

PSE VIEW VIEWα VWα+


PSE PSE PSE
for plain numeric output for complex alphanumeric messages

Ask (‘prompt’) for numeric input employing one of these four:

STOP VIEWα VWα+ PROMPT combines VWα+ X and


STOP STOP STOP in one command

Whatever you key in will be in X when you continue the program by pressing .
If you want it elsewhere, take care of it.

Prompt for alphanumeric input by the following steps:

αON sets alpha mode and prepares for showing the final part of alpha.
PROMPT displays this part and waits for user input, terminated by .
αOFF returns to the numeric mode previously set.

Whatever you key in will be appended to alpha here. The program will continue when
you press .

See the IOP for more information about these commands and their parameters.

If you press – instead of or after keying in numeric data – one of the hotkeys to in
input, the program will call the next routine beginning with a label carrying this name.

The following example shows how a typical structure of such a program might look:

001 LBL ’MYP’


002 CLα
003 α 'Hel' Sets up a message …
004 α 'lo!'
005 LBL 00
006 PROMPT … and stops waiting for user input.
007 GTO 00 does nothing, it simply returns to the prompt.
008 LBL A Called if user input after step 006 was terminated by .
009 ENTRY? Any new numeric data entered by the user?
010 GTO 03 Then go to step 012 where local label 11 lives.
011 XEQ 01 Else call subroutine 01 for computing a new number instead.
012 LBL 03
013 STO 01 Store the new number (input or computed).
014 RTN Return to prompting via step 007.
WP 34S Owner’s Manual Edition 3.1 Page 70 of 211
015 LBL 01
... ... Compute the new number for missing numeric user input.
... ...
... RTN
... LBL B Called if user input after step 006 was terminated by .
... ...
... ...
... RTN
... ...
... END

This is the way the TVM application is implemented.

If there is more than one program using labels A to D in RAM or FM, you must move the
program counter (PC) into the desired program and stop there – provided programs are
separated by END (see p. 67 to learn which label will be found else).

Keyboard Codes and Direct Keyboard Access


Sometimes, the four hotkeys might not suffice.
There is, however, an easy way to extend the 11 12 13 14 15 16
number of directly callable subroutines:
shorthand addressing of numeric labels using
21 22 23 24 25 26
keyboard codes as defined at right. Each key
gets a code simply given by its row and column
on the keyboard. 31 32 33 34 35

Whenever you are asked for the entry of a two-


41 42 43 44 45
digit label, any of the keys highlighted green in
this picture may be used for direct input. The
label will then be replaced by the row/column 51 52 53 54 55
code of the respective key. Keys not available
this way (since they have another fixed 61 62 63 64 65
meaning in this context) may still be used for a
short address by pressing before. Only
itself cannot be used for shorthand addressing. 71 72 73 74 75

If, for example, you want to link a program to the key , just put label 21 at the
beginning of the routine; then it can be called via by the user
conveniently.

The same keyboard codes are returned by the KEY? command, which allows ‘real time’
response to user input from the keyboard. KEY? takes a register argument (X is allowed
but does not lift the stack) and stores the key most recently pressed during program
execution in the register specified. R/S and EXIT cannot be queried; they stop program
execution immediately. The keyboard is active during program execution – but it is
desirable to display a message and suspend the program by PSE while waiting for user
input. Since PSE will be terminated by a key press, simply use PSE 99 in a loop to wait for

WP 34S Owner’s Manual Edition 3.1 Page 71 of 211


input. Since KEY? acts as a test as well, a typical user input loop may well look like this
example:

001 LBL ’USR’


002 CLα
003 α 'KEY' Sets up a message …
004 α ?
005 LBL 00
006 VIEWα … and displays it.
007 PSE 99 Waits 9.9 s for user input unless a key is pressed.
008 KEY? 00 Test for user input and put the key code in R00.
009 GTO 00 If there was none then go back to step 005.
010 LBL?→00 If a label corresponding to the key code has been defined …
011 XEQ→00 … then call it,
012 GTO 00 … else return to step 005.

Instead of the dumb waiting loop, the program can do some computations and update the
display before the next call to PSE and KEY? – think of e.g. a lunar landing game.

To be even more versatile, KTP? nn is designed to return the type of the key pressed if its
row / column code is given in register nn. It will return:
0 to 9 for the respective digit keys;
10 for , , and ;
11 for , , and ;
and 12 for the other keys.
An invalid code in the input register will throw an ‘Invalid Range’ error.

If you decide not to handle the key in your program you may feed it back to the main
processing loop of the WP 34S with the PUTK nn command. It will cause the program to
halt, and the key will be handled as if pressed after the stop. This is especially useful if you
want to allow numeric input while waiting for some special keys like the arrows. This allows
writing a vector or matrix editor in user code. After execution of the PUTK command you
are responsible for letting the program continue its work by pressing or a hotkey.

Flash Memory (FM) and XROM

In addition to the RAM provided, your WP 34S allows you to access FM for voltage-fail
safe storage of user programs and data. The first section of FM is a 2kB backup region,
holding the image of the entire user program memory, registers, and WP 34S states as
soon as you completed a SAVE. The remaining part of FM (up to some 12kB depending
on setup) will hold programs only. Alphanumeric labels in FM can be called via XEQ like in
RAM. This allows creating program libraries in FM. Use CAT to see the global labels
already defined – labels in FM are tagged with Lib there.

WP 34S Owner’s Manual Edition 3.1 Page 72 of 211


FM is ideal for backups or other long-living data, but should not be used for repeated
transient storage like in programmed loops. 49 Conversely, registers and standard user
program memory residing in RAM are designed for data changing frequently but will not
hold data with the batteries removed. So both kinds of memory have advantages and
disadvantages you shall take into account for optimum benefit and longevity of your
WP 34S. Find more about FM in Appendix A.

Furthermore, there is a memory section called XROM (for ‘extended ROM’), where some
command routines live. These, though written in user code, are read-only and thus can be
called as commands and executed but not edited. For you, it makes no difference whether
a preprogrammed WP 34S routine executes in ROM or XROM. The XROM routines,
however, are easy to read and understand for users knowing HP keystroke programming –
see http://wp34s.svn.sourceforge.net/viewvc/wp34s/trunk/xrom/ for the routines present
and Appendix H for additional information.

49
FM may not survive more than some 10,000 flashes. Thus, we made commands writing to FM (like SAVE
or PSTO) non-programmable.
WP 34S Owner’s Manual Edition 3.1 Page 73 of 211
INDEX OF OPERATIONS (IOP)

All commands available (more than 600) are found below with their names and their
necessary keystrokes. Names printed in bold face in this list belong to functions directly
accessible on the keyboard; the other commands may be picked from catalogs. The
command names will show up identically in catalogs and program listings unless specified
otherwise explicitly. Sorting in index and catalogs is case insensitive and works in the
following order:
␣ 0…9, A…Z, … , ( ) + – × / ± , . ! ? : ; ‘ “ * @ _ ~
< ≤ = ≈ ≠ ≥ > % $ € £ ¥ √ ∫ ∞ & \ ^ | [ ] { } ⎙ #

Superscripts and subscripts are handled like normal characters in sorting. The ⎙ near the
end of the sorting order list above is the printer symbol heading all print commands.

Generally, functions and keystroke-programming will work as on the HP-42S, bit and
integer functions as on the HP-16C, unless specified otherwise. There are also functions
inspired by other vintage calculators as mentioned in the index. Their manuals may
contain helpful additional information about those traditional commands.

A C heads the names of complex operations (see p. 29). is a legal prefix for all
functions whose names are printed in italics in this list.

Some 300 functions are available on your WP 34S for the first time ever on an RPN
calculator. They got their remarks printed on yellow background. Operations carrying a
familiar name but deviating in their functionality from previous RPN calculators are marked
light red.

The vast majority of remarks for the respective operation start with a number:
(0) represents functions without effects on the stack,
(1) and (2) are for real or complex one- or two-number functions as defined above,
and
(3) is for real three-number functions;
(-1) and (-2) stand for functions pushing numbers on the stack thus lifting it by 1 or 2
levels, respectively.
Operations disabling stack lift got a special remark.

Parameters will be taken from the lowest stack level(s) unless mentioned explicitly in the
2nd column of the index – then they must follow the command. Some parameters of
statistical distributions must be given in registers J and K as specified. Some examples of
the notation used will be given overleaf.

WP 34S Owner’s Manual Edition 3.1 Page 74 of 211


For the following three examples, assume R12 contains 15.67 (i.e. r12 = 15.67) generally.

n represents an arbitrary integer number which must be keyed in directly, while


n represents such a number which may be specified indirectly via a register as well
(as shown in the tables on pp. 26 and 33); and
n stands for the respective number itself;
Example: RSD 12 rounds x to twelve significant digits, while RSD 12
rounds x to fifteen significant digits.

s represents an arbitrary register address which must be keyed in directly, while


s represents such an address which may be specified indirectly as well; and
s stands for the respective contents (i.e. the contents of register s if specified
directly, or the contents of the address where s is pointing to if specified indirectly);
Example: STO 12 stores x into R12, while STO 12 stores x into R15.

label represents an arbitrary label which must be keyed in directly, while


label represents such a label which may be specified indirectly as well (as shown in
the table following p. 67); and
label stands for the respective label itself.
Example: GTO 12 goes to local label 12, while GTO 12 goes to local label
15.

In the following, each function is listed stating the mode(s) it will work in, indicated by the
corresponding annunciators (see p. 36) or abbreviated by their names. In this column,
‘integer’ stands for an arbitrary integer mode, ‘&’ for a Boolean AND, a comma for an OR,
and ‘¬’ for NOT. So e.g. 2X works in all modes but alpha, even in complex domain. All
operations may also be entered in programming mode unless stated otherwise explicitly.
Many functions contained in P.FCN make sense in programs only. Red print in the modes
column indicates commands working with the quartz crystal and the respective firmware
only (see App. A and H).

Name Keys to press in modes Remarks (see p. 74 for general information)

10 x ¬INPUT (1)

(0) Sets 12h time display mode: e.g. 1:23 will


12h 12h All become 1:23 AM, 23:45 will become 11:45 PM.
This will make a difference in αTIME only.

1COMPL 1COMPL All (0) Sets 1’s complement mode for integers.

DECM (1)
1/x
DECM (1) Shortcut working if label B is not defined.

24h 24h All (0) Sets 24h time display mode. Compare 12h.

2COMPL 2COMPL All (0) Sets 2’s complement mode for integers.

WP 34S Owner’s Manual Edition 3.1 Page 75 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)
x
2x ¬INPUT (1)

. ³√x ¬INPUT (1)


A

¬INPUT (1) Returns the absolute value.


ABS
DECM (1) Returns r x2 y 2 in X and clears Y.

ACOS DECM (1) Returns .50

(1) Returns . Note there is no need for


ACOSH DECM
pressing here.

(2) Returns the arithmetic-geometric mean of x and


AGM AGM DECM
y. See Appendix I for more.

(0) Sets the numeric display format to show all


decimals whenever possible. ALL 0 works almost
like ALL in the HP-42S.
13
For x ≥ 10 , however, display will switch to SCI or
ENG with the maximum number of digits
necessary (see SCIOVR and ENGOVR). The
ALL n ¬INPUT same will happen if and more than 12
digits are required to show x completely.
Example: Input: Display:
700
ALL 03
1/x
10 /

Integer (2) Works bitwise as in the HP-16C. See p. 56.

AND (2) Works like AND in the HP-28S, i.e. x and y are
DECM interpreted before executing this operation. Zero is
‘false’, any other real number is ‘true’.

ANGLE ANGLE DECM (2) Returns .

ASIN DECM (1) Returns .51

(1) Returns . Note there is no need for


ASINH DECM
pressing here.

50
Precisely, it returns the principal value of it, i.e. a real part ∈ [0, ] in RAD or [0, 180] in 360 or [0, 200] in Ⓖ.
Compare ISO/IEC 9899.
51
Precisely, it returns the principal value of it, i.e. a real part ∈ [- /2, /2] in RAD or [-90, 90] in 360 or [-100,
100] in Ⓖ if flag D is set. Else this result interval becomes e.g. (- /2, /2) for the arc tangent. Compare
ISO/IEC 9899.
WP 34S Owner’s Manual Edition 3.1 Page 76 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)

(1) Works like n (≤ 63) consecutive ASR


commands in the HP-16C, corresponding to a
ASR ASR n Integer n
division by 2 . ASR 0 executes as NOP, but loads
L. See p. 56 for more.

ATAN DECM (1) Returns .51

(1) Returns . Note there is no need for


ATANH DECM
pressing here.
B

(0) Jumps n steps backwards (0 ≤ n ≤ 255). E.g.


BACK 1 goes to the previous program step. If
BACK attempts to cross an END, an error is
thrown. Reaching step 000 stops program
BACK BACK n ¬INPUT execution. Compare SKIP.
ATTENTION: If you edit a section of your routine
that is crossed by one or more BACK, SKIP, or
CASE jumps, this may well result in a need to
manually maintain all those statements individually.

BASE BASE n All (0) Sets the base n for integer calculations, with
2 ≤ n ≤ 16 (see pp. 52ff for more). Popular bases
BASE 10 are directly accessible on the keyboard.
Furthermore, BASE 0 sets DECM, and BASE 1
BASE 16 calls FRACT. See there.
¬INPUT
ATTENTION: Stack contents are converted when
BASE 2 switching from an integer mode to DECM, and are
truncated vice versa. Other register contents stay
BASE 8 as they are (see p. 159). – BASE 10 is not DECM.

Integer (-1) Returns the integer base set.


BASE? BASE?
DECM (-1) Returns the integer base set before DECM.

(0) Measures the battery voltage in the range


DECM
between 1.9V and 3.4V and returns this value.
BATT BATT
(0) Works like in DECM but returns the value in
Integer
units of 100mV.

BC? BC? n Integer (0) Tests the specified bit in x.

(0) Selects the best curve fit model, maximizing


BestF BestF All
the correlation like BEST does in the HP-42S.

WP 34S Owner’s Manual Edition 3.1 Page 77 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

Binom Binom (1) Binomial distribution with the number of


successes g in X, the probability of a success p0 in
BinomP Binomₚ J and the sample size n in K. See Appendix I for
DECM the formula.
Binomu Binomᵤ Binom–1 returns the maximum number of
successes m for a given probability p in X, with p0
Binom –1 Binom⁸ in J and n in K.

Bn Bₙ (1) Bn returns the Bernoulli number for an integer


DECM n > 0 given in X. Bn* works with the old definition
Bn* Bₙ⁺ instead. See Appendix I for details.

BS? BS? n Integer (0) Tests the specified bit in x .


C

(0) Works like SKIP below but takes the number of


steps to skip from s.
Example: Assume the following program section:
...
100 CASE 12
101 GTO 01
102 GTO 02
103 GTO 07
104 GTO 05
105 LBL 01
...
132 LBL 02
...
153 LBL 05
...
CASE CASE s ¬INPUT 234 LBL 07
...
In program execution, r12 will be checked in step
100: if r12 ≤ 1 then the program will proceed to
step 101 and continue with a jump to step 105, for
r12 = 2 the program will go to step 102, etc.,
resulting in a nice controlled dispatcher for
1 ≤ r12 ≤ 4 .
ATTENTION 1: CASE might surprise you for
r12 > 4 in the example above. Take care of the
input you provide in s !
ATTENTION 2: If you edit a section of your routine
that is crossed by one or more BACK, SKIP, or
CASE jumps, this may well result in a need to
manually maintain all those statements individually.

¬INPUT Browser. See p. 125.

WP 34S Owner’s Manual Edition 3.1 Page 78 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

Cauch Cauch (1) Cauchy-Lorentz distribution (also known as


Lorentz or Breit-Wigner distribution) with the
CauchP Cauchₚ location x0 specified in J and the shape in K. See
DECM Appendix I for details.
Cauchu Cauchᵤ
Cauch–1 returns x for a given probability p in X,
–1
Cauch Cauch⁸ with x0 in J and in K.

CB CB n Integer (1) Clears the specified bit in x .

CEIL CEIL ¬INPUT (1) Returns the smallest integer ≥ x .

CF n ¬INPUT (0) Clears the flag specified.

CFALL CFALL ¬INPUT (0) Clears all user flags.

(0) Clears all registers, flags, and programs in


CLALL CLALL ¬INPUT RAM if confirmed. Not programmable. Compare
RESET.

(0) Clears the current program, i.e. the one the


CLP All
program pointer is in.

(0) Clears all programs in RAM if confirmed. Not


CLPALL CLPALL ¬INPUT
programmable.

(0) Clears all global and local general purpose


CLREGS CLREGS ¬INPUT registers allocated (see REGS and LOCR). The
contents of the stack, L, and I are kept.

Clears all stack registers currently allocated (i.e. X


CLSTK ¬INPUT through T or X through D, respectively). All other
CLSTK register contents are kept.

CLx ¬INPUT Clears register X only, disabling stack lift as usual.

INPUT
(0) Clears the alpha register like CLA does in the
CL HP-42S.
CLα ¬INPUT

(0) Clears the summation registers and releases


CLΣ DECM
the memory allocated for them.

(-1) Allows indirect addressing of the content at


CNST ¬INPUT
CNST n position n in CONST (see p. 134).

C (-2) Works like CNST but does a complex recall as


CNST ¬INPUT
©CNST n described on p. 128.

WP 34S Owner’s Manual Edition 3.1 Page 79 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Checks for convergence by comparing x and y


as determined by the lowest five bits of s.
a) The very lowest two bits set the tolerance limit:
0 = 10–14,
1 = 10–24,
2 = 10–32,
3 = Choose the best for the mode set, resulting
in taking 0 for SP and 2 for DP (see
Appendix H).
b) The next two bits determine the comparison
mode using the tolerance limit set:
DECM
CNVG? 0 = compare the real numbers x and y relatively,
CNVG? s 1 = compare them absolutely,
2 = check the absolute difference between the
complex values x + i y and z + i t ,
3 = works as 0 so far.
c) The top bit tells how special numbers are
handled:
0 = NaN and infinities are considered con-
verged,
1 = they are not considered converged.
Now, s .

Integer (0) Tests for x = y .

(2) Returns the number of possible sets of y items


taken x at a time. No item occurs more than once
COMB ¬INPUT in a set, and different orders of the same x items
are not counted separately. See Appendix I for the
formula. Compare PERM.

C (1) Flips the sign of y, thus returning the complex


CONJ DECM
©CONJ conjugate of xc .

¬INPUT Catalog. See p. 125.

DECM Catalog. See p. 125.

(-1) Returns the correlation coefficient for the


CORR DECM current statistical data and curve fitting model. See
Appendix I for more.

COS DECM (1) Returns the cosine of the angle in X.

WP 34S Owner’s Manual Edition 3.1 Page 80 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

COSH DECM (1) Returns the hyperbolic cosine of x.

(-1) Returns the population covariance for two data


COV COV DECM sets. It depends on the fit model selected. See sXY
for the sample covariance and Appendix I for more.

(2) Interprets x and y as Cartesian components of


C a first vector, and z and t as those of a second one,
CROSS DECM
©CROSS and returns [ ∙ t – y ∙ z , 0, …] , dropping two stack
levels.
D

(-1) Recalls the date from the real time clock into
the numeric section in the format selected. See
D.MY, M.DY, and Y.MD. In addition, DATE shows
DATE DATE DECM the day of week in the dot matrix.
The function DATE of the HP-12C corresponds to
DAYS+ in your WP 34S (see below).

(-2) Assumes x containing a date in the format


DATE DATE→ DECM selected and pushes its three components on the
stack.

(1) Assumes x containing a date in the format


DAY DAY DECM
selected and extracts the day.

(2) Adds x days on a date in Y in the format


selected and displays the resulting date including
DAYS+ DAYS+ DECM
the day of week in the same format as WDAY
does. Works like DATE in the HP-12C.

DBLOFF DBLOFF (0) Toggles DP mode. Setting becomes effective in


All DECM only and is indicated by D in the dot
DBLON DBLON matrix. See Appendix H.

DBL? DBL? ¬INPUT (0) Checks if DP mode is turned on.

Double word length commands for remainder,


DBLR DBLR multiplication and division. DBLR and DBL/ accept
a double size dividend in Y and Z (most signi-
ficant bits in Y), the divisor in X as usual, and
DBL× DBL× Integer
return the result in X. DBL× takes x and y as
factors as usual but returns their product in X and
DBL / DBL/ Y (most significant bits in X). See the HP-16C
Owner’s Handbook, Section 4 (pp. 52ff).

(0) Decrements s by 1. Does not load L even for


DEC DEC s ¬INPUT
target address X.

DECM ¬INPUT (0) Sets default decimal floating point mode.

WP 34S Owner’s Manual Edition 3.1 Page 81 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(-1) Decomposes x (after converting it into an


improper fraction, if applicable), returning y/x =
in top row and a stack [denominator(x),
DECOMP DECM numerator(x), …] . Reversible by division.
DECOMP
Example: If X contains 2.25 then DECOMP will
return x = 4 and y = 9 , pushing previous content
of Y to Z etc.

DEG DECM (0) Sets angular mode to degrees.

(1) Takes x as degrees and converts them to the


DEG DEG→ DECM
angular mode currently set.

(0) Sets default fraction format like in the HP-35S,


allowing maximum precision in fraction display with
the startup default – any denominator up to the
DENANY All value set by DENMAX may appear.
DENANY
Example: If DENMAX = 5 then DENANY allows
denominators 1, 2, 3, 4, and 5.

(0) Sets ‘factors of the maximum denominator’, i.e.


all integer factors of DENMAX may appear.
DENFAC All Example: If DENMAX = 60 then DENFAC will
DENFAC allow denominators 1, 2, 3, 4, 5, 6, 10, 12, 15, 20,
30, and 60. Now you know why 60 was a holy
number in ancient Babylon.

(0) Sets fixed denominator format, i.e. the one and


DENFIX DENFIX All only denominator allowed is the value set by
DENMAX.

(0) Works like /c in the HP-35S, but the maximum


denominator settable is 9,999. It will be set to this
DENMAX All
DENMAX value if x < 1 or x > 9,999 at DENMAX execution
time. For x = 1 the current setting is recalled.

(1) Takes a descriptor of a square matrix in X and


DET DET DECM returns the determinant of the matrix. The matrix
itself is not modified.

(0) Changes the number of decimals shown while


keeping the basic display format (FIX, SCI, ENG)
DISP DISP n All
as is. With ALL set, DISP will change the
switchover point (see ALL).

(2) Interprets x and y as Cartesian components of


C a first vector, and z and t as those of a second one,
DOT DECM
©DOT and returns [ ∙ z + y ∙ t , 0, …] , dropping two stack
levels.

WP 34S Owner’s Manual Edition 3.1 Page 82 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(-1) Assumes the source contains DP data and


dRCL ¬INPUT
dRCL s recalls them as such. See Appendix H.

DROP DROP ¬INPUT Drops x. See p. 20 for details.

C
DROP DECM Drops xc. See p. 31 for details.
©DROP
(0) Given cccccc.fffii in the source, DSE
decrements s by ii, skipping next program line if
then ccccccc ≤ fff . If s features no fractional
DSE s ¬INPUT part then fff is 0 and ii is set to 1. DSE does
not load L even for target address X.
Note that neither fff nor ii can be negative, and
DSE makes only sense with cccccc > 0 .

DSL DSL s ¬INPUT (0) Works like DSE but skips if ccccccc < fff .

(0) Decrements s by 1, and skips the next step if


DSZ DSZ s ¬INPUT s 1 thereafter. Does not load L even for target
address X. Known from the HP-16C.

D.MY D.MY All (0) Sets the format for date display.

(1) Takes x as a date in the format set and con-


D J D→J DECM
verts it to a Julian day number according to JG…

(1) See the catalog of conversions for conversions


D R DECM
from degrees to radians.
E

E3OFF E3OFF (0) Toggle the thousands separators for DECM


All
(points or commas depending on radix setting).
E3ON E3ON

(0) Last command in a routine and terminal for


END END ¬INPUT searching local labels as described on p. 67.
Works like RTN in all other aspects.

ENG n ¬INPUT (0) Sets engineer’s display format (see p. 40).

(0) Defines that numbers exceeding the range


ENGOVR ¬INPUT displayable in ALL or FIX will be shown in
engineer’s format. Compare SCIOVR.

(-1) Pushes x on the stack, disabling stack lift as


ENTER ¬INPUT
usual. See pp. 20 and 31 for details.

WP 34S Owner’s Manual Edition 3.1 Page 83 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Checks the entry flag. This internal flag is set if:
any character is entered in alpha mode, or
ENTRY? ENTRY? ¬INPUT any command is accepted for entry (be it via
, a function key, or with a partial
command line).

erf erf (1) Returns the error function or its complement.


DECM
See Appendix I for more.
erfc erfc

(0) Raises the error specified. The consequences


are the same as if the respective error really
ERR ERR n ¬INPUT occurred, so e.g. a running program will be
stopped. See Appendix C for the respective error
codes. Compare MSG.

EVEN? EVEN? ¬INPUT (0) Checks if x is integer and even.

ex ¬INPUT (1)

All See p. 120.

(0) Selects the exponential curve fit model


ExpF ExpF All
Rx a0 e a1x .

Expon Expon
(1) Exponential distribution with the rate in J. See
ExponP Exponₚ Appendix I for more.
DECM
Exponu Exponᵤ Expon–1 returns the survival time ts for a given
probability p in X, with in J.
Expon –1 Expon⁸

(1) Returns the exponent h of the number


EXPT EXPT DECM
displayed . Compare MANT.

(1) For x ≈ 0 , returns a more accurate result for


ex -1 eᵡ-1 DECM
the fractional part than eX does.

WP 34S Owner’s Manual Edition 3.1 Page 84 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)
F

(0) Sets the processor speed to ‘fast’. This is start-


FAST FAST All up default and is kept for fresh batteries. Compare
SLOW.

FB FB n Integer (1) Inverts (‘flips’) the specified bit in x .

FC? FC? n ¬INPUT (0) Tests if the flag specified is clear.

FC?C
FC?C n (0) Tests if the flag specified is clear. Clears, flips,
FC?F ¬INPUT
etc. or sets this flag after testing, respectively.
FC?S

FF FF n ¬INPUT (0) Flips the flag specified.

Integer (1) Returns the Fibonacci number.


See App. I
FIB FIB (1) Returns the extended Fibonacci for more.
DECM
number.

FILL ¬INPUT (0) Copies x to all stack levels. See p. 31 for CFILL.

FIX n ¬INPUT (0) Sets fixed point display format (see p. 40).

FLASH? FLASH? ¬INPUT (-1) Returns the number of free words in FM.

FLOOR FLOOR ¬INPUT (1) Returns the largest integer ≤ x .

FP ¬INPUT (1) Returns the fractional part of x .

FP? FP? ¬INPUT (0) Tests x for having a nonzero fractional part.

(0) Sets fraction mode like in the HP-35S, but


FRACT FRACT All
keeps the format as set by PROFRC or IMPFRC.

FS? FS? n ¬INPUT (0) Tests if the flag specified is set.

FS?C
FS?C n (0) Tests if the flag specified is set. Clears, flips, or
FS?F ¬INPUT
etc. sets this flag after testing, respectively.
FS?S

FP(x) Fₚ(x)
(1) Fisher’s F-distribution. F(x) equals 1 – Q(F) ,
Fu(x) Fᵤ(x) Fu(x) equals Q(F) and F–1(p) equals F p in the
DECM
HP-21S. The degrees of freedom are specified in J
F(x) F(x)
and K. See Appendix I for more.
F –1(p) F⁸(p)

WP 34S Owner’s Manual Edition 3.1 Page 85 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

Returns the first derivative of the function f(x) at


position x. This f(x) must be specified in a routine
starting with LBL label . On return, Y, Z, and T
will be cleared and the position x will be in L.
ATTENTION: f’(x) will look for a user routine
f ’(x) DECM labeled 'δx', returning a fixed step size dx in X. If
f’(x) label that routine is not defined, dx = 0.1 is set for
default (arbitrary, but it had to be specified).
Then, f’(x) will evaluate f(x) at ten points equally
spaced in the interval x ± 5 dx . If you expect any
irregularities within this interval, change dx to
exclude them.

Works like f’(x) but returns the second derivative.


f ”(x) DECM
f”(x) label The remarks above apply in analogy.
G

(2) Returns the Greatest Common Divisor of x


GCD GCD ¬INPUT
and y 52. This will always be positive.

(0) Clears the pixel at position x, y in the graphic


block starting at register address s. Valid ranges
gCLR gCLR s ¬INPUT
are 0 ≤ x ≤ w – 1 and 0 ≤ y ≤ h – 1 . Pixel 0, 0 is
top left. See gDIM for more.

gd g⒟ (1) Returns the Gudermann function or its inverse.


DECM
See Appendix I for details.
gd–1 g⒟⁸

(0) Initiates a set of registers (called a graphic


block) starting at address s, allowing for graphic
data featuring x (≤ 166) pixel columns and y pixel
rows. For x ≤ 0 , the width w will be set to 166. For
y ≤ 0 , the height h will be set to 8. The first two
bytes are reserved to take w and .
gDIM gDIM s ¬INPUT The number of registers needed for the set is
in startup standard mode. E.g. 21
registers are required for maximum width and
standard height.
The command can be exactly emulated in integer
mode by storing in the first register
and clearing the rest. See FLOOR and ⎙PLOT.

(-2) Assumes a graphic block starting at address s


gDIM? ¬INPUT
gDIM? s and recalls h and w for it. See gDIM for more.

52
GCD translates to “ggT” in German.
WP 34S Owner’s Manual Edition 3.1 Page 86 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)

Geom Geom (1) Geometric distribution: The cdf returns the


probability for a first success after m = x Bernoulli
GeomP Geomₚ experiments. The probability p0 for a success in
DECM each such experiment must be specified in J. See
Geomu Geomᵤ Appendix I for more.
Geom –1 returns the number of failures f before 1st
–1
Geom Geom⁸ success for given probabilities p in X and p0 in J.

gFLP gFLP s (0) Flips or tests the pixel at position x, y in the


¬INPUT graphic block starting at address s. See gCLR for
gPIX? gPIX? s more.

(0) Displays the top left sector of the graphic block


gPLOT ¬INPUT (starting at address s) in the dot matrix section of
gPLOT s the LCD. See gDIM for more.

GRAD DECM (0) Sets angular mode to gon (grad).

(1) Takes x as given in gon (grad) and converts


GRAD GRAD→ DECM
them to the angular mode currently set.

(0) Sets the pixel at position x, y in the graphic


gSET gSET s ¬INPUT
block starting at address s. See gCLR for more.

STO (0) Inserts an unconditional branch to label .


label
¬STO,
(0) Positions the program pointer to label .
¬INPUT

,
… to one of these labels, if defined.
, , or
(0)
GTO nnn Positions … to step nnn .
the program
pointer … … directly after previous END, going
¬INPUT to the top of current program.
(not
programm- … directly after next END, going to
able) the top of next program.

… to step 000, i.e. top of RAM.

(0) Takes the first three characters of alpha (or all


GTO GTOα ¬INPUT if there are fewer than three) as a label and
positions the program pointer to it.

WP 34S Owner’s Manual Edition 3.1 Page 87 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)
H

Hn Hₙ (2) Hermite polynomials for probability (Hn) and for


DECM
physics (Hnp). See Appendix I for details.
Hnp Hₙₚ

(1) Assumes X containing decimal hours or


degrees, and displays them converted in the format
H.MS DECM
hhhh°mm’ss.dd” temporarily as shown on p.
44.

H.MS+ H.MS+ (2) Assumes X and Y containing times or degrees


DECM in the format hhhh.mmssdd , and adds or
H.MS– H.MS- subtracts them, respectively.
I

(2) Integer division, working like in DECM


IDIV IDIV s ¬INPUT
and like in integer modes.

(1) Sets fraction mode allowing improper fractions


in display (e.g. 5/3 instead of 1 2/3). Converts x
according to the settings by DEN… into an
IMPFRC ¬INPUT
improper fraction if possible. Absolute decimal
equivalents of x must not exceed 100,000.
Compare PROFRC.

(0) Increments s by 1. Does not load L even for


INC INC s ¬INPUT
target address X.

INTM? INTM? ¬INPUT (0) Tests if your WP 34S is in an integer mode.

(0) Tests x for being an integer, i.e. having a


INT? INT? ¬INPUT
fractional part equal to zero. Compare FP?.

IP ¬INPUT (1) Returns the integer part of x .

(-1) Assumes the source contains integer data and


iRCL iRCL s ¬INPUT
recalls them as such. See p. 159.

ISE ISE s ¬INPUT (0) Works like ISG but skips if ccccccc ≥ fff .

(0) Given cccccc.fffii in the source , ISG


increments s by ii, skipping next program line if
then ccccccc > fff . If s has no fractional part
ISG s ¬INPUT then fff = 0 , of course, and ii is set to 1. ISG
does not load L even for target address X.
Note that neither fff nor ii can be negative, but
cccccc can.

(0) Increments s by 1, skipping next program line if


ISZ ISZ s ¬INPUT then s 1 . Does not load L even for target
address X. Known from the HP-16C.

WP 34S Owner’s Manual Edition 3.1 Page 88 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

Iβ Iβ
Returns the regularized (incomplete) beta or
IΓp IΓₚ DECM gamma function (one of two kinds). See Appendix I
for details.
IΓq IΓₛ
J

JG1582 JG1582 (0) These two commands reflect different dates the
Gregorian calendar was introduced in different
All
large areas of the world. D J and J D will be
JG1752 JG1752 calculated accordingly. See p. 39.

(1) Takes x as a Julian day number and converts it


J D J→D DECM
to a date according to JG... in the format selected
K

(0) Tests if a key was pressed while a program


was running or paused. If no key was pressed in
that interval, the next program step after KEY? will
KEY? KEY? s ¬INPUT be executed, else it will be skipped and the code of
said key will be stored in the address specified.
Key codes reflect the rows and columns on the
keyboard (see p. 71).

(-1) Assumes a key code in the address specified


(see KEY?). Checks this code and returns the key
type:
0 … 9 if it corresponds to a digit … ,
KTP? KTP? s ¬INPUT 10 if it corresponds to , , or ,
11 if it corresponds to , , or ,
12 if it corresponds to any other key.
May help in user interaction with programs.
L

(-1) See pp. 20 and 31 for details. In fact, this


LASTx ¬INPUT
command will be recorded as RCL L in programs.

(0) Identifies programs and routines for execution


LBL label STO and branching. Read more about specifying labels
on p. 67f.

(0) Tests for the existence of the label specified,


LBL? ¬INPUT
LBL? label anywhere in program memory. See LBL for more.

(2) Returns the Least Common Multiple of x


LCM LCM ¬INPUT
and y 53. This will always be positive.

(0) Takes x as a date in the format selected,


LEAP? LEAP? DECM
extracts the year, and tests for a leap year.

53
LCM translates to “kgV” in German.
WP 34S Owner’s Manual Edition 3.1 Page 89 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)

LgNrm LgNrm
(1) Lognormal distribution with ln x g specified
LgNrmP LgNrmₚ in J and ln in K. See xg and below. See
DECM Appendix I for details.
LgNrmu LgNrmᵤ LgNrm-1 returns x for a given probability p in X,
with µ in J and σ in K.
LgNrm –1 LgNrm⁸

(3) Solves a system of linear equations


. Takes a base register number in X, a vector
LINEQS LINEQS ¬INPUT
descriptor in Y, and a descriptor of a square matrix
in Z. Returns the filled in vector descriptor in X.

(0) Selects the linear curve fit model


LinF LinF All R x a0 a1 x .

(-1) Left justifies a bit pattern within its word size as


in the HP-16C: The stack will lift, placing the left-
justified word in Y and the count (number of bit-
LJ LJ Integer shifts necessary to left justify the word) in X.
Example for word size 8: 101102 LJ results in
x = 3 and y = 101100002 .

LN ¬INPUT (1) Returns the natural logarithm of x.

Ln Lₙ DECM (2) Laguerre polynomials. See Appendix I for more.

(1) For x ≈ 0 , this returns a more accurate result


LN1+x LN1+x DECM
for the fractional part than ln(x) does.

(3) Laguerre's generalized polynomials. See


Ln Lₙα DECM
Appendix I for more.

(2) Returns the natural logarithm of Euler’s Beta


LN β LNβ DECM
function. See β.

(1) Returns the natural logarithm of Γ(x) . Allows


also for calculating really large factorials.
Example: What is 5432! ?
LN Γ LNΓ DECM Remember Γ(x + 1) = x! So, entering 5433 LNΓ
10 returns 17 931.480374 being the
decadic logarithm of the result; now calling
returns its mantissa 3.0225535984. Thus
5432! = 3.02∙1017 931 .

Restores the entire backup from FM, i.e. executes


LOADP, LOADR, LOADSS, LOADΣ, and returns
LOAD LOAD ¬INPUT Restored. Not programmable. Compare SAVE.
See the commands mentioned below and
Appendix A for more.

WP 34S Owner’s Manual Edition 3.1 Page 90 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Loads the complete program memory from the


backup and appends it to the programs already in
LOADP LOADP ¬INPUT
RAM. This will only work if there is enough space –
else an error will be thrown. Not programmable.

(0) Recovers numbered general purpose registers


from the backup (see SAVE). Lettered registers will
LOADR LOADR ¬INPUT not be recalled. The number of registers copied is
the minimum of the registers held in the backup
and RAM at execution time.

(0) Recovers the system state from the backup.


LOADSS LOADSS ¬INPUT See Appendix B for more.

(0) Recovers the summation registers from the


LOADΣ LOADΣ ¬INPUT backup. Throws an error if there are none. See
Appendix B for more.

(0) Allocates n local registers (≤ 144) and 16 local


LocR LocR n ¬INPUT
flags for the current program. See p. 69.

(-1) Returns the number of local registers currently


LocR? LocR? ¬INPUT
allocated.

LOG10 ¬INPUT (1) Returns the logarithm of x for base 10.

LOG2 ¬INPUT (1) Returns the logarithm of x for base 2.

(0) Selects the logarithmic curve fit model


LogF LogF All R x a0 a1 ln x .

Logis Logis

LogisP Logisₚ
DECM (1) Logistic distribution with μ given in J and s in
Logisu Logisᵤ K. See Appendix I for details.

Logis –1 Logis⁸

LOGx x ¬INPUT (2) Returns the logarithm of y for the base x .

C (2) Returns the complex logarithm of z + i t for the


LOGx x DECM
complex base x + i y .

LZOFF LZOFF (0) Toggles leading zeros like flag 3 does in the
All
HP-16C. Relevant in bases 2, 4, 8, and 16 only.
LZON LZON

WP 34S Owner’s Manual Edition 3.1 Page 91 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(-2) Returns the parameters a1 and a0 of the fit


curve through the data points accumulated in the
summation registers, according to the curve fit
L.R. L.R. DECM model selected (see LINF, EXPF, POWERF, and
LOGF).
For a straight line (LINF), a0 is the y-intercept and
a1 the slope. See Appendix I for more.
M

(1) Returns the mantissa m of the number


MANT MANT DECM
displayed . Compare EXPT.

(-1) Work like MASKL and MASKR on the HP-16C,


MASKL but with the mask length (or its address) following
MASKL n the command instead of taken from X. Thus, the
Integer mask is pushed on the stack.
Example: For WSIZE 8, MASKL 3 returns a mask
MASKR word 111000002. Use it e.g. for extracting the three
MASKR n
most significant bits from an arbitrary byte by AND.

DECM Catalog. See p. 125.

MAX MAX ¬INPUT (2) Returns the maximum of x and y .

(-1) Returns the number of free words in program


MEM? MEM? ¬INPUT memory, taking into account the local registers
allocated.

MIN MIN ¬INPUT (2) Returns the minimum of x and y .

(1) Reflects the bit pattern in x (e.g. 000101112


MIRROR MIRROR Integer
would become 111010002 for word size 8).

(2) Returns y mod x (see p. 56 for examples).


MOD MOD ¬INPUT
Compare RMDR.

All Catalog. See p. 125.

(1) Assumes x containing a date in the format


MONTH MONTH DECM
selected and extracts the month.

(0) Takes a matrix descriptor x, a destination row


number y, a source row number z, and a real
MROW+× DECM number t. It multiples each element mzi of (X) by t
MROW+× and adds it to myi . The stack remains unchanged.
M.ROW+× is similar to PPC M3.

(0) Takes a matrix descriptor x, a row number y,


and a real number z. It multiples each element myi
MROW× MROW× DECM
of (X) by z. The stack remains unchanged.
M.ROW× is similar to PPC M2.

WP 34S Owner’s Manual Edition 3.1 Page 92 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Takes a matrix descriptor x and two row


numbers y and z. It swaps the contents of rows y
MROW MROW⇆ DECM
and z in (X). The stack remains unchanged.
M.ROW is similar to PPC M1.

(0) Throws the error message specified. This will


MSG MSG n ¬INPUT be a temporary message. See Appendix C for the
respective error codes. Compare ERR.

(3) Takes two matrix descriptors x and y, and a


real number z. Returns . Thus a
M+× M+× DECM scalar multiple of one matrix is added to another
matrix. The multiply adds are done in internal high
precision and results should be exactly rounded.

(0) Takes a descriptor of a square matrix in X and


M -1 M⁸ DECM
inverts the matrix in-situ. Doesn't alter the stack.

(1) Takes a matrix descriptor x 54 and returns a


value suitable for ISG or DSL looping in that
M-ALL M-ALL DECM matrix. The loop shall process all elements in (X).
The loop counter is for DSL if the descriptor was
negative and for ISG otherwise.

(2) Takes a matrix descriptor x (positive or


negative as for M-ALL) and a column number y.
M-COL M-COL DECM
Returns a loop counter for processing all elements
miy of that matrix column only.

(1) Takes a matrix descriptor x (as for M-ALL) and


returns a loop counter. The loop shall processes all
M-DIAG DECM
M-DIAG elements along the matrix diagonal, i.e. all
elements mii in (X).

(2) Takes a matrix descriptor x (as for M-ALL) and


M-ROW DECM a row number y. Returns a loop counter for
M-ROW processing all elements myi of that matrix row only.

(3) Takes two matrix descriptors y and z, and the


integer part of x as the base address of the result.
Returns . All calculations are done
M× M× DECM
in internal high precision (39 digits). The fractional
part of x is updated to match the resulting matrix –
no overlap checking is performed.

(2) Takes a matrix descriptor y and a base register


number x. Copies the matrix (Y) into registers
M.COPY DECM
M.COPY starting at Rx. Returns a properly formatted matrix
descriptor in X.

54
That descriptor will be saved in L as usual. This applies for the other matrix operations as well.
WP 34S Owner’s Manual Edition 3.1 Page 93 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)

M.DY M.DY All (0) Sets the format for date display.

Takes a matrix descriptor x and a register number


y. Returns the column that register represents in Y
M.IJ M.IJ DECM
and the row in X. M.IJ is similar to PPC M4.
Compare M.REG.

(1) Takes a descriptor of a square matrix in X.


Transforms (X) into its LU decomposition in-situ.
The value in X is replaced by a descriptor that
M.LU M.LU DECM defines the pivots that were required to calculate
the decomposition. The most significant digit is the
pivot for the first diagonal entry, the next most
significant for the second and so forth.

(3) Takes a matrix descriptor x, a row number y,


and a column number z. M.REG returns the
M.REG M.REG DECM
register number in X. It is similar to PPC M5.
Compare M.IJ.

(0) Tests a matrix descriptor x and returns true if


M.SQR? M.SQR? DECM
the matrix is square.
N

NAND NAND ¬INPUT (2) Works in analogy to AND. See there.

NaN? NaN? ¬INPUT (0) Tests x for being ‘Not a Number’.

(1) Counts bits set in x like #B does on the


nBITS nBITS Integer
HP-16C.

(1) Takes a matrix descriptor x and returns the


nCOL nCOL DECM
number of columns of (X).

(2) Returns the nearest machine-representable


number to x in the direction toward y in the mode
DECM
set 55. For x < y (or x > y ), this is the machine
NEIGHB NEIGHB successor (or predecessor) of x ; for x = y it is y.

(2) Returns x + 1 for x < y , y for x = y , or x – 1


Integer
for x > y .

NEXTP NEXTP ¬INPUT (1) Returns the next prime number greater than x.

NOP NOP ¬INPUT (0) ‘Empty’ step FWIW.

NOR NOR ¬INPUT (2) Works in analogy to AND. See there.

55
You may find NEIGHB useful investigating numeric stability. See NEIGHBOR in the HP-71 Math Pac.
WP 34S Owner’s Manual Edition 3.1 Page 94 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)

Norml Norml
(1) Normal distribution with an arbitrary mean µ
NormlP Normlₚ given in J and a standard deviation σ in K. See
DECM Appendix I for details.
Normlu Normlᵤ Norml–1 returns x for a given probability p in X,
with µ in J and σ in K.
Norml –1 Norml⁸

Integer (1) Inverts x bit-wise as on the HP-16C.


NOT
DECM (1) Returns 1 for x = 0 , and 0 for x ≠ 0 .

(1) Takes a matrix descriptor x and returns the


nROW nROW DECM
number of rows of (X).

(-1) Recalls the number of accumulated data


nΣ nΣ DECM
points.
O

ODD? ODD? ¬INPUT (0) Checks if x is integer and odd.

STO
(0) Inserts a step to turn your WP 34S off under
program control.
OFF
¬STO (0) Turns your WP 34S off.

OR ¬INPUT (2) Works in analogy to AND. See there.


P

(2) Returns the number of possible arrangements


of y items taken x at a time. No item occurs more
PERM ¬INPUT than once in an arrangement, and different orders
of the same x items are counted separately. See
Appendix I for the formula. Compare COMB.

Pn Pₙ DECM (1) Legendre polynomials. See App. I for more.

Poiss Poiss (1) Poisson distribution with the number of


successes g in X, the gross error probability p0 in
PoissP Poissₚ J, and the sample size n in K. The Poisson
DECM parameter n p0 is calculated automatically.
Poissu Poissᵤ See Poisλ below.
Poiss–1 returns the maximum number of successes
–1
Poiss Poiss⁸ m for a given probability p in X, p0 in J and n in K.

Poisλ Poisλ
(1) Poisλ works like Poiss but with λ in J and
PoisλP Poisλₚ without using K. See Appendix I for more.
DECM
Poisλu Poisλᵤ Poisλ–1 returns m for a given probability p in X and
λ in J.
Poisλ –1 Poisλ⁸

WP 34S Owner’s Manual Edition 3.1 Page 95 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Pops the local registers allocated to the current


PopLR PopLR ¬INPUT
routine without returning. See LOCR and RTN.

(0) Selects the power curve fit model


PowerF PowerF All a1
Rx a0 x .

(0) Copies the current program (from FM or RAM)


PRCL ¬INPUT and appends it to RAM, where it can then be
edited (see p. 72). Allows duplicating programs in
RAM. Will only work with enough space at
PRCL destination.
By recalling a library routine from FM, editing it,
CAT open and PSTO you may modify this part of the FM
library.

(0) Checks if the absolute value of the integer part


PRIME? PRIME? ¬INPUT of x is a prime. The method is believed to work for
integers up to 9∙1018.

DECM Catalog. See p. 125.

(1) Sets fraction mode like in the HP-35S, allowing


only proper fractions or mixed numbers in display.
Converts x according to the settings by DEN… into
PROFRC ¬INPUT
a proper fraction, e.g. 1.25 or 5/4 into 1 1/4 .
Absolute decimal equivalents of x must not exceed
100,000. Compare IMPFRC.

(0) Displays alpha and stops program execution.


PROMPT PROMPT ¬INPUT
See p. 69 for more.

(0) With a program running, refreshes the display


and pauses program execution for n ticks (see
PSE n ¬INPUT
TICKS), with 0 ≤ n ≤ 99 . The pause will terminate
early when you press a key.

(0) Copies the current program from RAM and


appends it to the FM library. Not programmable.
The program must have at least one LBL
statement with an alphanumeric global label
(preferably at its beginning). If a program with the
same label already exists in the library it will be
PSTO PSTO ¬INPUT
deleted first. Alphanumeric labels contained in FM
may be browsed by CAT (see p. 126) and called
by XEQ.
By recalling a library routine from FM via PRCL,
editing it, and PSTO you may modify this part of
the FM library.

WP 34S Owner’s Manual Edition 3.1 Page 96 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Assumes a key code in the address specified.


Stops program execution, takes said code and
puts it in the keyboard buffer resulting in immediate
PUTK PUTK s ¬INPUT
execution of the corresponding call. is
required to resume program execution. May help in
user interaction with programs.

¬INPUT Catalog. See p. 125.


R

RAD DECM (0) Sets current angular mode to radians.

(1) Takes x as radians and converts it to the


RAD RAD→ DECM
angular mode currently set.

(-1) Returns a random number between 0 and 1


DECM
like RAN does in the HP-42S.
RAN#
Integer (-1) Returns a random bit pattern for the word size set.

RCL s ¬INPUT (-1) See the addressing table above for CRCL.

(0) Recalls modes stored by STOM (see p. 37).


RCLM s ¬INPUT
No need pressing here.

Recalls 4 or 8 values from a set of registers


RCLS RCLS s ¬INPUT starting at address s, and pushes them on the
stack. This is the converse command of STOS.

RCL+ s (1) Recalls s, executes the specified operation and


pushes the result on the stack.
RCL– s
¬INPUT E.g. RCL–12 subtracts r12 from x and displays
RCL× s the result (acting like , but without
losing a stack level). In analogy, CRCL–12 sub-
RCL/ s tracts r12 from x and r13 from y.

RCL s
(1) RCL ( ) replaces x with the maximum
¬INPUT
RCL (minimum) of s and x.
s

(1) Rounds x to n decimal places (0 ≤ n ≤ 99),


RDP RDP n DECM taking the RM setting into account. See RM and
compare RSD.

RDX, All (0) Sets the decimal mark to a comma.


RDX,

DECM (0) Toggles the radix mark.

RDX.
RDX. All (0) Sets the decimal mark to a point.

REALM? REALM? ¬INPUT (0) Tests if your WP 34S is in real mode (DECM).

WP 34S Owner’s Manual Edition 3.1 Page 97 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Prepares your WP 34S for receiving data via


RECV RECV ¬INPUT
serial I/O. See SEND… and Appendix A for more.

(0) Specifies the number of global general purpose


registers wanted. With REGS 100 you get the
REGS REGS n All
default state (R00 – R99), REGS 0 leaves not
even a single such register for use.

(-1) Returns the number of global general purpose


REGS? REGS? ¬INPUT
registers allocated (0 … 100).

If confirmed, executes CLALL and resets all modes


to start-up default, i.e. 24h, 2COMPL, ALL 0,
DBLOFF, DEG, DENANY, DENMAX 0, D.MY,
RESET RESET ¬INPUT E3ON, LinF, LocR 0, LZOFF, PROFRC, RDX.,
REGS 100, SCIOVR, SEPON, SSIZE4, WSIZE 64,
and finally DECM. See these commands for more.
Not programmable.

(-1) Right justifies, in analogy to LJ on the HP-16C.


RJ RJ Integer Example: 1011002 RJ results in y = 10112 and
x = 2 . See LJ.

RL RL n (1) Works like n consecutive RLs / RLCs on the


HP-16C, similar to RLn / RLCn there. For RL,
Integer
0 ≤ n ≤ 63 . For RLC, 0 ≤ n ≤ 64 . RL 0 / RLC 0
RLC RLC n execute as NOP, but load L. See p. 56 for more.

(0) Sets floating point rounding mode. This is only


used when converting from the high precision
internal format to packed real numbers. It will not
alter the display nor change the behavior of
ROUND. The following modes are supported:
0: round half even: ½ = 0.5 rounds to next even
number (default).
RM RM n All 1: round half up: 0.5 rounds up (‘business-
man’s rounding’ 56).
2: round half down: 0.5 rounds down.
3: round up: rounds away from 0.
4: round down: rounds towards 0 (truncates).
5: ceiling: rounds towards +∞.
6: floor: rounds towards –∞.

(2) Returns the remainder of a division. Equals


RMDR ¬INPUT RMD on the HP-16C but works for real numbers as
well. See p. 56 for examples. Compare MOD.

56
Translates to “kaufmännische Rundung” in German.
WP 34S Owner’s Manual Edition 3.1 Page 98 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)

(-1) Returns the floating point rounding mode set.


RM? RM? ¬INPUT
See RM for more.

… display format like RND does


¬INPUT
in the HP-42S.
(1) Rounds x
ROUND
using the current
… denominator like RND does
FRC
in the HP-35S fraction mode.

ROUNDI ROUNDI ¬INPUT (1) Rounds x to next integer. ½ rounds to 1.

RR RR n (1) Works like n consecutive RRs / RRCs on the


HP-16C, similar to RRn / RRCn there. For RR,
Integer
0 ≤ n ≤ 63 . For RRC, 0 ≤ n ≤ 64 . RR 0 / RRC 0
RRC RRC n execute as NOP, but load L. See p. 56 for more.

(1) Rounds x to n significant digits, taking the RM


RSD RSD n DECM
setting into account. See RM and compare RDP.

(0) Last command in a typical routine. Pops the


local data (like PopLR) and returns control to the
calling routine in program execution, i.e. moves the
STO program pointer one step behind the XEQ
instruction that called said routine. If there is none,
RTN program execution halts and the program pointer is
set to step 000.

(0) Resets the program pointer to the beginning of


¬STO current program. If the current program is in FM,
the program pointer will be set to step 000 in RAM.

(0) Works like RTN, but moves the program pointer


RTN+1 RTN+1 ¬INPUT two steps behind the XEQ instruction that called
said routine. Halts if there is none.

(0) Interprets x in the form sss.nn . Clears nn


registers starting with address sss . If nn = 0 , it
will clear the maximum available.
Example: For x = 34.567 , R-CLR will clear R34
R-CLR R-CLR DECM through R89.
ATTENTION for nn = 0 : For ,
clearing will stop at the highest allocated global
numbered register. For , clearing
will stop at K. For , clearing will stop at
the highest allocated local register.

WP 34S Owner’s Manual Edition 3.1 Page 99 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Interprets x in the form sss.nnddd . Takes


nn registers starting with address sss and copies
their contents to ddd etc. If nn = 00 , it will take
the maximum available.
Example: For x = 7.03045678 , r07, r08, r09 will
R-COPY DECM be copied into R45, R46, R47, respectively.
R-COPY
For x < 0 , R-COPY will take nn registers from FM
instead, starting with register number |sss| .
Destination will be in RAM always.
ATTENTION: The advice at R-CLR applies, but
with ‘clearing’ replaced by ‘copying’.

(0) Interprets x in the form sss.nn . Sorts the


contents of nn registers starting with address sss .
If nn = 0 , it will sort the maximum available.
Example: Assume x = 49.0369 , r49 = 1.2 ,
R-SORT DECM
R-SORT r50 = –3.4 , and r51 = 0 ; then R-SORT will return
r49 = -3.4 , r50 = 0 , and r51 = 1.2.
ATTENTION: The advice at R-CLR applies, but
with ‘clearing’ replaced by ‘sorting’.

(0) Works like R-COPY but swaps the contents of


R-SWAP DECM
R-SWAP source and destination registers.

(1) See the catalog of conversions for conversions


R D DECM
of radians to degrees.

R
Rotates the stack contents one level up or down,
¬INPUT
respectively. See pp. 20 and 31 for details.
R
S

(-2) Takes the statistical sums accumulated,


calculates the sample standard deviations sy and
s DECM
sx and pushes them on the stack. See Appendix I
for the formula.

(0) Saves user program space, registers and


system state to FM, and returns Saved. Recall
SAVE SAVE ¬INPUT
your backup by the different flavors of LOAD. Not
programmable. See Appendix A for more.

SB SB n Integer (1) Sets the specified bit in x .

SCI n ¬INPUT (0) Sets scientific display format (see p. 40).

(0) Defines that numbers exceeding the range


SCIOVR ¬INPUT displayable in ALL or FIX will be shown in scientific
format. Compare ENGOVR, see RESET.

WP 34S Owner’s Manual Edition 3.1 Page 100 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

SDL SDL n (1) Shifts digits left (right) by n decimal positions,


DECM equivalent to multiplying (dividing) x by 10n .
SDR SDR n Compare SL and SR for integers.

SEED SEED DECM (0) Stores a seed for random number generation.

SENDA (0) SENDA sends all RAM data,


SENDP the program memory,
SENDP
SENDA SENDR the global general purpose registers, and
¬INPUT
etc. SENDΣ the summation registers, respectively,
SENDR
to the device connected via serial I/O.
SENDΣ See RECV and Appendix A for more.

SEPOFF All (0) Toggle the digit group separators for integers.
SEPOFF Points or commas will be displayed every …
Integer … four digits in bases 2 and 4,
… two digits in base 16,
SEPON
SEPON All … three digits in all other integer bases.

(-2) Takes the statistical sums accumulated,


SERR SERR DECM calculates and returns the standard errors (i.e. the
standard deviations of x and ).

(-1) Returns the standard error for weighted data,


SERRw SERR⒲ DECM
i.e. the standard deviation of xw.

SETCHN SETCHN All (0) Sets some regional preferences (see p. 39).

(0) Sets the date for the real time clock (the
SETDAT SETDAT All emulator takes this information from the PC clock).

SETEUR
SETEUR
SETIND All (0) Sets some regional preferences (see p. 39).
etc.
SETJPN

(0) Sets the time for the real time clock (the
SETTIM SETTIM All
emulator takes this information from the PC clock).

SETUK SETUK
All (0) Sets some regional preferences (see p. 39).
SETUSA etc.

SF n ¬INPUT (0) Sets the flag specified.

SHOW ¬INPUT Browser. See p. 125.

WP 34S Owner’s Manual Edition 3.1 Page 101 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(1) Returns 1 for x > 0 , –1 for x < 0 , and 0 for


SIGN SIGN ¬INPUT x = 0 or non-numeric data. Corresponds to
signum(x) for numeric input.

C
SIGN DECM (1) Returns the unit vector of x + i y in X and Y.
©SIGN
(0) Sets sign-and-mantissa mode for integers. See
SIGNMT SIGNMT All p. 53ff.

SIN DECM (1) Returns the sine of the angle in X.

sin x
SINC SINC DECM (1) Returns ; returns 1 for x = 0.
x

SINH DECM (1) Returns the hyperbolic sine of x.

(0) Skips n program steps forwards (0 ≤ n ≤ 255).


So e.g. SKIP 2 skips over the next two steps,
going e.g. from step 123 to step 126. If SKIP
attempts to cross an END, an error is thrown.
SKIP SKIP n ¬INPUT
ATTENTION: If you edit a section of your routine
that is crossed by one or more BACK, SKIP, or
CASE jumps, this may well result in a need to
manually maintain all those statements individually.

(1) Works like n (≤ 63) consecutive SLs on the


SL SL n Integer HP-16C. SL 0 executes as NOP, but loads L. See
p. 56 for more.

(0) Sets the processor speed to ‘slow’. This is also


SLOW SLOW All automatically entered for low battery voltage (see
p. 36). Compare FAST.

Solves the equation f(x) = 0 , with f(x) calculated


by the routine specified. Two initial estimates of the
root must be supplied in X and Y when calling
SLV. For the rest, the user interface is as in the
HP-15C. This means SLV returns xroot in X, the
SLV label DECM second last x-value tested in Y, and f(xroot) in Z.
It also means SLV acts as a test, so the next
program step will be skipped if SLV fails.
Please refer to the HP-15C Owner’s Handbook
(Section 13 and Appendix D) for more information
about automatic root finding.

WP 34S Owner’s Manual Edition 3.1 Page 102 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

Solves the quadratic equation ax 2 bx c 0 ,


with its real parameters on the input stack
[ c, b, a, …] , and tests the result.
b r
If r : b2 4ac 0 , SLVQ returns
2a
in Y and X. In a program, the step after SLVQ
will be executed.
SLVQ SLVQ DECM Else, SLVQ returns the real part of the first
complex root in X and its imaginary part in Y
(the 2nd root is the complex conjugate of the first
– see CONJ). If run directly from the keyboard,
the complex indicator C is lit then – in a
program, the step after SLVQ will be skipped.
In either case, SLVQ returns r in Z. Higher stack
levels are kept unchanged. L will contain equation
parameter c.

(-1) Returns the integer sign mode set, i.e.


2 (meaning ‘true’) for 2’s complement,
SMODE? ¬INPUT 1 (‘true’ again) for 1’s complement,
SMODE?
0 (i.e. ‘false’) for unsigned, or
-1 (i.e. ‘true’) for sign and mantissa mode.

SPEC? SPEC? ¬INPUT (0) True if x is ‘special’, i.e. infinite or non-numeric.

(1) Works like n (≤ 63) consecutive SRs on the


SR SR n Integer HP-16C. SR 0 executes as NOP, but loads L. See
p. 56 for more.

(-1) Assumes the source contains SP data and


sRCL sRCL s ¬INPUT
recalls them as such. See pp. 159 and 186.

Set the stack size to 4 or 8 levels, respectively.


SSIZE4 SSIZE4
See p. 17. Note register contents will remain un-
All changed in this operation. The same will happen if
stack size is modified by any other operation (e.g.
SSIZE8 SSIZE8
by RCLM, see p. 37).

SSIZE? SSIZE? ¬INPUT (-1) Returns the number of stack levels allocated.

DECM Catalog. See p. 125.

¬INPUT Browser. See p. 125.

STO s ¬INPUT (0) See p. 33 for CSTO.

(0) Stores mode settings for later use as described


STOM s ¬INPUT on p. 37 (no need pressing here). RCLM recalls
such data.

WP 34S Owner’s Manual Edition 3.1 Page 103 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

STO
(0) Stops program execution. May be used to wait
STOP
for input, for example.

Stopwatch application based on the real time clock


and following the timer of the HP-55. See p. 145
for a detailed description.
STOPW DECM
ATTENTION: This command will be part of the
STOPW function set only if the proper firmware is loaded.
See Appendix A.

(0) Stores all current stack levels in a set of 4 or 8


STOS STOS s ¬INPUT registers, starting at destination address s. See
RCLS.

STO+ s (0) Executes the specified operation on s and


stores the result into said address. See p. 33 for
STO– C
s STO...
¬INPUT
STO× s E.g. STO–12 subtracts x from r12 like the key-
strokes 12 12 would do, but
STO/ s the stack remains unchanged.

STO s
(0) STO ( ) takes the maximum (minimum) of s
¬INPUT
STO and x and stores it at the address specified.
s

(-2) Recalls the linear sums Σy and Σx . Useful


SUM SUM DECM
for elementary vector algebra in 2D.

DECM Catalog. See p. 125.

(-1) Calculates the standard deviation for weighted


sw s⒲ DECM data (where the weight y of each data point x was
entered via Σ+). See Appendix I for the formula.

(-1) Calculates the sample covariance for the two


sXY data sets entered via Σ+. It depends on the fit
sₓᵧ DECM
model selected. See Appendix I for the formula
See COV for the population covariance.

WP 34S Owner’s Manual Edition 3.1 Page 104 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)
T

TAN DECM (1) Returns the tangent of the angle in X.57

TANH DECM (1) Returns the hyperbolic tangent of x.

DECM Catalog. See p. 125.

(-1) Returns the number of ticks from the real time


clock at execution time. With the quartz crystal
built-in, 1 tick = 0.1 s. Without, it may be some 10%
TICKS TICKS ¬INPUT
more or less. So the quartz crystal is inevitably
required if the clock is to be useful in long range.
TICKS also works without the quartz, however.

(-1) Recalls the time from the real time clock at


execution, displaying it in the format hh.mmss in
TIME TIME DECM
standard 24h-mode. Choose FIX 4 for best
results.

(2) Chebychev polynomials of first kind. See


Tn Tₙ DECM
Appendix I for details.

(0) Returns false if TOP? is called in a


TOP? TOP? STO subroutine, true if the program-running flag is set
and the subroutine return stack pointer is clear.

(1) Takes a matrix descriptor x and returns the


descriptor of its transpose. The transpose is done
TRANSP DECM
TRANSP in-situ and does not require any additional
memory.

tP(x) tₚ(x)
(1) Student’s t distribution. The degrees of freedom
tu(x) tᵤ(x) are stored in J. tu(x) equals Q(t) and t –1(p)
DECM equals t p in the HP-21S. See p. 49 for an
t(x) t(x) application example and Appendix I for more
mathematical details.
t –1(p) t⁸(p)

t t⇆ s ¬INPUT Swaps t and s, in analogy to x .


U

(1) Returns 1 times the smallest power of ten


which can be added to x or subtracted from x to
ULP ULP ¬INPUT
actually change the value of x in your WP 34S in
the mode set. Thus 1 is returned in integer mode.

(2) Chebychev polynomials of second kind. See


Un Uₙ DECM
Appendix I for details.

(0) Sets unsigned mode like UNSGN does on the


UNSIGN UNSIGN All
HP-16C.

57
Note that TAN returns “not numeric” for x = ±90° or equivalents if flag D is set.
WP 34S Owner’s Manual Edition 3.1 Page 105 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)
V

VERS VERS ¬INPUT (0) Shows your firmware version and build number.

(0) Shows s until the next key is pressed. See p. 69


VIEW s ¬INPUT
for more.

VIEWα ¬INPUT (0) Displays alpha in the top row and - - - in the
VIEW bottom row until next key is pressed (compare to
INPUT AVIEW in the HP-42S). See p. 69 for more.

s INPUT (0) Displays alpha in the top row and s in the


VW + bottom row until the next key is pressed. See p. 69
VWα+ ¬INPUT for more.
W

(0) Displays credits to the brave men who made


WHO WHO ¬INPUT
this project work.

(1) Takes x as a date in the format selected,


returns the name of the day in the dot matrix and a
WDAY WDAY DECM
corresponding integer in the numeric display
(Monday = 1, Sunday = 7) 58.

Wm Wₘ (1) W p returns the principal branch of Lambert’s W


DECM for given x ≥ –1/e . Wm returns its negative branch.
Wp Wₚ See Appendix I for more.

W –1 W⁸ DECM (1) Returns x for given Wp ( ≥ -1). See there.

Weibl Weibl
(1) Weibull distribution with its shape parameter b
WeiblP Weiblₚ in J and its characteristic lifetime T in K. See
DECM Appendix I for details.
Weiblu Weiblᵤ Weibl –1 returns the survival time ts for a given
probability p in X, with b in J and T in K.
Weibl –1 Weibl⁸

(0) Works almost like on the HP-16C, but with the


All (but parameter 1 ≤ n ≤ 64 trailing the command
becomes instead of taken from X. Reducing the word size
truncates the values in the stack registers
effective
WSIZE WSIZE n in integer employed and in L. All other memory content stays
as is (see Appendix H for details). Increasing the
modes word size will add empty bits to each stack level.
only) WSIZE 0 sets the word size to maximum, i.e. 64
bits.

WSIZE? WSIZE? ¬INPUT (-1) Recalls the word size set.

58
These numbers correspond to Chinese weekdays 1 to 6 directly. For Portuguese weekdays (‘segunda
feira’ etc.), add 1 to days 1 to 5.
WP 34S Owner’s Manual Edition 3.1 Page 106 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)
X

x2 ¬INPUT (1)
3
x x³ ¬INPUT (1)

STO (0) Calls the subroutine with the label specified.


label
¬STO,
(0) Executes the program with the label specified.
¬INPUT
XEQ
, , , or (0) Calls the respective subroutine, so e.g. XEQ C
STO
will be inserted when is pressed.
(you may need for
reaching these hotkeys ¬STO,
in integer bases >10.) ¬INPUT (0) Executes the respective program if defined.

(0) Takes the first three characters of alpha (or all


XEQ XEQα ¬INPUT if there are fewer than three) as a label and calls or
executes the respective routine.

XNOR XNOR ¬INPUT (2) Works in analogy to AND. See there.

XOR ¬INPUT (2) Works in analogy to AND. See there.

(0) Tests for presence of the crystal necessary for


XTAL? XTAL? ¬INPUT
a precise real time clock (think of Xmas).

(-2) Returns the arithmetic means of the x- and y-


DECM
data accumulated. See also s, SERR, and .

(-2) Returns the geometric means of the data


xg Ẁg DECM accumulated. See Appendix I for the formula. See
also , m , and P .

(-1) Returns the arithmetic mean for weighted data


(where the weight y of each data point x was
xw Ẁ⒲ DECM
entered via ). See Appendix I for the formula.
See also sw and SERRw .

(1) Returns a forecast x for a given y (in X)


x ẋ DECM
following the fit model chosen. See L.R. for more.

All Catalog. See p. 125.

DECM (1) Returns Γ(x + 1) .


x!
Integer (1) Returns the factorial n! .

(0) Interprets x as a character code. Appends the


x x→α All respective character to alpha, similar to XTOA in
the HP-42S.

WP 34S Owner’s Manual Edition 3.1 Page 107 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

Swaps x and s, in analogy to x y . See p. 33 for


x s ¬INPUT C
x . Will be listed like x⇆ J, x⇆.12, x⇆→12, etc.

Swaps the stack contents x and y, performing


x Y ¬INPUT Re Im if a complex operation was executed
immediately before.

C Swaps the complex stack contents xc and yc . See


x Z ¬INPUT
p. 31 for details.

x< ? x<_? s

x≤ ? x≤_? s (0) Compare x with s. E.g. x< ?


compares x with k , and will be listed as x<? K in
x= ? s a program. See the examples on p. 26 for more.
x ≈ ? will be true if the rounded values of x and s
x = +0 ? x=+0?
are equal (see ROUND).
x = –0 ? x=-0? ¬INPUT The signed tests x = + 0 ? and x = – 0 ? are
meant for integer modes 1COMPL and SIGNMT,
x≈ ? x≈_? s and for DECM if flag D is set. In all these cases,
x≠ ? s e.g. 0 divided by –7 will display –0.
s and s compare
x≥ ? x≥_? s the complex number x + i y as explained on p. 33.
x> ? x>_? s

ᵡ√y ¬INPUT (2)


Y

(1) Assumes x containing a date in the format


YEAR YEAR DECM
selected and extracts the year.

¬INPUT (2) In integer modes, x must be ≥ 0.

yx ¬(INPUT,
-3, -4, (2) Shortcut working if label C is not defined.
-5, h)

(1) Returns a forecast y (in X) for a given x


ŷ DECM
following the fit model chosen. See L.R. for more.

Y.MD Y.MD All (0) Sets the format for date display.

y y⇆ s ¬INPUT Swaps y and s, in analogy to x .


Z

z z⇆ s ¬INPUT Swaps z and s, in analogy to x .

WP 34S Owner’s Manual Edition 3.1 Page 108 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)
α

Turns on alpha mode for keyboard entry of alpha


constants. INPUT is set and the previous program
step stays displayed until a character is entered.
STO &
Each such character (e.g. ‘?’) will be stored in one
¬INPUT program step (such as α ? here) and will be
appended to alpha in program execution. To start
a new string, use CL first.

Turns on alpha group mode for direct entry of up to


three characters in one program step taking two
words. Your WP 34S will display α' in the top line.
Now enter the characters you want to append to
alpha. To start a new string, use CL before.
STO & Example: Entering
INPUT will result in two program
steps stored:
α'Tes'
α't 1'
and Test 1 appended to alpha in program
execution.59

¬STO, Enters alpha mode for appending characters to


¬INPUT alpha. To start a new string, use CL first.

¬STO &
Leaves alpha mode.
INPUT

(0) Takes x as a date and appends it to alpha in


the format set. See DATE.
DATE αDATE ¬integer To append a date stamp to alpha, call DATE
DATE. For a short European date stamp, set
FIX 2, RDX. and call DATE RC# X.

(0) Takes x as a date, recalls the name of the


DAY αDAY ¬integer respective day and appends its first three letters to
alpha.

(0) Interprets s as character code. Takes the first


three characters of the converted code (or all if
GTO αGTO s ¬INPUT
there are fewer than three) as an alpha label and
positions the program pointer to it.

(0) Appends the integer part of x to alpha, similar


IP αIP All
to AIP in the HP-42S.

59
Note alpha group mode is left automatically after three characters are put in, so it must be called again for
continuation. Some characters must not be entered at third position (see Appendix E).
WP 34S Owner’s Manual Edition 3.1 Page 109 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)

(-1) Returns the number of characters found in


LENG αLENG All
alpha, like ALENG in the HP-42S.

(0) Takes x as a date, recalls the name of the


MONTH ¬integer respective month and appends its first 3 letters to
αMONTH
alpha.

OFF αOFF
(0) Work like AOFF and AON in the HP-42S,
¬INPUT
turning alpha mode off and on.
ON αON

s INPUT
(0) Interprets s as characters and appends them to
RCL
αRCL s ¬INPUT alpha.

(0) Interprets s as a number, converts it to a string


in the format set, and appends this to alpha.
RC# αRC# s All Example: If s is 1234 and ENG 2 and RDX. are
set, then 1.23e3 will be appended.
See also DATE for an application.

(0) Rotates alpha by n characters like AROT in the


HP-42S, but with n ≥ 0 and the parameter trailing
RL αRL n All
the command instead of taken from X. RL 0
executes as NOP, but loads L.

RR αRR n All (0) Works like RL but rotates to the right.

(0) Shifts the n leftmost characters out of alpha,


SL αSL n All like ASHF in the HP-42S. SL 0 equals NOP, but
loads L.

All (0) Works like SL but takes the s rightmost


SR αSR n
characters instead.

s INPUT (0) Stores the first (i.e. leftmost) 6 characters of


STO alpha in the destination specified. Stores all if there
αSTO s ¬INPUT are fewer than six.

(0) Takes x as a decimal time and appends it to


alpha in the format hh:mm:ss according to the
TIME αTIME ¬integer time mode selected. See 12h, 24h, and TIME.
To append a time stamp to alpha, call TIME
TIME.

(0) Interprets s as character code. Takes the first


three characters (or all if there are fewer than
XEQ αXEQ s ¬INPUT
three) of the converted code as an alpha label and
calls or executes the respective routine.

WP 34S Owner’s Manual Edition 3.1 Page 110 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(-1) Returns the character code of the leftmost


x α→x All character in alpha and removes this character from
alpha, like ATOX in the HP-42S.
β

x y
(2) Returns Euler’s Beta B x, y with
x y
β DECM
Re x 0 , Re y 0 . Called here to avoid
ambiguity.
γ

Γ Γ DECM (1) Returns Γ(x) . Additionally, calls x 1.


See also LNΓ.

γXY γₓᵧ
(2) Returns the lower or upper incomplete gamma
DECM
function, respectively. See Appendix I for details.
ΓXY Γₓᵧ
δ

(2) Assumes X and Y containing dates in the


ΔDAYS ΔDAYS DECM format chosen and calculates the number of days
between them like in the HP-12C.

x y
Δ% DECM (1) Returns 100 like %CH in the HP-42S.
y
ε

(-2) Returns the scattering factors εy and εx for log-


normally distributed sample data. This εx works for
ε DECM the geometric mean xg in analogy to the standard
deviation s for the arithmetic mean x but multipli-
cative instead of additive. See Appendix I for more.

DECM (-2) Works like but returns the scattering factors


m εₘ
of the two geometric means.

DECM (-2) Works like but returns the scattering factors


P εₚ
of the two populations.
ζ

ζ ζ DECM (1) Returns Riemann’s Zeta. See App. I for more.


π

π DECM (-1) Recalls .


C
π DECM (-2) Recalls into X and clears Y.

WP 34S Owner’s Manual Edition 3.1 Page 111 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(1) Computes a product using the routine specified.


Initially, X contains the loop control number in the
format cccccc.fffii , and the product is set to
1. Each run through the routine specified by s
Π label DECM computes a factor. At its end, this factor is
multiplied with said product; the operation then
decrements cccccc by ii and runs said routine
again if then cccccc ≥ fff , else returns the
resulting product in X.
σ

(-2) Works like s but returns the standard


σ DECM deviations of the two populations instead. See
Appendix I for the formula.

(1) Computes a sum using the routine specified.


Initially, X contains the loop control number in the
format cccccc.fffii , and the sum is set to 0.
Σ label DECM Each run through the routine specified by s
computes a summand. At its end, this summand is
added to said sum; the operation then decrements
cccccc by ii and runs said routine again if then
cccccc ≥ fff , else returns the resulting sum.

Σln2x

Σln2y (-1) Recall the respective statistical sums. These


sums are necessary for curve fitting models
Σlnx beyond pure linear. Calling them by name
Σln²x significantly improves program readability. Note
Σlnxy DECM these sums are stored in special registers in your
etc.
WP 34S.
Σlny
ATTENTION: Depending on input data, some or all
Σxlny of these sums may become non-numeric.

Σylnx

(-1) Works like sw but returns the standard


w σ⒲ DECM deviation of the population instead. See Appendix I
for the formula.

Σx

Σx2
(-1) Recall the respective statistical sums. These
2
Σx y sums are necessary for basic statistics and linear
Σx
DECM curve fitting. Calling them by name significantly
Σxy etc. improves program readability. Note these sums are
stored in special registers of your WP 34S.
Σy

Σy2

WP 34S Owner’s Manual Edition 3.1 Page 112 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

Adds a data point to the


DECM
statistical sums.
Both functions disable
Σ+
stack lift as usual.
Shortcut working if label
DECM Both may be used for
A is not defined.
2D vector adding and
subtracting as well.
Subtracts a data point
Σ– DECM
from the statistical sums.
φ

(1) Standard normal error probability, equaling Q


Φu(x) Φᵤ(x) DECM in the HP-32E and Q(z) in the HP-21S.

φ(x) φ(x) DECM (1) Standard normal pdf.


See Appendix I for
(1) Standard normal cdf. the formula.
Φ(x) Φ DECM

(1) Equals Q-1 in the HP-32E and zp in the


Φ –1(p) Φ DECM
HP-21S.
χ

2
χ²
2
(1) Chi square distribution. The cdf 2 (with its
INV χ²INV degrees of freedom given in J) equals 1 – Q( 2) ,
DECM 2 2
2
χ²ₚ u equals Q( 2) and 2
INV equals p in the
P
HP-21S. See Appendix I for more.
2
u χ²ᵤ
The Rest

(-1) X (-1)ᵡ ¬INPUT (1) If x is not a natural number, returns cos x .


¬INPUT (2) Returns y + x .
+
(2) Returns [ x + z , y + t , … ] .
DECM
May be used for adding 2D vectors as well.

¬INPUT (2) Returns y – x .



(2) Returns [ x – z , y – t , … ] .
DECM
May be used for subtracting 2D vectors as well.

¬INPUT (2) Returns y · x .


×
(2) Returns [ x · z – y · t , x · t + z · y , … ] .
DECM
Look at CROSS or DOT for multiplying 2D vectors.

×MOD ×MOD Integer (3) Returns (z ∙ y) mod x for x > 1 , y > 0 , z > 0 .

WP 34S Owner’s Manual Edition 3.1 Page 113 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

DECM (2) Returns y / x .

Integer (2) Returns y div x . Compare IDIV.


/

DECM (2) Returns [ , , …].

(1) ‘Unary minus’, corresponding to x · (-1) or


+/– ¬INPUT
xc · (-1) , respectively.

(3) Assumes the three components of a date (year,


month, and day) supplied on the stack in proper
DATE →DATE DECM
order for the date format selected and converts
them to a single date in x. Thus inverts DATE .

(1) Takes x as an angle in the angular mode


DEG DECM currently set and converts it to degrees. Prefix
may be omitted.

GRAD DECM (1) Works like DEG, but converts to gon (grad).

(1) Takes x as hours or degrees in the format


hhhh.mmssdd as in vintage HP calculators and
H DECM
converts it to a decimal time or angle, allowing for
using standard arithmetic operations then.

(1) Takes x as decimal hours or degrees and


converts it to the format hhhh.mmssdd as in
H.MS DECM
vintage HP calculators. For calculations, use
H.MS+ or H.MS– then.

Assumes X and Y containing 2D Cartesian


coordinates (x , y) of a point or components of a
POL DECM vector and converts them to the polar coordinates /
components (r , ) with the radius r x2 y2

RAD DECM (1) Works like DEG, but converts to radians.

Assumes X and Y containing 2D polar coordinates


REC DECM (r , ) of a point or components of a vector and
converts them to the Cartesian coordinates or
components (x , y) .

WP 34S Owner’s Manual Edition 3.1 Page 114 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

Shuffles the contents of the bottom four stack


levels at execution time. Examples:
⇆XXYZ works like ENTER (but does not
disable stack lift!),
⇆YZTX works like R ,
⇆ZTXY works like Cx y,
⇆____ ¬INPUT
but ⇆ZZZX is possible as well.

ATTENTION: This is a very powerful command


though it does not look it. Remember it will not
affect the higher levels in an 8-level stack. If you
play with this command, you may lose some level
contents and can easily make a mess of the stack.

x y
% DECM (1) Returns , leaving Y unchanged.
100
x y
(2) Returns the margin 60 100 in % for a
%MG %MG DECM x
price x and cost y , like %MU-Price in the HP-17B.

(3) Returns the mean rate of return in percent per


1
z
period, i.e. 100 x 1 with x = future
%MRR %MRR DECM y
value after z periods, y = present value.
For z = 1 , Δ% returns the same result easier.

x
%T %T DECM (2) Returns 100 , interpreted as % of total.
y
%Σ x
%Σ DECM (1) Returns 100 .
%Σ x

(2) Calculates a sales price by adding a margin of


x % to the cost y , as %MU-Price does in the
HP-17B.
y
Formula: p sale
x
1
%+MG %+MG DECM 100
You may use %+MG for calculating net amounts
as well. Just enter a negative percentage in x.
Example: Total billed = 221,82 €, VAT = 19%.
What is the net?
221,82 19 returns 186,40.

60
Margin translates to „Handelsspanne“ in German.
WP 34S Owner’s Manual Edition 3.1 Page 115 of 211
Name Keys to press in modes Remarks (see p. 74 for general information)

¬INPUT (1)

√‾ ¬(INPUT,
-4, -5, (1) Shortcut working if label D is not defined.
h)

Integrates the function given in the routine


specified. Lower and upper integration limits must
be supplied in Y and X, respectively. Otherwise,
∫ label DECM the user interface is as in the HP-15C.
Please turn to the HP-15C Owner’s Handbook
(Section 14 and Appendix E) for more information
about automatic integration and some caveats.

∞? ∞? ¬INPUT (0) Tests x for infinity.

(3) Returns (z y ) mod x for x > 1 , y > 0 , z > 0 .

^MOD ^MOD Integer Example:


73 55 31
returns 26.

1
1 1
(2) Returns , being very useful in
|| DECM x y
electrical engineering especially. Returns 0 for x or
y being zero.

(0) Prints the current contents of the print buffer


plus a linefeed.
ATTENTION: Any printing will only work with an
⎙ADV ⎙ADV ¬INPUT hardware modification (see Appendix H) or using
the WP 34S emulator in combination with a printer
emulator (see Appendix D). Else print commands
will be ignored. – The printer will actually print only
when a line feed is sent to it.

(0) Sends a single character (with the code


specified) to the printer. Character codes n > 127
⎙CHR ¬INPUT
⎙CHR n can only be specified indirectly. ⎙MODE setting
will be honored. See ⎙ADV.

(-1) Sends the graphic block starting at address s


⎙PLOT ¬INPUT to the printer. If its width is 166, the data will be
⎙PLOT s
trailed by a line feed. See ⎙ADV and gDIM.

WP 34S Owner’s Manual Edition 3.1 Page 116 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Prints the register specified and the next one,


i.e. prints an entire complex number. A semicolon
will separate both components. Works like ⎙r
otherwise.
⎙CrXY ⎙©rₓᵧ s ¬INPUT
Example: Assume ⎙MODE 1, SCI 1, x = –1.2 and
y = 0.34 . Then the output of ⎙©rₓᵧ X would
look like
-1.2e0 ; 3.4e-1
(0) Sets a delay of s ticks (see TICKS) to be used
⎙DLAY All
⎙DLAY n with each line feed on the printer.

(0) Sets print mode. Legal print modes are:


0: Use the printer font and character set wherever
possible (default). All characters feature the
same width (5 columns + 2 columns spacing).
1: Use the variable pitch display font, resulting in
some jitter on the printout but packing more
⎙MODE All characters in a line.
⎙MODE n
2: Use the small display font, which allows for
packing even more info in a line.
3: Send the output to the serial channel. Works for
plain ASCII only – no characters will be
translated. Line setup is the same as for serial
communication: 9600 baud, 8 bits, no parity.

(0) Prints the listing of the current program, one


⎙PROG ⎙PROG ¬INPUT line per step. The current program is the one the
program pointer is in at execution time. See ⎙ADV.

(0) Prints the register specified, right adjusted,


⎙r ⎙r s ¬INPUT without labeling the output. If you want a heading
label, call ⎙α+ first or use ⎙REGS. See ⎙ADV.

(1) Interprets x in the form sss.nn . Prints the


contents of nn registers starting with number sss .
Each register takes one line starting with a label.
ATTENTION for nn = 0 :
For , printing will stop at the highest
⎙REGS ⎙REGS ¬INPUT
allocated global numbered register.
For , printing will stop at K.
For , printing will stop at the highest
allocated local register.
See also ⎙ADV.

(0) Prints the stack contents. Each level prints in a


⎙STK ⎙STK ¬INPUT
separate line starting with a label. See ⎙ADV.

WP 34S Owner’s Manual Edition 3.1 Page 117 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

(0) Positions the print head to print column n (0 to


165, where n > 127 can only be specified
indirectly). Useful for formatting (in ⎙MODE 1 or 2
⎙TAB ¬INPUT
⎙TAB n in particular). Allows also for printer plots. If n is
less than current position, a linefeed will be
entered to reach the new position. See ⎙ADV.

(-1) Returns the number of print columns alpha


would take in the print mode set. See ⎙ADV and
⎙MODE.
⎙WIDTH ⎙WIDTH ¬INPUT
Second use: in ⎙MODE 1 or 2, ⎙WIDTH returns
the width of alpha in pixels (including the last
column being always blank) in the specified font.

(0) Appends alpha to the print line, trailed by a line


⎙α ⎙α ¬INPUT
feed. Compare ⎙α+ and ⎙+α. See ⎙ADV.

(0) Sends alpha to the printer without a trailing line


feed, allowing further information to be appended
⎙α+ ⎙α+ ¬INPUT
to this line. May be repeated. See also ⎙ADV, ⎙r
and ⎙+α.

(0) Prints the summation registers. Each register


⎙Σ ⎙Σ ¬INPUT
prints in one line starting with a label. See ⎙ADV.

(0) Appends alpha to the print line, adjusted to the


right and trailed by a line feed. Compare ⎙α and
⎙α+. See ⎙ADV.
Example: The following program section
CLα
α'Lef'
⎙+α ⎙+α ¬INPUT α t
⎙α+
CLα
α'Rig'
α'ht'
⎙+α
will print, if ⎙MODE 1 is set:
Left Right
(0) Tests if the quartz crystal and the necessary
⎙? ⎙? ¬INPUT
firmware are installed for printing.

(0) Sends a single byte, without translation, to the


printer (e.g. a control code). n > 127 can only be
⎙# ⎙# n ¬INPUT
specified indirectly. ⎙MODE setting will not be
honored. See ⎙ADV.

Inserts an integer 0 ≤ n ≤ 255 in a single step,


# # n STO
thus saving up to two steps and an ENTER.

WP 34S Owner’s Manual Edition 3.1 Page 118 of 211


Name Keys to press in modes Remarks (see p. 74 for general information)

C # n Works like # but also clears y. The shortcut works


# DECM
only for 1 ≤ n ≤ 9 .
n

Nonprogrammable Control, Clearing and Information Keys

Besides the commands CLALL, CLPALL, GTO., LOAD, LOADP, PSTO, RESET, and
SAVE, the following keystrokes cannot be programmed under the conditions stated:

Key(s) pressed in modes Remarks

In CONV Inverts the current conversion (see p. 139).

Asking for Answers the question Sure? with N for ‘no’. Any other response
confirmation except =Y, or for ‘no’ will be ignored.

Will be honored as described in the respective sections. See the


In a browser
note below this list.

CAT, SHOW, Will be honored as described in the respective sections. See the
STOPW note below this list.

↕ Selects the current item like below.


61

INPUT Leaves alpha mode.

Else Calls the programmable command ENTER described in the IOP.

61
The mode conditions specified here will be checked top down for this command at execution time:
If CAT or STOPW is open, it will be served as described there;
else if there is an open catalog, the current item will be selected;
else if alpha mode is set, it will be left;
else the programmable command ENTER will be executed or inserted.
This method holds for all commands listed here using this triangular symbolic.
WP 34S Owner’s Manual Edition 3.1 Page 119 of 211
Key(s) pressed in modes Remarks

Clears the temporary message displayed (see p. 38) returning to


¬RPN
the calculator state as was before that message was thrown.

Asking for Answers the question Sure? with N for ‘no’. Any other response
confirmation except =Y, or for ‘no’ will be ignored.

In STOPW Leaves the application (see p. 145).

In a catalog Leaves the current catalog or browser without executing anything,


or browser returning to the WP 34S status as it was before.

Command Cancels the execution of the pending operation, returning to the


input pending WP 34S status as it was before.

RCL Stops the running program like below.

STO Leaves programming mode like below.

INPUT Leaves alpha mode like above.

Else Does nothing.

Calculator off Turns your WP 34S on.

-key combinations are found in the appendices. Most impor-


Else
tant is + or for adjusting display contrast.

¬INPUT Toggles programming mode (STO).

Asking for Answers the question Sure? with Y for ‘yes’. Any other response
confirmation except or or for ‘no’ will be ignored.

Will be honored as described in the respective sections. See the


CAT/STOPW
note below this list.

RCL
Stops program execution immediately. Stopped will be
displayed until the next keystroke.

Runs the current program or resumes its execution starting with


¬(STO, INPUT)
the current step.

INPUT Appends a Y or y to alpha.

STO Enters the command STOP described in the IOP.

Will be honored as described in the respective section. See the


In CAT
note below this list.

Selects the item currently displayed and exits, executing the


↕ respective command.

Else Calls the programmable command XEQ described in the IOP.

WP 34S Owner’s Manual Edition 3.1 Page 120 of 211


Key(s) pressed in modes Remarks

Clears the temporary message displayed (see p. 38) returning to


¬RPN
the calculator state as was before that message was thrown.

Asking for Answers the question Sure? with N for ‘no’. Any other response
confirmation except =Y, or for ‘no’ will be ignored.

In STOPW Resets the timer (see p. 145).

In a catalog or
Leaves the catalog or browser like above.
browser

Command Deletes the last digit or character put in. If there is none yet,
input pending cancels the pending command like above.

INPUT Deletes the rightmost character in alpha.

STO Deletes current program step.

Else Calls the programmable command CLx described in the IOP.

Shifts the display window to the left / right like in the HP-16C.
Integer
Helpful while working with small bases. See p. 53.
/
Shows the full mantissa and the full exponent until the next key is
DECM
pressed (see p. 40).62 Compare SHOW in previous calculators.

Shows x as an integer to base 2, 8, 10, or 16, respectively.


DECM Returns to the base set with the next keystroke. Prefix may be
omitted here.

↕ Calls the character ⎙.


INPUT Toggles upper and lower case (the latter is indicated by ).

Else Calls the programmable command ⎙r X (see there).

62
These keys act differently in DP mode as described in Appendix H below.
WP 34S Owner’s Manual Edition 3.1 Page 121 of 211
Key(s) pressed in modes Remarks

In a browser Will be honored as described in the respective sections. See the


or STOPW note below this list.

↕ Goes to previous / next item in the current catalog.

Scrolls the display window six characters to the left/right in alpha


if possible. If fewer than six characters are beyond the limits of
/ INPUT
the display window on this side, the window will be positioned to
the beginning/end of string. Useful for longer strings.

Acts like the command BST / SST in the HP-42S. I.e. browses
programs in programming mode, where / will repeat with
Else 5Hz when held down for longer than 0.5s. – Out of programming
mode, SST will also execute the current program step, but the
keys will not repeat.

See the section about catalogs and browsers (beginning on p. 125) to learn more about
them. Note that the browsers CAT, SHOW, STATUS, and the application STOPW (see p.
145f) use some of the keys above as well as some more for particular control purposes
(e.g. , , , and some digit keys). Check the respective entries below.

WP 34S Owner’s Manual Edition 3.1 Page 122 of 211


Alphanumeric Input

Character Keys to press in modes Remarks

␣ INPUT Appends a blank space to alpha.

Separates degrees or hours from minutes and


° DECM seconds, so input format is hhhh.mmssdd. The
user has to take care where an arbitrary real number
represents such an angle or time.

Standard numeric input. For integer bases <10,


¬INPUT input of illegal digits is blocked. Note you cannot
enter more than 12 digits in the mantissa.

0…9 Register input. See the tables above for the number
in addressing
ranges.

, , ,
INPUT Appends the respective digit to alpha.
…,
… -1, -2, -3, Numeric input for digits >10. See pp. 52ff for more.
A…F
(grey print) -4, -5, h

Register input. See the virtual keyboard on p. 25 for


in addressing
the letters applicable.

A…Z
(grey print) Appends the respective Latin letter to alpha. Use
INPUT
to toggle between cases.

DECM &
E
¬FRC Works like in the Pioneers.

DECM &
i Enters complex number i , i.e. x = 0 and y = 1 .
¬FRC

… Appends the respective Greek letter to alpha. Use


A…Ω INPUT
(grey print) to toggle between cases. See p. 62 for more.

( (

) )

+ INPUT Appends the respective symbol to alpha.

WP 34S Owner’s Manual Edition 3.1 Page 123 of 211


Character Keys to press in modes Remarks

A second in input turns to fraction mode and is


interpreted as explained below. Note you cannot
DECM
enter after you entered twice – but you may
Second delete the second dot while editing the input line.
/
FRC First is interpreted as a space, second as a frac-
tion mark. See p. 41 for some examples.

INPUT Appends a slash to alpha.

+/- ¬INPUT Works like in the Pioneers.

, INPUT Appends the respective symbol to alpha.

‘.’ or ‘,’ DECM Inserts a radix mark as selected.


INPUT Appends the respective symbol to alpha.

&

| INPUT Appends a vertical separator to alpha.

Enters the print character for fast access to the


⎙ In a catalog
respective commands (see p. 133).

WP 34S Owner’s Manual Edition 3.1 Page 124 of 211


CATALOGS AND BROWSERS

Due to the large set of operations your WP 34S features, most of them are stored in
catalogs as was discussed from p. 22 on. Numeric constants (in CONST and CONV) or
characters (in the alpha catalogs) are other items contained in catalogs. Opening a
catalog on your WP 34S will set alpha mode to allow for typing the first character(s) of the
item wanted for rapidly accessing it. A subset of the full alpha keyboard shown on p. 62 is
sufficient for catalog browsing as pictured here. But there are three differences:

1/x (= ) is for
reverting conver-
sions easily (see p.
139).
A B C D E F just calls the
g character ‘ ’, and
f h calls the
print character in
G H I
catalog browsing
(since case switch-
ENTER ing is not needed
⇆ here).
J ( K L
and will
XEQ browse the open
catalog.
M N O P
or
▲ × select the item
displayed, recall or
Q R S T execute it, and exit
the catalog.
▼ 1 2 – or leave
the catalog without
U V W executing anything,
i.e. they cancel the
EXIT 0 ● + catalog call.
See p. 133 for some
⎙ X Y Z
examples.

You may switch catalogs easily by just calling a new one accessible in current mode
directly from the catalog you are browsing – no need to first.
When the last catalog called is reopened, the last item viewed therein is immediately
displayed for easy repetitive use. A single function may be contained in more than one
catalog.

WP 34S Owner’s Manual Edition 3.1 Page 125 of 211


There are also three browsers featured for checking memory, flags, program labels and
registers (i.e. , , and ). They may be called in all modes except alpha.
Therein, and work as in catalogs. and operate in mode,
however. And some special keys and special rules may apply in browsers as explained in
the following table:

Keys to press Contents and special remarks

Defined alphanumeric (i.e. global) labels. The first item displayed is the top global
label of the current program63 – if there is no such label, the end of this program is
shown.
and browse global labels, while the location of the label displayed is
indicated in the lower line (rAMm for RAM, Lib for the library in FM, or bup for
the backup region therein). Duplicate labels also show the primary address,
e.g. CALLS 012 when found a second time in RAM, or e.g. CALLS Lib
when found a second time elsewhere.
and browse programs, i.e. show the first label in previous or next
program (with programs separated by END statements).
, , and allow quick jumps to the top of rAMm, Lib, or bup, respectively.
goes to the alpha label displayed, while executes it. Both key-
strokes will perform a label search as described on p. 67. In programming
mode, will insert a GTO and an XEQ with said label trailing.
starts the current program without performing a label search first.
and execute PRCL or PSTO for the current program, respectively.
The latter makes sense if called in rAMm or bup only.
enters mode STO, allowing for browsing or editing the current program. The
latter will work in rAMm only. Note that will delete a step, while or
another will leave CAT immediately.
deletes the current program, be it in rAMm or Lib.

Browses all allocated stack and general purpose registers showing their contents,
starting with X.
goes up the stack, continuing with the other lettered registers, then with R00,
R01, etc.
browses the registers going down from the highest allocated numbered register
(R99 in startup default) to R00 if applicable, then continuing with K, J, etc.
turns to local registers if applicable, starting with R.00. Then, and browse
local registers up and down until another returns to X. Local register
addresses may exceed .15 here!
Input of any legal letter or two-digit number jumps to the corresponding register
(see p. 26).
or recall the register displayed. In programming mode, they enter a
corresponding step RCL …
(In your WP 34S, the keys and do what SHOW did in vintage calculators –
see p. 40.)

63
The current program is the one the program pointer is in, as mentioned above.
WP 34S Owner’s Manual Edition 3.1 Page 126 of 211
Keys to press Contents and special remarks

Displays the memory status and browses all user flags, similar to STATUS on the
HP-16C. It shows the amount of free memory words in RAM and FM first, e.g.:

.
Press and read if there are summation registers used, plus the number of
global numbered registers and local registers allocated:

or .

Another presents the status of the first 30 user flags in one very concise
display as explained below.

Example:
If flags 2, 3, 5, 7, 11, 13, 14, 17, 19, 20, 26, and X are set, and labels B and D
are defined in program memory, will display this:

Where the mantissa is usually displayed, there are now three rows of horizontal
bars. Each row shows the status of 10 flags. If a particular flag is set, the
corresponding bar is lit. So here the top row of bars indicates flags 0 and 1 are
clear, 2 and 3 set, and 4 clear. Then a II separates the first five flags from the
next. The following top-row bars indicate flag 5 set, 6 clear, 7 set, 8 and 9 clear.
The next two rows show the status up to flag 29.

Pressing once again will increment the start address by ten; so the display
will look like:

Another will show flags 20 - 49 etc. until 70 - 99, 80 - 99, 90 - 99. A final
will display the last 12 global flags in rows of four – note flag X is shown being
set as we expect:

The exponent section indicates the status of the four hotkeys – if all four labels
are defined in programs then ALL will be shown there.

browses backwards.
Alternatively, pressing a digit (e.g. ) will display up to 30 flags starting with 10
times this digit (flags 50 – 79 here). Pressing a legal letter such as will display
the top 12 flags.

WP 34S Owner’s Manual Edition 3.1 Page 127 of 211


Finally, see the following list of catalogs featured:

Keys to in modes Contents and special remarks


press

Constants like in the HP-35s, but more. See them listed on p. 134ff.
DECM &
While browsing this catalog, the values of the constants are displayed.
¬STO
Picking a constant will recall it.

Integer Calls the command # (see the IOP).

Picking a constant will insert a program step containing the name of the
DECM &
STO
constant selected, preceded by ¦#¦. This step will then recall the value
of said constant in program execution.

Opens the same catalog of constants as , but picking a


DECM &
constant will execute a complex recall. So, a stack looking like [x, y, …]
¬STO
before will contain [constant, 0, x, y, …] after picking.

Picking a constant will insert a program step containing the name of the
DECM &
STO
constant selected, preceded by ¦©#¦. This step will then perform a
complex recall of the value of said constant in program execution.

Conversions as listed in a table below. While browsing this catalog, the


DECM converted content of X is displayed. Picking a conversion will return
this value.

INPUT
‘Complex’ letters mandatory for many languages (see p. 132). Case
may be toggled here (see above).

DECM Matrix operations library.

¬INPUT Mode setting functions.

DECM Probability distributions beyond the standard normal and its inverse.

¬INPUT Extra programming and I/O functions.

INPUT Superscripts and subscripts (see p. 132).

DECM Extra statistical functions.

All summation registers. While browsing this catalog, register contents


DECM
are displayed. Picking a register will recall its contents.

¬INPUT All tests except the two on the keyboard (see p. 130).

INPUT Comparison symbols and brackets, except , , and (see p. 132).

WP 34S Owner’s Manual Edition 3.1 Page 128 of 211


Keys to in modes Contents and special remarks
press

DECM Extra real functions.


These three catalogs are
Integer Extra integer functions. merged in mode STO to
ease programming.
INPUT Extra alpha functions.

DECM Extra complex functions.

INPUT Punctuation marks and text symbols (see p. 132).

INPUT Arrows and mathematical symbols (see p. 132).

See the next pages for detailed contents of the various catalogs. Items are sorted
alphabetically within the catalogs (see p. 74 for the sorting order). You may access a
particular item quickly by typing the first characters of its name – see p. 133 for examples
and constraints.

Note neither catalog nor browser calls can be programmed.

WP 34S Owner’s Manual Edition 3.1 Page 129 of 211


Catalog Contents in Detail
BACK COV nΣ BC?
2
BASE? L.R. Σln x BS?
DET 12h Binom CASE SEED Σln2y CNVG?
LINEQS 1COMPL BinomP CFALL SERR Σlnx DBL?
MROW+× 24h Binomu CLALL SERRW Σlnxy ENTRY?
MROW× 2COMPL Binom –1 CLPALL SUM Σlny EVEN?
MROW BASE Cauch CLREGS sW Σx FC?
M+× BestF … CLSTK sXY Σx2 FC?C
M-1 DBLOFF Expon CL xg Σx2y FC?F
M-ALL DBLON … CNST xW Σxlny FC?S
M-COL DENANY FP(x) DEC x Σxy FP?
M-DIAG DENFAC Fu(x) DROP Σy FS?
M-ROW DENFIX F(x) DSL m Σy2 FS?C
M× DENMAX F –1(p) DSZ P Σylnx FS?F
M.COPY DISP Geom END FS?S
M.IJ D.MY … ERR W gPIX?
M.LU E3OFF Lgnrm FF %Σ INTM?
M.REG E3ON … FLASH? INT?
nCOL ExpF Logis f ’(x) KEY?
nROW FAST … f ”(x) PUTK WSIZE? LBL?
TRANSP FRACT Norml gCLR RCLS XEQ LEAP?
JG1582 … gDIM RECV y M.SQR?
JG1752 Poiss gDIM? REGS? z NaN?
LinF … gFLP RESET GTO ODD?
LogF Poisλ gPLOT RM? OFF PRIME?
LZOFF … gSET RTN+1 ON REALM?
LZON tP(x) GTO R-CLR XEQ SPEC?
M.DY tu(x) INC R-COPY TOP?
PowerF t(x) ISE R-SORT ⎙ADV XTAL?
RCLM t –1(p) ISZ R-SWAP ⎙CHR x< ?
RDX, Weibl LOADP SAVE ⎙CrXY x≤ ?
SETUK RDX. … LOADR SENDA ⎙PLOT x = +0?
SETUSA REGS Φu(x) LOADSS SENDP ⎙PROG x = -0?
SIGNMT RMF φ(x) LOADΣ SENDR ⎙r x≈ ?
2
SLOW SEPOFF LocR SENDΣ ⎙REGS x≥ ?
2
SSIZE4 SEPON INV LocR? SKIP ⎙STK x> ?
2
SSIZE8 SETCHN p MEM? SMODE? ⎙TAB ∞?
2
STOM SETDAT u MSG SSIZE? ⎙WIDTH ⎙?
UNSIGN SETEUR NOP STOS ⎙α
WSIZE SETIND PopLR TICKS ⎙α+
Y.MD SETJPN PRCL t ⎙Σ
⎙DLAY SETTIM PROMPT VIEW ⎙+α
⎙MODE PSTO VW + ⎙#

WP 34S Owner’s Manual Edition 3.1 Page 130 of 211


varies with the mode set, except in programming.64 It contains in …
… alpha … decimal mode: … integer modes:
mode:
3 3 C3
√x LCM Wm √x RLC √x
C
VERS AGM Ln Wp ASR ROUNDI AGM
–1 C
x ANGLE LN1+x W BATT RR CNST
C
DATE BATT Ln XNOR CB RRC CONJ
3 C
DAY Bn LN x CEIL SB CROSS
C
IP Bn* LNΓ x DBLR SEED DOT
x C
LENG CEIL MANT √y DBL× SIGN DROP
C x
MONTH DATE MAX YEAR DBL / SL e -1
C
RC# DATE MIN DATE dRCL SR FIB
C
RL DAY MOD DAY DROP sRCL gd
C
RR DAYS+ MONTH IP FB ULP gd –1
C
SL DECOMP NAND LENG FIB VERS IDIV
C
SR DEG NEIGHB MONTH FLOOR WHO LN1+x
3 C
TIME dRCL NEXTP RCL IDIV x LN
C
x DROP NOR RC# GCD XNOR LNΓ
C
D J Pn RL LCM x SIGN
x C
erf RAD RR LJ √y SINC
C
erfc RDP SL MASKL IP Wp
C
EXPT RESET SR MASKR LENG W –1
x C 3
e -1 ROUNDI STO MAX RCL x
Cx
FIB RSD TIME MIN RC# √y
C
FLOOR SDL x MIRROR RL
C
GCD SDR MOD RR Γ
gd SIGN Γ NAND SL C
(-1)x
gd –1 SINC γXY nBITS SR
GRAD SLVQ ΓXY NEIGHB STO
Hn sRCL DAYS NEXTP x
X
Hnp STOPW ζ NOR (-1)
X
H.MS+ TIME (-1) RESET ×MOD
H.MS– Tn DATE RJ ^MOD
IDIV ULP %MG RL
iRCL Un %MRR
Iβ VERS %T
IΓp WDAY %Σ
IΓq WHO %+MG
J D

64
In programming mode, these three contents are merged.
WP 34S Owner’s Manual Edition 3.1 Page 131 of 211
À À À à à à Here are the contents of the alpha catalogs. Small
Á Á Á á á á font is printed on light grey background on this page.
The catalog is listed left. Use to toggle
ÂÃĀĂ Ã Ã âãāă ã ã
between cases. Accented letters are as wide as plain
Ä Ä Ä ä (ă) ä ä ones wherever possible.
Æ Æ Æ æ æ æ
Å Å Å å å å
Ć Ć Ć ć ć ć , : ; ‘ “ * @ _ ~ % $ € £ ¥ ⊙ ⊕ ⎙ ` #
Č Č Č č č č , : ; ' “ * @ _ ~ % $ € £ ¥ ☉ ♁ ⎙ ` #
Ç Ç Ç ç ç ç , : ; ‘ “ * @ _ ~ % $ € £ ¥ ☉ ♁ ⎙ ` #
Ð Ð Ð đ ð ð
È È È è è è
É É É é é é < ≤ = ≈ ≥ > [ ] { } √ ∫ ∞ ^
ÊĒĔĚ Ē Ē êēĕě ē ē < ≤ = ≈ ≥ > [ ] { } →←↑↓√∫∞^↕
Ë Ë Ë ë (ĕ) ë ë < ≤ = ≈ ≥ > [ ] { } → ← ↑ ↓ √ ∫ ∞^↕
ħ ħ ℏ ħ ħ
Ì Ì Ì ì ì ì
Í Í Í í í í 0° 1 2 2 3 A B c c d e k m n p q u w x x Y µ -1 * ∞
ÎĨĪĬ Ĩ Ĩ î ĩ ī ĭ ĩ ĩ ₀°₁₂²³ₐ⒝⒞©⒟ₑₖₘₙₚₛᵤ⒲ₓᵡᵧ¶⁸⁺₈
Ï Ï Ï ï(ĭ) ï ï ₀°₁₂ ² ³ₐ⒝⒞ © ⒟ₑₖₘₙₚₛᵤ⒲ₓᵡᵧ¶⁸ ⁺ ₈
ÑŇ Ñ Ñ ñň ñ ñ
The letters provided in your WP 34S allow for correct
Ò Ò Ò ò Ò Ò 9
writing the languages of more than 3·10 people using
Ó Ó Ó ó ó ó Greek or simple variants of Latin alphabets, i.e. the
ÔÕŌŎ Õ Õ ôõōŏ õ õ following languages:
Ö Ö Ö ö (ŏ) ö ö Afrikaans, Català, Cebuano, Česky, Cymraeg, Dansk,
Ø Ø Ø ø ø ø Deutsch, Eesti, English, Español, Euskara, Français,
Ř Ř Ř ř ř ř Gaeilge, Galego, Ελληνικά, Hrvatski, Bahasa
Š Š Š š Š Š Indonesia, Italiano, Basa Jawa, Kiswahili, Kreyòl
ß ß ß ß ß ayisyen, Magyar, Bahasa Melayu, Nederlands, Norsk,
Ù ù Português, Quechua, Shqip, Slovensky, Slovenščina,
Ù Ù ù ù
Srpski, Basa Sunda, Suomi, Svenska, Tagalog,
Ú Ú Ú ú ú ú Winaray, and Zhōngwén (with a little trick explained
ÛŨŪŬ Ũ Ũ ûũūŭ ũ ũ below). This makes the WP 34S the most versatile
Ü Ü Ü ü (ŭ) ü ü global calculator known. If you know further living
Ů Ů Ů ů ů ů languages covered, please tell us. Find the full
ẋ ẋ / Ẁ Ẁ character set provided in Appendix E.
Ý Ý Ý ý ý ý Mandarin Chinese (Zhōngwén) features four tones, usually
transcribed like e.g. mā, má, mă, and mà. So we need different
ŷ ŷ / ȳ ȳ letters for ā and ă here, and for e, i, o, and u as well. With six
Ÿ Ÿ Ÿ ÿ ÿ ÿ pixels total character height, we found no way to display these in
both fonts nicely, keeping letters and accents separated for easy
Ž Ž Ž ž ž Ž reading. For an unambiguous solution, we suggest using a
dieresis (else not employed in Hànyŭ pīnyīn) representing the
third tone here. Pinyin writers, we ask for your understanding.

WP 34S Owner’s Manual Edition 3.1 Page 132 of 211


Accessing Catalog Items the Fast Way

You can browse each and every catalog just using the cursors and as explained
on p. 125. You may reach your target significantly faster, however, taking advantage of the
alphabetical method demonstrated in the left columns of the table below:

1 User , , , , , or , , or
input , , , , or in alpha mode in alpha mode

Dot Shows the first item in this catalog


matrix
display (e.g. BC? in ) (e.g. à in ) (e.g. , in )

2 User 1st character of command desired Desired basic letter


input (e.g. ) (e.g. )
Dot Shows the first item starting with this character*
matrix
display (e.g. FC? ) (e.g. Ù )

3 User
2nd character (e.g. )
input
Dot Shows the first item starting
matrix with this sequence*
display (e.g. FS? )

… Continue browsing with until you reach the item desired


(e.g. FS?C ). (e.g. Ü ). (e.g. € ).

n User or
input
Your WP 34S leaves the catalog returning to the mode set before …
… and executes or inserts the command … and appends the selected character
chosen, or recalls the constant selected. to alpha.

Dot Result Contents of alpha register


matrix
(e.g. true ) (e.g. 3 Rüben à 0,25€ )
display

* Like in browsing a dictionary, it may be faster to search the letter following alphabetically instead
and then browse backwards. In our example, finds FS?C but
would do as well.
Remember you can enter Greek letters in a search using prefix , e.g. + for (compare
p. 62).
If a character or sequence specified is not found in the catalog chosen then the first item following
alphabetically will be shown – see the sorting order on p. 74. If there is no such item, then the last
item in this catalog is displayed.
You may key in even more than two characters – after 3 seconds, however, or after or ,
the search string will be reset and you may start with a first character again.

WP 34S Owner’s Manual Edition 3.1 Page 133 of 211


Constants (CONST)

Your WP 34S contains a rich catalog of constants. Navigation therein works as explained
above. Names of astronomical and mathematical constants are printed on colored
background below. Values of physical constants (including their relative standard
deviations given in parentheses below) are from CODATA 2010, copied in July 2011,
unless stated otherwise. Green background denotes exact or almost exact numeric values.
The redder the background, the less precisely the particular constant is known, even by
the national standards institutes and the international scientific community. 65
2
Wb V s kg m
For the units, remember tesla with 1T 1 1 2 , joule with 1J 1N m 1
m 2
m s2
J
and on the other hand 1J 1W s 1V A s . Thus 1 1A m 2 .
T

Employ the constants stored here for further useful equivalences, like expressing joules in
1
electron-volts ( 1A s V eV 6.24 1018 eV ), or calculating the wavelength from the
e
frequency of electromagnetic radiation via c , or whatever else crosses your mind.
f

Now here is the complete content of CONST:

Numeric value Remarks

1/2 0.5 Trivial but helpful constant in some iterations.

a 365.242 5 d (per definition) Gregorian year

a0 5.291 772 109 2E-11 m (3.2E-10) Bohr radius a0


4 R

am 384.4E6 m (1E-3) Semi-major axis of the Moon’s orbit

Semi-major axis of the Earth’s orbit. Within the


a⊕ 1.495 979E11 m (1E-6)
uncertainty stated here, it equals 1 AU.

c 2.997 924 58E8 (per definition) Speed of light in vacuum ≈ 300 000 km/s

c1 3.741 771 53E-16 m 2 W (4.4E-8) First radiation constant c1 2 h c2

c2 0.014 387 770 m K (9.1E-7) Second radiation constant c2 hc


k

65
The numbers printed in parentheses allow for determining the precision of results you may obtain using
these constants, through the process of ‘error propagation’ going back to C. F. Gauß (1777 – 1855). This
procedure is essential if your results are to be trustworthy – not only in science. Consult a suitable
reference. Yardstick measurements cannot yield results precise to four decimals.
WP 34S Owner’s Manual Edition 3.1 Page 134 of 211
Numeric value Remarks

2
e 1.602 176 565E-19 C (2.2E-8) Electron charge e 0 G0
K J RK

Euler’s e. Note the letter e represents the electron


eE 2.718 281 828 459 045…
charge elsewhere in this table.

C
F 96 485.336 5 (2.2E-8) Faraday’s constant F e NA
mol

Fα 2.502 907 875 095 892 8...


Feigenbaum’s and
Fδ 4.669 201 609 102 990 6...

g 9.806 65 (per definition) Standard earth acceleration

m3 Newtonian constant of gravitation. See GM below


G 6.673 84E-11 (1.2E-4)
kg s 2 for a more precise value.

G0 7.748 091 734 6E-5 / (3.2E-10) Conductance quantum G0 2e 2 2


h RK

GC 0.915 965 594 177… Catalan’s constant

ge -2.002 319 304 361 53 (2.6E-13) Landé’s electron g-factor

3 Newtonian constant of gravitation times the Earth’s


GM 3.986 004 418E14 m (2.0E-9) mass with its atmosphere included (according to
s2 WGS84)

h 6.626 069 57E-34 J s (4.4E-8) Planck constant

ℏ 1.054 571 726E-34 J s (4.4E-8) h


2

k 1.380 648 8E-23 J (9.1E-7) Boltzmann constant k R


K NA

KJ 4.835 978 70E14 Hz (2.2E-8) Josephson constant K j 2e


V h

lP 1.616 199E-35 m (6.0E-5) Planck length l P G t pc


c3

me 9.109 382 91E-31 kg (4.4E-8) Electron mass

Mm 7.349E22 kg (5E-4) Mass of the Moon

mn 1.674 927 351E-27 kg (4.4E-8) Neutron mass

mp 1.672 621 777E-27 kg (4.4E-8) Proton mass

WP 34S Owner’s Manual Edition 3.1 Page 135 of 211


Numeric value Remarks

MP 2.176 51E-8 kg (6.0E-5) Planck mass M P c 22 g


G
-3
mu 1.660 538 921E-27 kg (4.4E-8) Atomic mass unit = 10 kg / NA

muc2 1.492 417 954E-10 J (4.4E-8) Energy equivalent of atomic mass unit

mµ 1.883 531 475E-28 kg (5.1E-8) Muon mass

M☉ 1.989 1E30 kg (5E-5) Mass of the Sun

M⊕ 5.973 6E24 kg (5E-5) Mass of the Earth

NA 6.022 141 29E23 / mol (4.4E-8) Avogadro’s number

‘Not a Number’, e.g. ln(x) for x ≤ 0 or tan(90°)


unless in complex domain. So NaN covers poles
as well as regions where a function result is not
NaN not nuMmeric defined at all. Note that infinities, however, are
considered numeric (see the end of this table).
Non-numeric results will lead to an error message
thrown – unless flag D is set.

p0 101 325 Pa (per definition) Standard atmospheric pressure

qP Planck charge q p 4 c 11.7e . This was


1,875 545 9E-18 A s (6.0E-5) 0
in CODATA 2006, but in 2010 no more.

J
R 8.314 462 1 (9.1E-7) Molar gas constant
mol K
2
re 2.817 940 326 7E-15 m (9.7E-10) Classical electron radius re a0

RK 25 812.807 443 4 Ω (3.2E-10) von Klitzing constant RK h


e2

Rm 1.737 530E6 m (5E-7) Mean radius of the Moon

2
R∞ 1.097 373 156 853 9E7 / m (5.0E-12) Rydberg constant R me c
2h

R☉ 6.96E8 m (5E-3) Mean radius of the sun

R⊕ 6.371 010E6 m (5E-7) Mean radius of the Earth

Semi-major axis of the model WGS84 used to


Sa 6.378 137 0E6 m (per definition) define the Earth’s surface for GPS and other
surveying purposes

Sb 6.356 752 314 2E6 m (1.6E-11) Semi-minor axis of WGS84

WP 34S Owner’s Manual Edition 3.1 Page 136 of 211


Numeric value Remarks

Se2 6.694 379 990 14E-3 (1.5E-12) First eccentricity squared of WGS84

Second eccentricity squared of WGS84 (it is really


Se’2 6.739 496 742 28E-3 (1.5E-12)
called e’2 in that article, sorry)

Sf –1 298.257 223 563 (per definition) Flattening parameter of WGS84

T0 273.15 K (per definition) = 0°C, standard temperature

G lp
tP 5.391 06E-44 s (6.0E-5) Planck time t P
c5 c

c2 c M pc2 Ep
TP 1.416 833E32 K (6.0E-5) Planck temperature TP
k G k k

Molar volume of an ideal gas at standard


3
Vm 0.022 413 968 m (9.1E-7) RT0 l
mol conditions Vm 22.4
p0 mol

Z0 376.730 313 461… Characteristic impedance of vacuum Z 0 0 c

e2 1
α 7.297 352 569 8E-3 (3.2E-10) Fine-structure constant
4 0 c 137

γEM 0.577 215 664 901 532 86… Euler-Mascheroni constant EM

1 2
γp 2.675 222 005E8 (2.4E-8) Proton gyromagnetic ratio p
P
s T

1
ε0 8.854 187 817…E-12 Electric constant or vacuum permittivity 0 2
0c

λC 2.426 310 238 9E-12 m (6.5E-10)


h
Compton wavelengths of the electron ,
λCn 1.319 590 906 8E-15 m (8.2E-10) C
me c
neutron, and proton, respectively
λCp 1.321 409 856 23E-15 m (7.1E-10)

Magnetic constant or vacuum permeability


µ0 1.256 637 061 4…E-6 : 4 10 7 V s
0
A m

µB 9.274 009 68E-24 J (2.2E-8) Bohr’s magneton e


T B 2me

µe -9.284 764 30E-24 J (2.2E-8) Electron magnetic moment


T

WP 34S Owner’s Manual Edition 3.1 Page 137 of 211


Numeric value Remarks

µn -9.662 364 7E-27 J (2.4E-7) Neutron magnetic moment


T

µp 1.410 606 743E-26 J (2.4E-8) Proton magnetic moment


T

5.050 783 53E-27 J Nuclear magneton e


µu T
(2.2E-8) u 2m p

µµ -4.490 448 07E-26 J (3.4E-8) Muon magnetic moment


T

W 2 5k 4
σB 5.670 373E-8 (3.6E-6) Stefan-Boltzmann constant
m2 K 4 B
15h 3 c 2

1 5
Φ 1.618 033 988 749 894… Golden ratio
2

Φ0 2.067 833 758E-15 V s (2.2E-8) Magnetic flux quantum h 1


0 2e KJ

ω 7.292 115E-5 rad (2E-8) Angular velocity of the Earth according to WGS84
s

–∞ -Infinity May the Lord of Mathematics forgive us calling


these ‘constants’. Note both are counted as
∞ Infinity numeric values in your WP 34S.

# See the very last command in the IOP.

WP 34S Owner’s Manual Edition 3.1 Page 138 of 211


Unit Conversions (CONV)

CONV mainly provides the means to convert local to common units.66 Navigation works as
in the other catalogs. There is one specialty, however: (i.e. ) will execute the
inverse of the conversion displayed and leave CONV.

Example: Assume the display set to FIX 3. Then keying in

will display

telling you that


4 acres equal 1.619 hectares.

Now press and you will get

being the
amount of acres equaling 4 hectares.

Press again and you see

confirming what was just said.

Leave CONV via and the display will return to 9.884.

The conversions provided are listed in the table below. The calculations are for your
orientation only; your WP 34S uses more precise numbers where applicable.

Conversion Calculation Remarks Class

°C→°F * 1.8 + 32 Temperature

°F→°C - 32 ) / 1.8 Temperature

Converts to gon , also known as grads


°→G / 0.9 Angle
or gradians

°→rad * π / 180 Equals D R, converting to radians Angle

acres→ha * 0.404 686 1 ha = 104 m2 Area

acreUS→ha * 0.404 687 Based on the U.S. Survey foot Area

ar.→dB Amplitude ratio Ratio

atm→Pa * 1.013 25E5 Pressure

66
The SI system of units is agreed on internationally. Meanwhile, it is adopted by all countries on this planet
except two. Thus for most readers, most of the units appearing in CONV may look obsolete at least. They
die hard, however, in some corners of this world (English is spoken in all of them). For symmetry reasons,
we think about adding some traditional Indian and Chinese units to CONV.
This table may also give you an idea of the mess we had in the world of measures before going metric.
WP 34S Owner’s Manual Edition 3.1 Page 139 of 211
Conversion Calculation Remarks Class

AU→km * 1.495 98E8 Astronomic units Length

bar→Pa * 1E5 Pressure

Btu→J * 1 055.06 British thermal units Energy

cal→J * 4.186 8 Energy

cft→l * 28.316 9 Cubic feet Volume

cm→inches / 2.54 Length

cwt→kg * 50.802 4 (Long) hundredweight = 112 lbs Mass

dB→ar. Amplitude ratio Ratio

dB→pr. Power ratio Ratio

fathom→m * 1.828 8 Length

feetUS→m * 0.304 801 1 U.S. Survey foot = 1200/3937 m exactly Length

feet→m * 0.304 8 The so-called ‘international feet’ Length

flozUK→ml * 28.413 1

flozUS→ml * 29.573 5
1 l = 1/1000 m3 Volume
galUK→l * 4.546 09

galUS→l * 3.785 42

G→° * 0.9 Angle

g→oz / 28.349 5 Mass

G→rad * π / 200 Angle

g→tr.oz / 31.103 5 Mass

ha→acres / 0.404 686 1 ha = 10000 m2 Area

ha→acreUS / 0.404 687 Based on the U.S. Survey foot Area

HPe→W * 746 Electric horsepower Power

hpUK→W * 745.700 British horsepower Power

inches→cm * 2.54 Length

inHg→Pa * 3 386.39 Pressure

WP 34S Owner’s Manual Edition 3.1 Page 140 of 211


Conversion Calculation Remarks Class

J→Btu / 1 055.06 Energy

J→cal / 4.186 8 Energy

J→kWh / 3.6E6 Energy

kg→cwt / 50.802 4 (Long) hundredweight = 112 lbs Mass

kg→lb / 0.453 592 Mass

kg→stones / 6.350 29 Mass

kg→s.cwt / 45.359 2 Short hundredweight = 100 lbs Mass

km→AU / 1.495 98E8 Astronomical units Length

km→l.y. / 9.460 73E12 Light years Length

km→miles / 1.609 344 Length

km→nmi / 1.852 Nautical miles Length

km→pc / 3.085 68E16 Parsec Length

kWh→J * 3.6E6 Energy

lbf→N * 4.448 22 Force

lb→kg * 0.453 592 Mass

l.y.→km * 9.460 73E12 Light years Length

l→cft / 28.316 9

l→galUK / 4.546 09 1 l = 1/1000 m3 Volume

l→galUS / 3.785 42

miles→km * 1.609 344 Length

ml→flozUK / 28.413 1
1 ml = 1 cm3 Volume
ml→flozUS / 29.573 5

mmHg→Pa * 133.322 Pressure

m→fathom / 1.828 8 Length

m→feet / 0.304 8 Length

m→feetUS / 0.304 801 Length

m→yards / 0.914 4 Length

WP 34S Owner’s Manual Edition 3.1 Page 141 of 211


Conversion Calculation Remarks Class

nmi→km * 1.852 Nautical miles Length

N→lbf / 4.448 22 Force

oz→g * 28.349 5 Ounces Mass

Pa→atm / 1.013 25E5 1 Pa = 1 N/m2 Pressure

Pa→bar / 1E5 Pressure

Pa→inHg / 3 386.39 Pressure

Pa→mmHg / 133.322 Pressure

Pa psi / 6 894.76 Pressure

Pa→torr / 133.322 Pressure

pc→km * 3.085 68E16 Parsec Length

pr.→dB 10 lg P1 Power ratio Ratio


P2

psi→Pa * 6 894.76 Pounds per square inch Pressure

PS(hp)→W * 735.499 Horsepower Power

rad→° * 180 / π Equals R D Angle

rad→G * 200 / π Angle

stones→kg * 6.350 29 Mass

s.cwt→kg * 45.359 2 Short hundredweight = 100 lbs Mass

s.tons→t * 0.907 185 Short tons Mass

tons→t * 1.016 05 Imperial tons Mass

torr→Pa * 133.322 1 torr = 1 mm Hg Pressure

tr.oz→g * 31.103 5 Troy ounces Mass

t→s.tons / 0.907 185


1 t = 1000 kg Mass
t→tons / 1.016 05

WP 34S Owner’s Manual Edition 3.1 Page 142 of 211


Conversion Calculation Remarks Class

W→HPe / 746 Power

W→hpUK / 745.700 Power

W→PS(hp) * 735.499 Power

yards→m * 0.914 4 Length

The constant To may be useful for conversions of temperatures, too; it is found in CONST
and is not repeated here because it is only added or subtracted.

You may, of course, combine conversions as you like. For example, filling your tires with a
maximum pressure of 30psi the following will help you at gas stations in Europe and
beyond:

resulting in 2,1bar.

Now you can set the filler and will not blow your tires.

In cases of emergency of a particular kind, remember becquerel equals hertz, gray is the
unit for deposited or absorbed energy ( 1Gy 1 J kg ), and sievert (Sv) is gray times a
radiation dependant dose conversion factor for the damage caused in human bodies.

In this area also some outdated units may be found in older literature: Pour les amis de
10 10
Mme. Curie, 1Ci 3.7 10 Bq 3.7 10 decays s . And for those admiring the very first
Nobel laureate in physics, Mr. Röntgen, for finding the x-rays (ruining his hands in these
experiments), the charge generated by radiation in matter was measured by the unit
1R 2.58 10 4 As kg . A few decades ago, the rem (i.e. roentgen equivalent in man)
measured what the sievert does today.

Predefined Global Alpha Labels (CAT)

In addition to the label ‘δx’ reserved for step size in calculation of derivatives (see f’(x) in
the IOP), additional labels may already be provided for particular tasks. You will find them
listed in CAT when the respective library routines are loaded in FM (see below). Thus they
will not take any steps from user program memory in RAM.

All library routines presently available are found on the WP 34S website in the directory
http://wp34s.svn.sourceforge.net/viewvc/wp34s/library/ . They are text files with extension
.wp34s by convention. These routines include a suite of basic 3D vector operations, a
TVM (time value of money) application, some matrix routines including an editor, and
more. You may open these text files using e.g. Notepad, and you should find the
necessary user information at the beginning of each file.

WP 34S Owner’s Manual Edition 3.1 Page 143 of 211


The library files are also included in the distribution ZIP file in source form (*.wp34s) as
well as in a precompiled library (wp34s-lib.dat – some 3kB) which is part of the
firmware files calc_full.bin, calc_xtal_full.bin, and calc_ir_full.bin – so
you get the full library when you load one of these firmware files into your WP 34S (see
Appendix A about how to do this).

When you copy wp34s-lib.dat into the directory your WP 34S emulator runs in, you
can access all those routines via CAT from your emulator as well. See Appendix D.

See Appendix H if you want to write such routines yourself.

WP 34S Owner’s Manual Edition 3.1 Page 144 of 211


THE STOPWATCH APPLICATION

As mentioned in the IOP, a stopwatch is


provided. It works on your WP 34S with
a quartz crystal and associated firmware
installed (see App. A and H). It will work
on the emulator as well.

Name Keys to press in modes Remarks

Stopwatch following the HP-55 timer.67


STOPW When STOPW is started, the display will look like this:
DECM,
STOPW
¬STO

unless the timer was already running.


Within STOPW, the following keys will work:

starts or stops the timer without changing its value.

resets the timer to zero without changing its status (running or stopped).
or

hides or displays tenths of seconds. Startup default is ‘display’.

sets the current register address (CRA, startup default is 00). Your numeric
input will already be displayed in the exponent section as shown here68:

67
There are two deviations: Your WP 34S will not take the content of X as start time. Start times are
supported by RCL here. And your WP 34S will display tenths instead of hundredth of seconds.
68
Attempts to specify a CRA beyond the allocated address range will be blocked and may cause ‘__’ or the
like being displayed in the exponent section.
On the HP-55, input of a single digit was sufficient for storing, since only 10 registers were featured for this
purpose there. Furthermore, there was no automatic address increment.
WP 34S Owner’s Manual Edition 3.1 Page 145 of 211
stores the present timer value in the current register at execution time in
format hhh.mmssd without changing the timer status or value. It then
increments the CRA and displays it as shown above.

hides or displays the CRA. Startup default is ‘hide’.

or increments or decrements the CRA, respectively.

combines and in one keystroke, but the total time since the
last explicit press of or is shown and updated in the top row like:

or .
Note this total time is volatile, however – it will disappear without a trace
when or is pressed.

adds the present timer value to the statistics registers like would do.
This allows for computing its arithmetic mean and standard deviation after
leaving STOPW.

combines and in one keystroke.

nn recalls rnn without changing the status of the timer. The value recalled may
be used e.g. as start time for further incrementing.

leaves the application. Unless already stopped, however, the timer


continues incrementing in the background (indicated by the small ‘=’
annunciator flashing) until
a) stopped explicitly by within STOPW or
b) your WP 34S is turned off.
ATTENTION: Since your WP 34S shuts down automatically after
some five minutes of non-usage, press some prefix keys twice in
suitable intervals if you need to measure longer times.
While the stopwatch display is limited to 99h59'59" 9, internal counting
will continue with the display showing the time modulo 100.

Note no other keys will work in STOPW – so e.g. for adding or subtracting split times
you have to leave this application.

Note STOPW is not programmable.

WP 34S Owner’s Manual Edition 3.1 Page 146 of 211


APPENDIX A: SETUP AND COMMUNICATION

How to Flash Your HP-20b or HP-30b

Unless you buy a WP 34S pre-flashed as explained on p. 5, you must do the flashing
yourself. Then you need an unmodified HP-20b or HP-30b calculator, a special cable, a
binary file to load on your computer, and software for the transmission to your calculator.

The transmission software is called MySamBa. Download MySamBa.zip from


http://sourceforge.net/projects/wp34s/files/FlashTool/ and unpack it.

The specific binary file you need to transmit to your calculator to make it your WP 34S
is called calc.bin and is included in the zipped release package you can download
from http://sourceforge.net/projects/wp34s/files/ . Alternatively, you may download
from http://wp34s.svn.sourceforge.net/viewvc/wp34s/trunk/realbuild/ one of the
following files alone:

calc.bin features maximum FM, but supports neither the real


time clock commands (DATE, TIME, SETDAT, and
SETTIM) nor STOPW nor the print commands,

calc_full.bin equals calc.bin but already includes the FM library;

calc_xtal.bin assumes a quartz crystal is installed and includes the


real time clock commands and STOPW but no print
commands,

calc_xtal_full.bin equals calc_xtal.bin but includes the FM library;

calc_ir.bin contains everything featured in calc_xtal.bin plus


the print commands requiring both the quartz crystal
and an IR diode built in,

calc_ir_full.bin equals calc_ir.bin but includes the FM library.

The amount of FM you have to pay for the STOPW application is 1.5kB, for printer
support some 3.5kB more. Make your choice! But check the hardware pre-
requisites (see Appendix H) or your calculator may hang.

For the cable, there are two alternatives, presented in order of appearance:

A. There was a limited production run of special programming cables supplied by HP.
This item – pictured overleaf – is almost out of stock now. Thus if you have got
one, keep it. Then you will need a computer with a traditional 9-pin serial port.

ATTENTION: If your computer does not feature such an hardware serial


interface (many modern computers do not), you will need a USB-to-serial
converter to connect this programming cable to your computer. Following our

WP 34S Owner’s Manual Edition 3.1 Page 147 of 211


experience, converters containing FTDI
chips work – others may not. Such a
converter is offered in the web e.g. at
http://commerce.hpcalc.org/usbserial.php .

WARNING: As long as this cable is


connected to your calculator, it may
draw a considerable current from its
batteries. If your calculator happens to
hang during the flashing procedure
described overleaf, the calculator
processor may be left running at full
speed, draining your coin cells while you
are trying to find out what is going
wrong. Thus, disconnect your cable
when you will not need it for the next
few minutes. For frequent flashing, an
external 3V DC power supply may quickly pay for itself. Take care to connect ‘+’
to the outer and ‘–‘ to the inner battery contact. The flashing will work best with
a stable 3V supply – but as long as the voltage supplied exceeds 2.5V it is ok.

B. You may use an ordinary micro USB cable for the connection to your PC if you
install a small custom board developed by Harald Pott, hosting a USB port. With
the board pictured here, your
calculator will be powered through
that USB line while connected.
Thus you will not need the special
programming cable nor an hardware
serial port, and avoid the battery drain observed – but you will need to install this
board. See Appendix H for more information and a detailed installation guide.

Having prepared your calculator, computer, software, file, and cable, follow one of the two
alternative procedures A or B below for transforming an HP-20b or HP-30b into a WP 34S.

GENERAL WARNING: Flashing your HP-20b or -30b will erase the HP firmware –
your business calculator will then be gone. Its firmware will be replaced by the file
you downloaded. Thereafter you will have a WP 34S RPN Scientific – i.e. your
calculator will react as documented in this manual.69
This also means your calculator will not do anything useful for you after step 6 and
before step 9 is successfully completed in the procedures described below. Your
calculator may even look dead – it is not, be assured. If the procedure is interrupted
at any time, don’t worry: remember RESET will always turn your calculator off, and
shall turn it on70 – simply start over at step 1.

69
If – for any reason whatsoever – you want to return to the original HP business calculator firmware, you
can do so by sending the respective binary file instead of calc.bin or its siblings. An HP-30b file is
hosted on our website for example – note it represents the firmware status as known to us today and will
not be maintained.
70
… as long as the batteries are good – even if you do not see anything on the LCD.
WP 34S Owner’s Manual Edition 3.1 Page 148 of 211
A1. Remove the black battery door of your
calculator. Connect the programming cable to
the programming port of the calculator and to
your computer.

A2. Start MySamBa. Pick the port you use for your
cable and select the file you want to transmit.
The window may look like:

Do not start sending yet! Just leave the window open.

A3. Press to turn the calculator on.

A4. Hold down the ERASE button on the cable (do


not release it until step A7).

A5. Press the RESET button on the cable. It will


turn the calculator off.

A6. Press to turn the calculator on again.

A7. Release ERASE now. Press RESET to turn the calculator off again.

A8. Press to turn the calculator on again. It will look dead. Do not worry – see
the note above.

A9. Now click "Send File" in the


MySamBa window and wait for it
to finish transmission (it will take
some 20s). If you have the FTDI
USB/serial adapter mentioned
above, you will see the blue TX
light blinking. If MySamBa
should fail71, return to A3.

A10. Press RESET to turn the calculator off once more. Exit MySamBa. Disconnect the
cable. Close the battery door.

A11. Press and release . Your calculator should turn on as a WP 34S now.

71
Then it throws (a) “Error, could not connect to calculator” or (b) “Unable to connect”. Typical causes are:
(a) a wrong COM port or not the proper driver for that port, (b) the cable makes no proper contact or your
calculator is not turned on. Anyway, this dead looking calculator is a most annoying state since you do not
get the slightest feedback from your system while you keep looping through the steps above. And with the
programming cable connected, your battery will be drained in this time … so press RESET, disconnect,
think, and check before reconnecting and retrying. If it fails repeatedly, turn to p. 182.
WP 34S Owner’s Manual Edition 3.1 Page 149 of 211
If you have installed the custom USB board mentioned above before transforming your
HP-20b or -30b into a WP 34S, your calculator will be powered through this port while
connected to your computer. The standard USB cables, however, do not feature the
buttons ERASE and RESET of the custom programming cable. Thus, the flashing
procedure will work here as follows:

B1. Remove the battery door. Connect


said micro USB cable to the USB port
of the calculator and to your computer.

B2. Start MySamBa. Enter the proper port


and file information as explained in
step A2 above (see also p. 197). Do
not start sending yet – just leave the
window open.

B3. Press to turn the calculator on.


B4. Shorten the top right and the bottom
left of the six pads below the RESET
hole by a removable link
(a paper clip works well).
B5. Use a suitable pin to
press the RESET button
behind the hole. It will
turn your calculator off.
B6. Ensure the link makes
good contact to both
pads. Then press to turn the
calculator on again. It will erase the old firmware.

B7. Remove the shorting link. Press RESET to turn your calculator off again.

B8. Press on the calculator. It will look dead like on the picture. Do not worry –
see the note above.

B9. Click "Send File" in the MySamBa window and wait for it to finish transmission (it will
take some 20s). If it fails71, return to B3.

B10. Press RESET to turn the calculator off once more. Exit MySamBa. Close the battery
door.

B11. Press and release . Your calculator should turn on as a WP 34S now.
You can disconnect the cable unless you want to use it for data transfer from your
emulator (see Appendix D).

WP 34S Owner’s Manual Edition 3.1 Page 150 of 211


Overlays – Where to Get Them and How to Make Them Yourself if Required

After flashing successfully, a keyboard overlay is


very helpful for further work since most labels
deviate from those on said business calculators.

You get fine adhesive vinyl overlays from Eric


Rechlin at http://commerce.hpcalc.org/overlay.php.

You can choose if you want them with white labels on


the prefix keys (the original design, shown left) or with f,
g, and h in their respective colours (shown above).
Supporters of the ‘Great American Divide’ can apply
them with a instead of a as well.
Eric’s vinyls come pre-stamped. Detach the key plate
part of it from the carrier paper first. Adjust it carefully
over your WP 34S and attach it starting with the section
between ENTER and STO, then proceeding towards the
top and bottom. The vinyl shall not contact the keys.
After covering the key plate, detach the individual key
labels and apply them to the respective keys. If you
need to lift a label once attached – e.g. for readjusting it
– a sharp knife will serve you best: put it between key
and label and lift a corner of the label, then take it with
your finger nails or tweezers and peel it off.

If Eric’s vinyls are not available for any reason, you


can make preliminary paper overlays most easily
using the picture shown left. Print this page to scale
(the picture shall be 68mm wide), cut it out, span it over your WP 34S using transparent
adhesive tape, and you are done – the flexibility of this setup allows actuating the keys
though there are no holes in your overlay. It will serve you well until the vinyls will arrive.

WP 34S Owner’s Manual Edition 3.1 Page 151 of 211


Updating Your WP 34S

Whether you bought your WP 34S pre-flashed or did the first flashing yourself, you may
want to keep it up-to-date when a new release is published or after you modified its hard-
ware. We recommend you SAVE your work each time before updating (your WP 34S
will show Saved when successful). Starting the new firmware, this backup will be restored
automatically – if you didn't accidently press the ERASE button on the programming cable.

With the proper binary downloaded as described on p. 147, updating works as follows:

1. Remove the battery door of your WP 34S.

2. Connect the programming cable to the 2. Connect the micro USB cable to the
programming port of your WP 34S and USB port of your WP 34S and to your
to your computer. Do not touch the computer.
cable buttons in the remaining
procedure!

3. Start MySamBa.
Enter the proper port
and file information
(see p. 147ff).
Do not start sending
yet! Just leave the window open.

4. Press to turn your WP 34S on. Then press +


72
simultaneously. It is echoed Debug ON. Release both keys.

5. Now press + simultaneously. It is echoed SAM-BA? boot.


Release only and press it a second time for confirmation, then
release both keys. This resets the Samba boot bit and turns your
WP 34S off.

6. Press . Your WP 34S will look dead. Do not worry – it is not.

7. Click "Send File" in the MySamBa window and wait for it to finish
transmission (it will take some 20s).73

8. Use a suitable pin to press the RESET button behind the hole on the
back of your WP 34S. Your device will turn off. Exit MySamBa.
Disconnect the cable. Close the battery door.

9. Press . Your WP 34S should turn on now with the last backup
Restored and the new firmware loaded. Check using VERS.

72
The next step will only work in debug mode. See Appendix H.
73
If MySamBa fails for any reason, press RESET and disconnect the programming cable. Then take your
time and check footnote 71. Return to step 6 when you want to retry.
WP 34S Owner’s Manual Edition 3.1 Page 152 of 211
I/O Overview

Looking to I/O from your WP 34S calculator, this matter now appears as follows
schematically:

HP82440A/B IR line WP 34S Cable Computer


Printer SEND… WP 34S
⎙… RAM
RECV Emulator

PSTO PRCL Host of


SAVE LOAD... wp34s-, dat-
and bin-files

FM Flashing

SAVE is for battery-fail-safe internal backup of your work as explained on p. 72. The
different flavors of LOAD are for recovering from backup. Alternatively to the commands
SAVE and LOAD, you may hold down (i.e. ) and press one of the following
keys twice:

for backup: Creates a copy of the RAM in FM like SAVE does.

for restore: Restores the most recent backup like LOAD does.

PRCL recalls an individual program from the FM library into RAM. PSTO stores a copy of
an individual program from RAM into the FM library. Note that by recalling a library routine
from FM via PRCL, editing it, and storing it back via PSTO you may modify this part of the
FM library. See the IOP for these commands and Appendix B for details about the RAM
and FM sections of your WP 34S.

The different flavors of SEND are for sending your work from WP 34S RAM to your
computer, where you will enjoy more editing comfort. See Appendix D for more.

WP 34S Owner’s Manual Edition 3.1 Page 153 of 211


APPENDIX B: MEMORY MANAGEMENT

Your WP 34S features 6kB of RAM (of which 4kB are volatile) and 128kB of FM.
The firmware takes some 90% of FM, depending on the file you loaded in flashing. The
remaining, user accessible part of FM may be almost 14kB. Of these, 2kB are reserved for
the backup region corresponding to the 2kB of non-volatile RAM featured; the remainder
may be filled by a collection (a.k.a. library) of different programs.
1 1 1
The remaining part of this appendix covers RAM only. It 2
discusses how the available memory is divided in program area,
local and global data. The 1024 words of non-volatile RAM are
shared by four sectors:
1. Status and configuration data 2

2. Global registers, i.e. general purpose registers and stack


3. Registers used for cumulative statistics (optional)
4. Program memory including subroutine return stack (SRS, 2
3
including local registers if applicable). 4

These sectors are arranged top down as pictured at right to scale.


The center bar shows the RAM configuration in startup default
(but with statistics registers allocated). The left bar shows it for
maximum register space allowed (but without statistics), the right 4
for maximum program space. This appendix covers the variable
boundaries between these sectors.74
4
Status and Configuration Data

This sector of 88 bytes is fixed at the top of available memory and is completely user
transparent. It contains status and modes data, the 30 byte alpha register, and 14 bytes
holding the 112 global user flags.

Global Registers

Global registers are placed near the end of available memory. In startup default memory
layout, the numbered registers R00 to R99 precede the twelve stack and special registers
X, Y, Z, T, A, B, C, D, L, I, J, and K as shown on p. 24. This totals to 112 global registers,
which is the maximum available. Their number can be reduced down to the 12 lettered
registers using REGS (see the IOP). REGS? will return an integer between 0 and 100
corresponding to the number of global numbered registers currently allocated.
REGS controls the lower boundary of the global register sector (abbreviated LBG in the
following). Reducing the number of registers will pull up LBG to higher absolute addresses;

74
A complete copy of the RAM as you configured and loaded it can be written to FM using SAVE. It will
remain safely preserved not only when your WP 34S turns off but also even when its batteries fail. See
Appendix A for more information about flashing and handling of FM in general.
WP 34S Owner’s Manual Edition 3.1 Page 154 of 211
increasing their number will push it down. The memory contents are moved accordingly,
thus preserving the data in the surviving registers. Contents of deallocated registers are
lost; newly added registers are cleared. The lettered registers do not move.

Example: See the global register sector at startup default in the following memory
table. The two rightmost columns show what happens after subsequent execution
of REGS 96 and REGS 98. The registers are loaded with arbitrary values here so
they can be traced easily. LBG is indicated by a red horizontal line.

Default startup memory After executing Then after executing


allocation REGS 96 REGS 98
Relative Relative Relative
Absolute register register register
address Contents address Contents address Contents address

X+11 k = 40.7 R111 = K 40.7 K = R111 40.7 K = R111


… … … … … … …
X+2 z = 4.5 R102 = Z 4.5 Z = R102 4.5 Z = R102
X+1 y = -33.8 R101 = Y -33.8 Y = R101 -33.8 Y = R101
X x = 123.0 R100 = X 123.0 X = R100 123.0 X = R100
X-1 r99 = -13.6 R99 23.1 R95 0.0 R97
X-2 r98 = 67.9 R98 6.4 R94 0.0 R96
X-3 r97 = -45.2 R97 4.8 R93 23.1 R95
X-4 r96 = 9.7 R96 … … 6.4 R94
X-5 r95 = 23.1 R95 … … 4.8 R93
X-6 r94 = 6.4 R94 … … … …
X-7 r93 = 4.8 R93 … … … …
… … … … … … …
X-94 r06 = 62.4 R06 5.7 R02 29.4 R04
X-95 r05 = -0.6 R05 -2.4 R01 81.3 R03
X-96 r04 = 29.4 R04 1.1 R00 5.7 R02
X-97 r03 = 81.3 R03 -2.4 R01
X-98 r02 = 5.7 R02 1.1 R00
X-99 r01 = -2.4 R01
X-100 r00 = 1.1 R00

Note the absolute addresses of R00 up to Rn-1 change after REGS n whenever n
is changed, while their contents are copied. The lettered registers stay at fixed
absolute addresses.
In indirect addressing, zero in the index register points to R00 always. Index values
exceeding the maximum set by REGS will throw an ‘out of range’ error, unless they
fall between 100 and 111 – where the lettered registers live.

WP 34S Owner’s Manual Edition 3.1 Page 155 of 211


The two sectors following in lower memory (summation registers and SRS) are tied to
LBG – their contents will be copied whenever LBG moves. This makes it possible to
execute REGS in the middle of a subroutine without disrupting the program.

Summation Registers

The memory needed for cumulative statistics is allocated separately – these data are no
longer held in global general purpose registers. This allows for higher internal precision
and prevents destroying these data inadvertently. The only way to update statistical data is
via Σ+ and Σ- . The accumulated data are evaluated and recalled by dedicated commands;
they are not accessible by STO or RCL.
The first invocation of Σ+ allocates 70 words for the 14 summation registers.75 They are
inserted between LBG and SRS, pushing the latter down in memory. Depending on the
competing requirements for program and data space, it may be necessary to make room
first (see overleaf).
After CLΣ, CLALL, or RESET, the memory allocated for the summation registers is
released. All pointers are automatically adjusted, so the memory allocation or release will
not disrupt a running program. Recall commands such as e.g. Σxy or SUM will return zero
if no data are allocated; other statistical operations will throw an error if not enough data
are present.

ATTENTION: The summation data will be cleared automatically when a long


program is loaded (from FM or via the serial interface) if the registers would no longer
fit in RAM after that load. You can avoid this by reducing the amount of numbered
registers using REGS before the load attempt. This should move the summation data
out of the way.

Subroutine Return Stack (SRS) and Program Memory

Both share the remaining space at lowest memory addresses.


The SRS is used for return addresses and local data. Its upper boundary is given by LBG
or the lowest summation register if applicable. There is no command to set the size of the
SRS – it fills all the space down to the top program step currently stored. When new
program steps are entered, the SRS is reset, not only to make room but because any
stored address may become invalid by changing the program.
Local data are pushed on the SRS. Thus they cannot overwrite global data; this greatly
increases the flexibility of programs. LOCR n allocates n local registers and a fixed
amount of 16 local flags. It does so by pushing a frame on the SRS containing a marker,
a flag word, and the registers requested (0 to 144). The marker contains the frame size in
words, depending on the precision mode set (see Appendix H). A pointer to this frame in
memory is initialized. If the pointer is zero, no local registers exist. Newly allocated
registers are cleared.

75 2 2 2
Herein, 2 words are employed for Σn, 4 × 8 words for Σx , Σy , Σxy, and Σx y, and 9 × 4 words for the
other sums. If memory allocation for these 70 words fails, an error will be thrown.
WP 34S Owner’s Manual Edition 3.1 Page 156 of 211
Calling LOCR again in the same subroutine will adjust the number of local registers. This
requires data copying since these registers are allocated from low to high addresses and
the SRS grows in the opposite direction. LOCR? will return the number of local registers
currently allocated in the routine you are in.
See overleaf for addressing local data, and for an example of recursive programming. The
SRS must be large enough to hold these data, however, so you may have to make room
first – see next paragraph.

Below the SRS, program memory holds the stored program steps. A typical program step
takes just one word. Multi-byte labels and multi-character alpha strings take two words
each. The total size of program memory depends on the number of global and local
registers allocated, as explained in the following.

Making Room for Your Needs

The 12 special (lettered) registers are always allocated. The SRS has a minimum size of
six words or levels. Everything else is user distributable within the 982 words left for
sections 2 to 4, so:
982 = r + s + p with
r= number of words allocated for global registers. These are 4 per standard
register. There are at least 12 and at most 112 of them. So r varies
between 48 and 896 (this maximum is explained in Appendix H); startup
default is 448.
s = number of words allocated for summation registers (70 if they are used;
startup default is 0).
p = number of words available for program steps and SRS. One step is
already taken by the inevitable final END statement; 6 words is the
minimum size of the SRS. So STATUS will show you a maximum of 933
free words in RAM, meaning up to 927 free program steps. Startup default
is 533 steps. Subroutine nesting and local registers expand the SRS, thus
reducing the program space available.
If, for instance, you need to do statistics and also use 20 global numbered registers,
there will be space for 777 program steps maximum.

You have several options for increasing the free space where you need it (see the picture
on p. 154):
1. Reduce the number of global numbered registers allocated. One register less
typically allows for four additional program steps.
2. Move programs to FM and clear the respective steps in RAM. Four cleared program
steps typically allow for one additional register.
3. Release the summation registers when you do not need them anymore. This space
may be distributed to up to 70 additional program steps, up to 17 additional
registers, or a mix.

WP 34S Owner’s Manual Edition 3.1 Page 157 of 211


Which solution serves you best depends on your application. You may of course combine
options. Use to monitor the free space available and the amount of global and
local numbered registers allocated.

Addressing and Accessing Local Data, Recursive Programming

Global data take relative addresses from 0 to 111 as described on p. 154. So, relative
addresses of local data begin with 112 and may go up to 255 if 144 local registers are
allocated. The first 16 local registers and all local flags may also be directly addressed
using a dot heading the number – the arguments go from .00 to .15, corresponding to
relative addresses from 112 to 127.76 Any registers beyond are only indirectly addressable.
This scheme allows for indirectly addressing
a global register via a global index register (e.g. STO 23 with r23 < 112),
a global register via a local index register (e.g. STO .15 with r.15 < 112),
a local register via a global index register (e.g. STO 47 with r47 ≥ 112), and
a local register via a local index register (e.g. STO .06 with r.06 ≥ 112).

Subroutine calls: XEQ – executed in a program – just pushes the return address on the
SRS before it branches to the target. The subroutine called will keep having access to the
caller's local data as long as it does not execute LOCR itself. As soon as it does, the
pointer to the local data is newly set, and the subroutine called cannot access the caller's
local data anymore.
RTN or POPLR – executed in a program – check if the current SRS pointer points to a
local frame (as explained on p. 156). If true then the pointer is moved above that frame,
and the SRS is searched from this point upwards for another local frame. If such a frame is
found then its pointer is stored; otherwise the pointer to the active local frame is cleared.
RTN will branch to the return address found, while POPLR will just continue execution. So
the current local frame is dropped and the next higher (or older) frame is reactivated if one
exists.
Manually executing RTN, starting a new program with XEQ, SLV, etc., or program editing
will clear the SRS and remove all local registers and flags by clearing the pointer. All such
data are lost then!

Recursive programming: Using local registers allows for creating a subroutine that calls
itself recursively. Each invocation deals with its local data only. Of course the RPN stack is
global so be careful not to corrupt it.
Here is a recursive implementation of the factorial. It is an example for demonstration
only, since this routine will neither set the stack correctly nor will it work for input
greater than some hundred:

76
Only arguments up to 127 are storable in an op-code, hence the limit.
WP 34S Owner’s Manual Edition 3.1 Page 158 of 211
LBL ’FAC’
IP Assume x = 4 when you call FAC. Then it will allocate 1 local register
x>1? (R.00) and store 4 therein. After decrementing x, FAC will call itself.
GTO 00 Then FAC2 will allocate 1 local register (R.002) and store 3 therein.
1 After decrementing x, FAC will call itself again.
RTN
LBL 00 Then FAC3 will allocate 1 local register (R.003) and store 2 therein.
LocR 001 After decrementing x, FAC will call itself once more.
STO .00 Then FAC4 will return to FAC3 with x = 1 . This x will be multiplied by
DEC X r.003 there, returning to FAC2 with x = 2 . This x will be multiplied by
XEQ ’FAC’ r.002 there, returning to FAC with x = 6 , where it will be multiplied by
RCL× .00 r.00 and will finally become 24.
RTN

Switching between Standard Real (SP) and Integer Modes

Your WP 34S starts in standard real mode (DECM) when you get it new. You may use it
for integer computations as well, as shown above many times. Going from DECM to any
integer mode, the values on the current stack will be truncated to integers. Going from
integer mode to DECM, the current stack contents (being all integers) will be converted to
decimal. All other memory contents will stay as they were!
See the fate of some register contents undergoing mode switches in the following
examples, where j, k, r00, and r01 will be checked by recalling them. :

X Y J K R00 R01
Contents at start e.g. 1.1 20.2 300.3 4,000.4 50,000.5 600,000.6
After 2COMP, WSIZE 32, BASE 10 1
d
20
d
3,075
d
40,964
d
512,005
d
6,291,462
d

d d d d
Recall the registers by sRCL 300 4,000 50,000 600,000
d d d d d d
567 STO J, -9 STO 00 -9 567 567 40,964 -9 6,291,462
DECM -396 -389
-9.0 567.0 5.7 4,000.4 3.0 600,000.6
Recall the registers by iRCL 567.0 40,964.0 -9.0 6,291,462.0

Note that identical register contents are interpreted quite differently in DECM and integer
modes. Even very small integers may lead to very large surprises:

R00 R01 R02


Contents at start e.g. 0 2 10
h h h
Then after WSIZE 64, BASE 16 0 2238000000000002 223C000000000001
h h h
Recall the registers by sRCL 0 2 A
h h h
2 STO 01, A STO 02 0 2 A
-398 -397
DECM 0 2 1
Recall the registers by iRCL 0 2 10

WP 34S Owner’s Manual Edition 3.1 Page 159 of 211


Thus take care with indirect addressing!

Example: Start with DECM, WSIZE 64, 2COMP, 0 STO 00, 2 STO 01, 10 STO 02 as
above. So RCL 01 shall recall r02. Let us check. Key in:

10. just for verification: OK


h
A turn to BASE 16
Out of range Error compare the table above
h
2
h
000000000001 now this works in hex mode, too
h
223C
18
2,46684669589 turn to DECM
0. compare the table above –
indirection has to ignore all
fractional components. Otherwise
ISG/DSE loops all simply would
not work.

All this is caused by the internal representation of numbers: While integers are simply
stored as such (allowing for n ≤ 1.84∙1019 in UNSIGN or |n| ≤ 9.22∙1018 in the other
modes using 64 bits), standard floating point numbers are stored using a format as follows:
Real zero is stored as integer zero, i.e. all bits cleared.
The mantissa of a real number (also known as significand in this context) is
encoded in five groups of three digits. Each such group is packed into 10 bits
straight forward, meaning e.g. 55510 = 10 0010 11002 or 99910 = 11 1110 01112 =
3E716 . So the 15 rightmost decimal digits of the significand take the least significant
50 bits. Trailing zeroes are omitted, so the significand will be right adjusted.
The most significant (64th) bit takes the sign of the mantissa.
The remaining 13 bits are used for the exponent and the leftmost digit of the
mantissa. Of those 13, the lowest 8 are reserved for the exponent. For the top 5 bits
it gets complicated77: if they read …
o 00ttt, 01ttt, or 10ttt then ttt takes the leftmost digit of the
significand (0 – 710), and the top two bits will be the most significant bits of
the exponent;
o 11uut then t will be added to 10002 and the result (810 or 910) will become
the leftmost digit of the significand. If uu reads 00, 01, or 10 then these two
will be the most significant bits of the exponent. If uu reads 11 instead, there
are codes left for encoding special numbers (e.g. infinities).
In total, we get 16 digits for the mantissa and a bit less than 10 bits for the
exponent: its maximum is 10 1111 11112 (i.e. 76710). For reasons becoming
obvious below, 398 must be subtracted from the value in this field to get the true
77
Don’t blame us – this part follows the standard IEEE 754.
WP 34S Owner’s Manual Edition 3.1 Page 160 of 211
exponent of the number represented. The 16 digits of the significand allow for a
range from 1 to almost 1016.

Rewarding your patience so far, we will show you some illustrative examples of the
encoding in your WP 34S instead of telling you more theory:

Floating point Hexadecimal value stored Bottom bits split Top 14 bits in binary Stored
number in groups of 10 representation expo-
nent
1. 22 38 00 00 00 00 00 01 0010 0010 0011 10 398
-1. A2 38 00 00 00 00 00 01 1010 0010 0011 10 398
111. 22 38 00 00 00 00 00 6F 0010 0010 0011 10 398
111.111 22 2C 00 00 00 01 bC 6F 06F 06F 0010 0010 0010 11 395
-123.000123 A2 20 00 00 07 b0 00 7b 07b 000 07b 1010 0010 0010 00 392
9.99∙1099 23 bC 00 00 00 00 03 E7 0010 0011 1011 11 495
1∙10–99 20 AC 00 00 00 00 00 01 0010 0000 1010 11 299
1∙10–383 00 3C 00 00 00 00 00 01 0000 0000 0011 11 15

The last number is the smallest that can be entered directly. Dividing it by 1015
results in 1∙10–398, being stored as hexadecimal 1. Divide this by 1.999 999 999 99
and the result will remain 1∙10–398 in default rounding mode (and in RM 1, 2, 3,
and 5, see RM). Divide it by 2 instead and the result will become zero.

Look at the other end:


9.999 999 77 FF E7 F9 FE 7F 78 00 9 3E7 3E7 0111 0111 1111 11 767
999 99∙10384 3E7 3dE 000

This number (featuring the digit 9 twelve times) is the greatest that can be entered
numerically. Adding 9.999∙10372 to it will display 1∙10385 …
1∙10385 77 FF E7 F9 FE 7F 9F E7 9 3E7 3E7 0111 0111 1111 11 767
3E7 3E7 3E7

… which is stored as 9.999 999 999 999 999∙10384. This is the greatest number
representable. All this follows Decimal64 floating point format, though not exactly.

Additionally, your WP 34S features three special numbers:


Infinity 78 00 00 00 00 00 00 00 0111 1000 0000 00 n/a
-Infinity F8 00 00 00 00 00 00 00 1111 1000 0000 00 n/a
not nuMmeric 7C 00 00 00 00 00 00 00 0111 1100 0000 00 n/a

These special numbers are legal results on your WP 34S if flag D is set.

WP 34S Owner’s Manual Edition 3.1 Page 161 of 211


APPENDIX C: MESSAGES AND ERROR CODES

There are some commands generating messages, be they in the numeric or in the dot
matrix section of the display. Of these, DAY, DAYS+, ERR, STATUS, VERS, and WDAY
were introduced above in the section about display. Others are PROMPT, αVIEW and
more alpha commands, and the test commands (see p. 67). Also two constants will return
a special message when called.
Furthermore, there are a number of error messages. Depending on error conditions, the
following messages will be displayed in the mode(s) listed:

Error
Message Mode(s) Explanation and examples
code

Invalid date format or incorrect date or time in


2 DECM
input, e.g. month >12, day >31 .

Invalid digit in integer input, e.g. 2 in binary or 9


9 Integer in octal mode. Will be displayed as long as the
respective key is pressed.

Caused by calling an operation in a mode


where it is not defined, e.g. calling a constant in
13 All
a program written in DECM but executed in an
integer mode.

An argument exceeds the domain of the


mathematical function called. May be caused by
roots of negative numbers or logs of x ≤ 0
1 ¬α (both if not preceded by ), by 0 / 0, (0),
tan(90°) and equivalents, by artanh(x) for
Re x 1 , by arcosh(x) for Re x 1 , etc.78

No more space in FM. Delete a program from


23 All
FM to regain space.

May appear in an attempt running an old


program containing a command which turned
7 All
nonprogrammable after said program was
written.

Set when there is a checksum error either in FM


18 All or as part of a serial download. It is also set if a
FM segment is otherwise not usable.

Similar to error 1 but a parameter specified in J


16 ¬α or K is out of valid range for the function called.
May appear e.g. if LgNrm is called with j < 0.

78
Note that e.g. tan(90°) and logs of 0 are legal if flag D is set. See the end of this appendix.
WP 34S Owner’s Manual Edition 3.1 Page 162 of 211
Error
Message Mode(s) Explanation and examples
code

17 ¬α See Appendix D.

A matrix isn't square when it should be.


21 DECM
Matrix sizes aren't miscible.

May be thrown by DATE, TIME, SETDAT, or


24 All
SETTIM. See Appendix H for curing.

20 DECM The solver did not converge.

6 All Attempt to address an undefined label.

A number exceeds the valid range. This can


be caused by specifying decimals >11, word
size >64, negative flag numbers, integers
≥2 64, hours or degrees >9000, invalid times,
denominators ≥9999, etc.
A register or flag address exceeds the valid
range of currently allocated registers. May
8 All also happen in indirect addressing or calling
nonexistent local addresses.
An R-operation (e.g. R-COPY) attempts
accessing invalid register addresses.
A matrix descriptor would go beyond the
registers available or a row or column index
is too large.

No more space in RAM. May be caused by


attempts to write too large programs, allocate
11 All too many registers, and the like. May happen
also in program execution due to too many local
data dynamically allocated (see p. 156).

Attempt to use a LU decomposed matrix for


solving a system of equations.
22 DECM
Attempt to invert a matrix when it isn't of full
rank.

STOS or RCLS attempt using registers that


would overlap the stack. Will happen with e.g.
12 All
SSIZE = 8 and STOS 94 (if REGS 100 is
set).

WP 34S Owner’s Manual Edition 3.1 Page 163 of 211


Error
Message Mode(s) Explanation and examples
code

A statistical calculation was started based on


15 DECM too few data points, e.g. regression or standard
deviation for < 2 points.

Keyboard input is too long for the buffer. Will


10 All happen e.g. if you try to enter more than 12
digits.

An instruction with an undefined operation code


3 All occurred. Should never happen – but who
knows?

Integer,
14 Register content is too big for the word size set.
¬STO

19 All Attempt to delete or edit program lines in FM.

Division of a number > 0 (or < 0) by zero.


4
Divergent sum or product or integral.
¬α,
¬STO, Positive (or negative) overflow in DECM (see
¬flag D p. 40).
5 Error 5 is also thrown for a logarithm of (+)0,
while a logarithm of -0 returns NaN.

This is not an error but only a message. It will


DECM
25 be returned by MSG 25 and works exclusively
& RCL
in programs.

Each error message is temporary (see p. 38), so or will erase it and allow
continuation. Any other key pressed will erase it as well, but will also execute with the
stack contents present. Thus, another easy and safe return to the display shown before
the error occurred is pressing an arbitrary prefix twice.

A final note about flag D: if it is set, errors 4 and 5 will not occur at all, and error 1 will
happen less frequently, since ±∞ and NaN are legal results then (see the respective entry
in CONST and the very end of Appendix B).

WP 34S Owner’s Manual Edition 3.1 Page 164 of 211


APPENDIX D: THE WP 34S EMULATOR ON YOUR COMPUTER

For the emulator, run \trunk\windows\bin\wp34sgui.exe – it features the identical


function set as your WP 34S calculator. While tactile feedback, boot speed, pocketability
and battery life suffer, some things become easier, e.g. printing (see below). What else is
different?

Typically the emula-


tor is operated by the
computer mouse.
You click on the
respective areas of
the keyboard image
instead of pressing
keys. Right clicking
on a key area is a
shortcut to its green
h-shifted label. And
there is one important
extra: a right click on
the WP logo top right
opens a menu as
seen here.

It contains links to the


project website and
the manual (under
<Help>) as well as to
three different skins
for the emulator.

The ‘medium’ skin is shown several times in this manual; the other two are smaller and
are displayed overleaf. All three skins are printed to scale here.

Furthermore, this menu allows you to import or export numbers, export alpha as text or
the complete ‘LCD’ screen as a picture to the clipboard for use in other computer
applications.

If your computer has a numeric keypad, the digit and arithmetic operation keys there
are shortcuts to the corresponding ‘keys’ of the emulator. The up and down cursor keys
are connected to and , Backspace and Del to , and Enter and CR to .

WP 34S Owner’s Manual Edition 3.1 Page 165 of 211


Data Transfer between Your WP 34S and Your Computer

The emulator and the calculator can talk to each other over the cable used for flashing.
Either one of the cables mentioned on p. 147 will do, or you need a modified calculator as
described here:
http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv020.cgi?read=186826 .
Remember the programming cable draws current from the batteries of your WP 34S, so
disconnect it from your WP 34S as soon as it is no longer needed.

In the emulator directory, a plain ASCII text file wp34s.ini must be placed that contains
just a single line (terminated by CR/LF) stating the name of the computer port used for
serial communication, such as e.g.
COM2:

The following commands (all stored in P.FCN catalog) allow for sending programs,
registers or all RAM from your WP 34S to your computer or vice versa. With a suitable
cable it would be even possible to transfer data directly between two WP 34S.

WP 34S Owner’s Manual Edition 3.1 Page 166 of 211


On the receiving device, enter RECV. It will display Wait... .

On the sender you have three alternatives:


1. SENDP will send the current program from RAM. It will display Wait... as well.
After successful transmission, the sender will show OK and the receiver will
display Program .
2. SENDR will send all the global numbered registers allocated, displaying
Wait... . After successful transmission, the sender will show OK and the
receiver will display Register .
3. SENDA will send the complete two kilobytes of non-volatile RAM, displaying
Wait... . After successful transmission, the sender will show OK and the
receiver will display All RAM .

If your WP 34S receives data from the emulator on your computer, these data
will be stored in the respective RAM section directly. If the emulator receives
data from your WP 34S, these data may be used immediately as well, and will
be part of the file wp34s.dat which is written when exiting the emulator (see
overleaf).

The little = annunciator is lit while the serial port is in use. Take to abort the
communication if necessary.

Each of the four commands RECV, SENDP, SENDR, and SENDA features a fixed
timeout of some 10s for setting up the connection. After an interval of inactivity of said
length, an I/O Error is thrown indicating no communication has occurred. If such
an error appears in the middle of a transmission, try again.

Using a WP 34S without the crystal installed, you may also get an I/O error
because of its baud rate setting being a bit too far off. To determine the speed,
use the loop
CLx
INC X
BACK 001
and let it run for 30s while connected to the computer. The expected result at
nominal speed is around 123,000 for v3.1. On a crystal-less WP 34S, the I/O
commands accept a correction factor in percent in X. Divide the counts you
measured by 123,000 and leave it in X for RECV or SEND… Already some 2%
may make a significant difference. Values between 80 and 120 are accepted –
all other are ignored. On the emulator or on a WP 34S with the crystal installed
and initialized, x will be ignored in those commands.

With your data transmitted successfully from your WP 34S to your computer, see
section 4 of the separate WP 34s Assembler Tool Suite User Guide for the way to
extract .wp34s-files from the file wp34s.dat. The .wp34s-files may be opened and
edited by a suitable text file editor, e.g. WordPad. After completion of editing and
assembling the routine(s) – and after successful tests on the emulator – you may
WP 34S Owner’s Manual Edition 3.1 Page 167 of 211
revert the transmission: issue RECV on your WP 34S and send the respective
modified data from your computer back to your WP 34S. What works on the emulator
shall work on your calculator as well (provided the necessary hardware is installed) –
both feature the same set of functions. Your WP 34S is one of the smallest
calculators in a commercial package allowing such a bidirectional data exchange and
backup.

The newer Qt-based emulators for Windows and MacOS contain a setup option for the
serial interface. They will eventually replace the current Windows emulator. with the
commands above.

Mapping of Memory Regions to Emulator State Files

Region State file Remarks

Backup Bup wp34s-backup.dat This file is created by SAVE.

Flash This file is written whenever a flash command


wp34s-lib.dat
library Lib is executed.

Backup of the emulator RAM area (registers,


RAM RAMm wp34s.dat state, and programs) – this file is written only
when exiting the emulator.

All these files are only read into memory at emulator startup.

Simulating a Printer on Your Computer

For simulated printing,


use Christoph Gieße-
link’s HP82240B Printer
Simulator, available
here:
http://hp.giesselink.com/h
p82240b.htm .
Installation and use are
self-explanatory. It is an
autonomous application,
so you have to start and
exit it independently. The
‘print’ shown here was
generated by ⎙STK and
⎙REGS. Guess what
stack size was set?

WP 34S Owner’s Manual Edition 3.1 Page 168 of 211


APPENDIX E: CHARACTER SETS

The following table shows the complete dot matrix character set as implemented in big and
small font sorted according to the hexadecimal character codes (Unicode). Characters with
codes < 2016 are for control purposes – some of them (4, 1010, 2710) may be useful for
HP82240B control.

0 1 2 3 4 5 6 7 8 9 A B C D E F 0 1 2 3 4 5 6 7 8 9 A B C D E F
000x ␀␁␂␃␄␅␆␇␈␉␊␋␌␍␎␏ ␀␁␂␃␄␅␆␇␈␉␊␋␌␍␎␏
001x ␐␑␒␓␔␕␖␗␘␙␚␛␜␝␞␟ ␐␑␒␓␔␕␖␗␘␙␚␛␜␝␞␟
002x ␠ ! “ #$%& ' ( ) *+ , - . / ␠ ! “ #$ % & ‘ ( ) * + , - . /
003x 0 1 2 3 4 5 6 7 8 9 : ; < = > ? 0 1 2 3 4 5 6 7 8 9 : ; < = > ?
004x @ A B C D E F G H I J K L M N O @ A B C D E F G H I J K L M N O
005x P Q R S T U V W X Y Z [ \ ] ^ _ P Q R S T U V W X Y Z [ \ ] ^ _
006x ` a b c d e f g h i j k l m n o ` a b c d e f g h i j k l m n o
007x p q r s t u v w x y z { ¦ } ~ p q r s t u v w x y z { ¦ } ~
008x
009x
00Ax £ ¥ ¦ © ¬ £ ¥ ¦ © ¬
00Bx ° ± ² ³ µ ¶ ° ± ² ³ µ ¶
00Cx À Á Ã Ã Ä Å Æ Ç È É Ê Ë Ì Í Î Ï À Á Ã Ã Ä Å Æ Ç È É Ê Ë Ì Í Î Ï
00Dx Ð Ñ Ò Ó Õ Õ Ö × Ø Ù Ú Û Ü Ý ß Ð Ñ Ò Ó Õ Õ Ö × Ø Ù Ú Û Ü Ý ß
00Ex à á ã ã ä å æ ç è é ê ë ì í î ï à á ã ã ä å æ ç è é ê ë ì í î ï
00Fx ð ñ ò ó õ õ ö ø ù ú û ü ý ÿ ð ñ ò ó õ õ ö ø ù ú û ü ý ÿ
010x Ā ā Ā ā Ć ć Č č Ā ā Ā ā Ć ć Č č
011x Ē ē Ē ē Ē ē Ē ē Ē ē Ē ē
012x ħ Ĩ ĩ Ĩ ĩ Ĩ ĩ ħ Ĩ ĩ Ĩ ĩ Ĩ ĩ
013x
014x Ň ň Ō ō Ō ō Ň ň Ō ō Ō ō
015x Ř ř Ř ř
016x Š š Ũ ũ Ũ ũ Ũ ũ Ů ů Š š Ũ ũ Ũ ũ Ũ ũ Ů ů
017x ŷ Ÿ Ž ž ŷ Ÿ Ž ž
023x ȳ ȳ
039x A B Γ Δ E Z H Θ I K Λ M N Ξ O A B Γ Δ E Z H Θ I K Λ M N Ξ O
03Ax Π P Σ T Y Φ X Ψ Ω Π P Σ T Y Φ X Ψ Ω
30Bx α β γ δ ε ζ η θ ι κ λ μ ν ξ o α β γ δ ε ζ η θ ι κ λ μ ν ξ o
03Cx π ρ σ τ υ φ χ ψ ω π ρ σ τ υ φ χ ψ ω
1D6x ᵡ ᵤ ᵧ ᵡ ᵤ ᵧ
1E8x Ẁ ẋ Ẁ ẋ
201x ' ' ' “ “ “ ‘ ‘ ‘ “ “ “

WP 34S Owner’s Manual Edition 3.1 Page 169 of 211


0 1 2 3 4 5 6 7 8 9 A B C D E F 0 1 2 3 4 5 6 7 8 9 A B C D E F
207x ⁸ ⁺ ⁸ ⁺
208x ₀ ₁ ₂ ₈ ₀ ₁ ₂ ₈
209x ₐ ₑ ₓ ₖ ₘ ₙ ₚ ₛ ₐ ₑ ₓ ₖ ₘ ₙ ₚ ₛ
20Ax € €
219x ← ↑ → ↓ ↕ ← ↑ → ↓ ↕
21Cx ⇆ ⇆
221x √ ∞ √ ∞
222x ∫ ∫
223x
224x ≈ ≈
225x
226x ≠ ≤ ≥ ≠ ≤ ≥
239x ⎙ ⎙
249x ⒝ ⒞ ⒟ ⒝ ⒞ ⒟
24Ax
24Bx ⒲ Ⓖ ⒲ Ⓖ
24Cx
24Dx ⓕ ⓖ ⓗ ⓕ ⓖ ⓗ
260x ☉ ☉
264x ♁ ♁

In addition, there are two raster fonts supplied. Characters printed on darker grey show
identical patterns in all four fonts. Characters from 00F0 to 00FF (printed on yellow here)
cannot be the last ones in a group of three (see LBL etc.). Note the internal character
sorting in your WP 34S differs from Unicode – the characters of these four dot-matrix fonts
are repeated in the internal order at codes greater than E000. Do not use these high codes
in texts, however, since some programs (e.g. Preview) may not read them correctly.

For the numeric seven-segment display and the annunciators top right, there is another
character set and font provided:

0 1 2 3 4 5 6 7 8 9 A B C D E F Remarks
002x ! " # $ % & ' ( ) * + , - . /
003x 0 1 2 3 4 5 6 7 8 9 : ; < = > ?
004x @ A B C D E F G H I J K L M N O Note M

005x P Q R S T U V W X Y [ \ ] ^ _ Note W

006x ` a b c d e f g h i j k l m n o Note m

007x p q r s t u v w x y Note w
008x
WP 34S Owner’s Manual Edition 3.1 Page 170 of 211
0 1 2 3 4 5 6 7 8 9 A B C D E F Remarks
009x
00Ax ¡ ¢ £ ¤ ¥ ¦ § ¨ For STATUS
For high bits
00Bx · ¸ ¹ in BASE 2
00Cx
00Dx

00Ex à áâ ãäåæçèé Annunciators

00Fx ñ ò ó ô õ ö ÷ ø ù ÿ Segments

All fonts are available at


http://wp34s.svn.sourceforge.net/viewvc/wp34s/trunk/windows/winfont/ .

WP 34S Owner’s Manual Edition 3.1 Page 171 of 211


APPENDIX F: CORRESPONDING OPERATIONS TO THE HP-42S AND -16C
FUNCTION SETS

In the IOP, the corresponding functions of vintage HP calculators were mentioned under
the respective entry of your WP 34S. The tables below revert this in a way: the first shows
the functions of the HP-42S and the corresponding ones of your WP 34S unless they carry
identical names and are either both keyboard accessible or both stored in a catalog or
menu. There is an analog table for the HP-16C functions on p. 178 below. Functional
differences of homonymous commands are covered in the IOP.

HP-42S

Remarks printed on light grey indicate commands being either default settings or keyboard
accessible on your WP 34S while you must use a menu on the HP-42S.

HP-42S WP 34S Remarks

ABS Press .
ACOSH Press .
ADV ⎙ADV
AGRAPH n/a The LCD of the HP-30b has only a very small dot matrix
section (see p. 35). Striving for full fledged graphics therein
would be futile.
AIP αIP
ALENG αLENG
ALL Press .
ALLΣ Superfluous Your WP 34S always runs in ALLΣ mode.
See the description of alpha mode on p. 61.
AND Press .
AOFF αOFF
AON αON
ARCL αRCL or αRC#
AROT αRL or αRR
ASHF αSL or αSR
ASINH Press .
n/a The LCD of the HP-30b does not allow soft keys. This
prevents us from offering you a CUSTOM menu.
ASTO αSTO Press in alpha mode.
ATANH Press .
ATOX α x
AVIEW αVIEW

WP 34S Owner’s Manual Edition 3.1 Page 172 of 211


HP-42S WP 34S Remarks

Superfluous HEXM, DECM, OCTM, BINM, as well as the LOGIC


operations AND, OR, XOR, and NOT are on the keyboard of
your WP 34S in all modes but alpha. A…F are provided in
integer modes if applicable. The functions corresponding to
BIT? and ROTXY are then in X.FCN.
BASE+ etc. Superfluous Your WP 34S executes these arithmetic commands in integer
modes automatically.
BEEP n/a The HP-30b does not contain the necessary hardware.
BINM BASE 2 Press .
BIT? BS?
Replaced This is split in various catalogs on your WP 34S.
CF Press .
CLA CLα Press in alpha mode.
CLD Superfluous Any keystroke will clear a temporary message.
CLΣ, CLP, CLST, CLA, and CLX are on the keyboard.
CLKEYS n/a See ASSIGN.
CLLCD n/a Use to clear the dot matrix section. See AGRAPH.
CLMENU n/a See ASSIGN.
CLP Press .
CLRG CLREGS
CLST CLSTK Press .
CLV n/a The hardware of the HP-30b does not provide space for
variables. Thus, variable-based commands (e.g. many matrix
commands) cannot be featured on your WP 34S.
CLx Press .
CLΣ Press .
COMB Press .
n/a The hardware of the HP-30b does not provide space for
different data types. Thus, complex numbers have to be put in
two registers each on your WP 34S as described on p. 29.

CORR Press .
COSH Press .
CPXRES
n/a See COMPLEX.
CPX?
n/a See ASSIGN.
DECM Press .
DEG Press .

WP 34S Owner’s Manual Edition 3.1 Page 173 of 211


HP-42S WP 34S Remarks

DEL n/a Not featured. Too dangerous, in our opinion.


DELAY ⎙DLAY
DELR
n/a Matrix commands: see CLV.
DIM & DIM?
Superfluous FIX, SCI, ENG, ALL, RDX., and RDX, are on the keyboard.
DSE Press .

EDIT & EDITN n/a Matrix commands: see CLV.


ENG Press .
EXITALL n/a See ASSIGN.
FCSTX x
FCSTY ŷ Press .
FIX Press .
SF and CF are on the keyboard.
FNRM n/a Matrix command: see CLV.
FP Press .
GAMMA Γ
GETKEY KEY?
GETM n/a Matrix command: see CLV.
GRAD Press .
GROW n/a Matrix command: see CLV.
HEXM BASE 16 Press .
I+ and I-
n/a Matrix commands: see CLV.
INDEX
INPUT PROMPT
INSR n/a Matrix command: see CLV.
INTEG ∫ Press .
INVRT M–1
IP Press .
ISG Press .
J+ and J- n/a Matrix commands: see CLV.
KEYASN
n/a See ASSIGN.
KEYG & KEYX
RCL L
LBL Press .

WP 34S Owner’s Manual Edition 3.1 Page 174 of 211


HP-42S WP 34S Remarks

LCLBL n/a See ASSIGN. Nevertheless, your WP 34S provides local


labels as described on p. 67.
LINΣ Superfluous Your WP 34S always runs in ALLΣ mode.
LIST n/a Use ⎙PROG instead.
LOG LOG10 Press .
MAN CF T This print mode is startup default on your WP 34S.

n/a Matrix commands: see CLV.


MAT?
MEAN x Press .
MENU n/a See ASSIGN.

MVAR n/a See ASSIGN.


N! x! Press .
NEWMAT n/a Matrix command: see CLV.
NORM n/a Not featured.
NOT Press .
OCTM BASE 8 Press .
OLD n/a Matrix command: see CLV.
ON n/a Not featured.
OR Press .
PERM Press .
LBL, RTN, VIEW, PSE, ISG, and DSE are on the keyboard.
PGMINT Contained in ∫
PGMSLV Contained in SLV
PI π
PIXEL gSET See AGRAPH.
POLAR n/a See COMPLEX.
POSA n/a The alpha register contains 30 characters maximum.
Automatic searching would be overkill, in our opinion.
PRA ⎙α
PRLCD ⎙PLOT See AGRAPH.

PRX is on the keyboard.


Replaced COMB, PERM, N!, and RAN are on the keyboard. SEED is in
STAT; GAM(MA) is in X.FCN.
PROFF CF T Press .

WP 34S Owner’s Manual Edition 3.1 Page 175 of 211


HP-42S WP 34S Remarks

PRON SF T Press .
PRP ⎙PROG
PRSTK ⎙STK
PRUSR
n/a See CLV.
PRV
PRX ⎙r X Press except in alpha mode.
PRΣ ⎙Σ
PSE Press .
PUTM n/a Matrix command: see CLV.
QUIET n/a See BEEP.
RAD Press .
RAN RAN# Press .
RCLEL & RCLIJ n/a Matrix commands: see CLV.
RDX,
Press except in alpha mode.
RDX.
REALRES
n/a See COMPLEX.
REAL?
RECT Superfluous This mode is permanent on your WP 34S. See COMPLEX.
RND ROUND Press .
RNRM n/a Matrix command: see CLV.
ROTXY RL, RLC, RR,
and RRC
RSUM
n/a Matrix commands: see CLV.
R<>R
SDEV s Press .
SF Press

SINH Press .
SIZE REGS
SLOPE L.R.
SOLVE SLV Press .
Contained in SLV
SQRT √
MEAN, SDEV, FCSTY, and CORR are on the keyboard; the
MODL setting commands are in MODE.
STOEL & STOIJ n/a Matrix commands: see CLV.

WP 34S Owner’s Manual Edition 3.1 Page 176 of 211


HP-42S WP 34S Remarks

STR? n/a See CLV.


TANH Press .
TONE n/a See BEEP.
n/a See ASSIGN.
TRACE SF T Press .
TRANS TRANSP
C
UVEC SIGN
VARMENU n/a See CLV.
VIEW Press .
WMEAN xw
WRAP n/a Matrix command: see CLV.
X<> Press .
X<0? & X<Y? x< ?
X≤0? & X≤Y? x≤ ?
X=0? & X=Y? x= ? Press .
X≠0? & X≠Y? x≠ ? Press .
X≥0? & X≥Y? x≥ ?
X>0? & X>Y? x> ?
XOR Press .
XTOA x α
YINT L.R. Alternatively, press .
∫f(x) Contained in ∫
ΣREG & ΣREG? Superfluous See Appendix B.
DEC Superfluous Conversions are done automatically in integer modes.
DEG ° rad Press .
H.MS Press .
HR Press .
OCT Superfluous Conversions are done automatically in integer modes.
POL Press .
RAD rad ° Press .
REC Press .
%CH Δ% Press .

WP 34S Owner’s Manual Edition 3.1 Page 177 of 211


HP-16C

The table for the functions of the HP-16C is sorted following their appearance on its
keyboard, starting top left. As for the HP-42S, only functions carrying different names on
both calculators are listed. HP-16C functions neither mentioned here nor printed on the
keyboard of your WP 34S are all found in the catalog X.FCN in integer modes.

HP-16C WP 34S Remarks

, RL In X.FCN.
, RR In X.FCN.

, RLC In X.FCN.
, RRC In X.FCN.
nBITS In X.FCN.

DBL / In X.FCN.

Superfluous Every register may be used for indirection.

BS? In TEST.

, LZON, LZOFF In MODE. Control display of leading zeros.


, , Carry.
, , Overflow.
FS? In TEST.

Superfluous Every register may be used for indirection.

Note there is also CLPALL.


CLREGS In P.FCN.
Superfluous See p. 35
n/a

WP 34S Owner’s Manual Edition 3.1 Page 178 of 211


HP-16C WP 34S Remarks

1COMP
2COMP In MODE. Note there is also SIGNMT.
UNSIGN

x≤ ?
x< ?
In TEST. Note there are more tests covered there.
x> ?

(
)

WP 34S Owner’s Manual Edition 3.1 Page 179 of 211


APPENDIX G: TROUBLESHOOTING GUIDE

Overcoming Troubles in Calculator Operation

Symptoms Possible Cause Recommended Therapy

After entering a decimal number such


as 32.4567, the display shows Inadvertently
entered an
integer mode. Press to return to
or similar. default floating point mode.

After entering a decimal number such


as 32.4567, the display shows Inadvertently
entered a fraction
mode. Press to return to
or similar. default floating point mode.

After entering a decimal number such


Inadvertently
as 32.4567, the display shows
entered
programming
Press to return to run
mode.
or similar. mode.

You are working in binary integer


mode and the display shows
Long integer
number displayed
or alike with See on p. 53 how to deal
such strange quotes top right. with such numbers.

Display looks like


Inadvertently
entered
or similar. programming
Calculation commands are not exe- mode. Press to return to run
cuted. mode.

Display looks strange, but not like the Inadvertently


patterns shown so far. entered a
browser or a Press to return to pre-
catalog. vious mode and display.

All input goes to the top line. Inadvertently


entered alpha Press to return to pre-
mode. vious mode.

WP 34S Owner’s Manual Edition 3.1 Page 180 of 211


Symptoms Possible Cause Recommended Therapy

After entering or alike, an


unwanted decimal comma shows up.
How to get the requested point?
Wrong radix mark Press .
set.
After entering or alike, an Alternatively, choose a suit-
unwanted decimal point is displayed able regional setting from
instead of a comma. the table on p. 39.

After entering a decimal number such


as 3.4567, the display just shows 3. Wrong output
(there is a radix mark, but decimals format set. Set 00 to see all
are missing). the decimals you entered.

Your WP 34S does not understand the Choose a suitable regional


Wrong date or
dates or times you enter (anymore). setting from the table on p.
time format set.
39.

After entering hours, minutes, and


Wrong output
seconds, only hours and minutes are Set 4 to see the
format set.
displayed. seconds as well.

After entering a date such as Wrong output Press 6 to see the


11.112012, the year is truncated. format set. complete year.

In long numbers, there are unwanted If in an integer mode, press


points or commas every four, three, or for
two digits. E3ON or SEPON SEPOFF;
is set. if in a floating point mode,
press
for E3OFF.

After converting a number such as e.g.


0.46875 into a fraction, you get
DENMAX is too
low. Press 0
. While being
true, this result is not as precise as to get the maximum
you want it. precision in fraction mode.

After converting a number such as e.g.


DENFAC or
32 into a fraction, you get a result Press
DENFIX is set.
deviating from 1/32. to return to DENANY.

WP 34S Owner’s Manual Edition 3.1 Page 181 of 211


If you experience a strange situation not mentioned here, we recommend you check your
display for annunciators as they are shown on p. 36.

Overcoming Flashing Troubles

Sometimes it seems the calculator just does not want to be flashed. You may have tried
the procedure described in Appendix A repeatedly, but the display stubbornly stays blank.
Do not despair79 – the following procedure will help:

1. If there is still a cable linking calculator and computer, remove it from both devices
a.s.a.p.

2. Check the batteries.


a. Open the battery door and remove both batteries.
b. Check their voltage. Both batteries shall produce
2.8 … 3.2 V. If it is less, replace both of them with
fresh coin cells type CR 2032.
c. Insert the batteries properly – with their positive
poles up – like pictured here. Now your calculator
is well prepared. Leave the battery door open.

3. Check your computer interface and connecting cable.


a. You have either (i) a programming cable or (ii) a micro USB cable. The latter
requires a custom board installed in your calculator – if this installation was done
and verified according to Appendix H you can be confident it is ok.
b. Check both sides of your cable: are there any bent or dirty pins? If not and the
cable also shows no signs of being cut or torn, it should be ok. If dubious, use
another cable of the same kind.
c. Connect your chosen cable to your computer. For case (i), use a suitable
converter if required (see p. 147). Using such a converter with a PC, you shall see
it listed in <System Control> <Device Manager> as a “USB Serial Converter”
(remember its port) – else this converter must be replaced.
d. Check the contacts at your calculator being dirt-free. Connect the cable to your
calculator. For case (ii) and using a PC, you shall see it listed in <System Control>
<Device Manager> as a “FT230X USB Half UART” (remember its port) – else
there is a problem with the cable or the custom board (see p. 197).

4. Now all the hardware is well prepared. Start MySamBa on your computer and enter
the port (as determined in 3.c or 3.d) and the file information (see p. 147). Do not
start sending yet, just leave the window open!

79
No such calculator ended as a paperweight just by flashing yet (but many batteries were drained for sure
by programming cables staying connected for too long and / or by calculators quietly continuing running full
speed).
WP 34S Owner’s Manual Edition 3.1 Page 182 of 211
5. Focus on your calculator again – it is connected to your computer now. Press
firmly for two seconds.
Does the calculator display react now? Fine – continue with step 4 on p. 149 for first
flashing, case (i), or on p. 150 for case (ii), or on p. 152 for updating your WP 34S.
If the display stays blank, however, press RESET and continue with the next step
here.

6. Verify the port is known to the operating system. On a


PC, open the start window and enter CMD – this will
open the good old DOS window; therein type MODE
COMx with x being the port number as determined in
3.c or 3.d. MODE should return data like these here
or the equivalent in your language. Do not care for
particular values since the parameters will be
controlled by MySamBa anyway.
If you get a similar message, continue with step 3 on
p. 149 for case (i), or on p. 150 for case (ii).
If the port is unknown, however, return to step 3.c above and try another one.

Two of our early HP-20b-based calculators (without USB boards built in) developed
broken serial input lines in service. These lines cannot be repaired easily. Those are the
only cases we know so far a WP 34S showing a blank display became unusable since it
cannot be flashed anymore.

WP 34S Owner’s Manual Edition 3.1 Page 183 of 211


APPENDIX H: ADDITIONAL INFORMATION FOR ADVANCED USERS

The information provided here may ease your work when you are going to do advanced
computations, especially programming. This section also covers optional hardware
modifications (see p. 192). All the topics collected here require special care and/or a
deeper understanding of the respective ‘mechanics’ of the WP 34S – else you might be
surprised by the consequences. Use the information provided in this appendix at your
own risk!

Changing Word Size in Integer Modes

Increasing or reducing the word size in integer modes will affect the current stack contents
only (see the description of WSIZE in the IOP). All other memory content of your WP 34S
will stay as is. This differs from the implementation as known from the HP-16C, where all
registers were remapped on word size changes.

See the following example starting with your WP 34S in BASE 16, WSIZE 16, SSIZE 4,
and with the stack filled with F12E:

I F12E F12E F12E F12E F12E F12E F12E


L old stuff old stuff E61 61 61 61 2E

T F12E F12E F12E 2E 2E 2E 2E


Z F12E F12E F12E 2E 2E 2E 2E
Y F12E F12E F12E 2E 8F 8F 2E
X F12E E61 FF8F 8f 2E 2E 19b2
Input E61 WSIZE 8 WSIZE 16

ATTENTION: Note that increasing the word size will just add empty bits to the
significant side. Thus, a negative integer will immediately become positive then.
There is no automatic sign extension! If you want it, you will have to take care
of it yourself.

Mode Storing and Recalling

The command STOM stores mode data in a register. The following table shows more
details. Types are coded G for general, F for fractions, D for decimal, and I for integer.

Bits Type Contents


0…3 G LCD contrast setting (0 … 7, always 4 for the emulator)
4, 5 F Fractions denominator mode (DENANY = 0, DENFAC = 1, DENFIX = 2)
6 … 19 F DENMAX (14 bits for 0 … 9999)
20 F Clear for PROFRC, set for IMPFRC
21 F Set if fraction mode is on

WP 34S Owner’s Manual Edition 3.1 Page 184 of 211


Bits Type Contents
22, 23 D Decimal display mode (ALL = 0, FIX = 1, SCI = 2, ENG = 3)
24 … 27 D Number of decimals (4 bits for 0 … 11)
28 D Clear for SCIOVR, set for ENGOVR
29 D Clear for RDX. – set for RDX,
30 D Set if E3OFF
31 I Set if SEPOFF
32 I Set if an integer mode is on
33 I Set if LZON
34, 35 I Integer sign mode (1COMPL = 1, 2COMPL = 0, UNSIGN = 2, SIGNMT = 3)
36 … 39 I Integer base (4 bits for 2 … 16 coded as 1 … 15)
40 … 45 I WSIZE (6 bits for 1 … 64 coded as 0 … 63)
46 D Set if DBLON
47 D Time format (24h = 0, 12h = 1)
48, 49 G Print mode (0 … 3, see ⎙MODE)
50 Not used yet
51 G Stack size (SSIZE4 = 0, SSIZE8 = 1)
52, 53 D Date format (Y.MD = 1, M.DY = 2, D.MY = 3)
54, 55 D Angular mode (DEG = 0 ,RAD = 1, GRAD = 2)
56 … 58 D Curve fit model (LINF = 0, EXPF = 1, POWERF = 2, LOGF = 3, BESTF = 4)
59 G Clear for FAST, set for SLOW (always clear on the emulator)
60 … 62 D Rounding mode (0 … 7, see RM)
63 D Clear for JG1782, set for JG1582

Commands for Advanced Users

There are three ON combinations not documented above:

+ : Tells the system a quartz crystal is installed for an accurate real time clock.
Find more about this hardware modification on p. 192ff. The quartz crystal
is an inevitable prerequisite for the clock being useful in medium to long
range (see DATE, TIME, STOPW, SETDAT, SETTIM), and it is required for
print operations as well. The system asks Crystal? InStALLEd – keep
held down and press a second time to activate the crystal. +
will work only if the crystal was not activated yet.
+ should be only necessary if you have loaded the file calc.bin
or calc_full.bin, while the other files enable the crystal automatically.

WARNING: If the crystal is not installed, the system will hang if


+ is entered or if a firmware is booted requiring the crystal, and
WP 34S Owner’s Manual Edition 3.1 Page 185 of 211
a hard reset (actuating the RESET button) or a battery pull will be
required!
Lack of the IR diode cannot cause a system hang since it cannot be
checked automatically. Note this diode may alternatively be connected
using one of the small USB boards mentioned in Appendix A. Find an
installation guide for such a board (version 2) at the end of this
appendix here.

+ : Toggles debugging mode, echoed by Debug ON or Debug OFF. The small


= annunciator is lit for debug mode set. Note your WP 34S will not shut down
automatically in this mode. + is used in the updating procedure on p.
152 (followed there by + ) but was not explained there.

+ Stands for MySamBa: the system asks SAM-BA? boot – keep held
down and press a second time to clear the GPNVM1 bit and turn the
calculator off. This will work in debugging mode only (see above) to prevent
accidental access to this potentially dangerous feature. It is used in the
updating procedure on p. 152 but was not explained there.

WARNING: After issuing this combination, you can only boot in


MySamBa boot mode! Without the MySamBa software and a proper
cable (as mentioned on p. 147), you are lost!

Double Precision (DP) Calculations and Mode Switching

Your WP 34S starts in SP mode per default, wherein 16 digit precision is reached in all
calculations. Switching between SP and integer modes was discussed on p. 159.
Additionally, you may use your WP 34S in DP mode. Each DP register will contain 16
bytes instead of eight, allowing for 34 digits instead of 16 (see below). Note matrix
commands will not work in DP.
DP allows for more precise calculations. While some computations will reach high
accuracy, we do not warrant 34 digit precision in all calculations in DP mode.
The following figure illustrates what happens in memory in transitions between SP and DP
modes, assuming startup in SP mode with REGS 16. RRA stands for ‘relative register
address’.

WP 34S Owner’s Manual Edition 3.1 Page 186 of 211


Absolute Startup memory allocation After executing DBLON Then after DBLOFF
address
Contents RRA Contents RRA Contents RRA
X+11 k = 1.40E-397 R111 = K 1.40E-397 K = R111
1.40E-397 K = R111
X+10 … … … …
… … … … … … …
X+1 y = -33.8 R101 = Y -33.8 Y = R101
… C = R106
X x = 123.0 R100 = X 123.0 X = R100
X-1 r15 = -43.6 R15 0.0 R15
… B
X-2 r14 = 167.9 R14 0.0 R14
X-3 … R13 0.0 R13
… A
X-4 … R12 0.0 R12
X-5 … R11 0.0 R11
… T
X-6 … R10 0.0 R10
X-7 … R09 0.0 R09
… Z
X-8 … R08 0.0 R08
X-9 … R07 0.0 R07
-33.8 Y = R101
X-10 … R06 0.0 R06
X-11 … R05 0.0 R05
123.0 X = R100
X-12 r04 = -12.9 R04 0.0 R04
X-13 r03 = -1234.89 R03 -1234.89 R03
-1.95E184 R01
X-14 r02 = 5.43E-396 R02 5.43E-396 R02
X-15 r01 = 6.6 R01 6.6 R01
1.03E182 R00
X-16 r00 = 0.54 R00 0.54 R00
X-17

Going from SP to DP mode via DBLON, the contents of the twelve registers X … K are
copied, cutting 48 bytes into the former SP numbered register sector. So the top twelve SP
numbered registers will be lost in such a transition. All other memory contents stay where
and as they were – just each DP RRA covers what were two SP registers before. The
space allocated for summation registers will not change in such transitions.
Starting with the default memory configuration and executing DBLON then will leave you
with 44 DP registers. Executing REGS with an argument >44 in DP is legal, but the sector
of global numbered registers will then cut into the former program sector.

Returning from DP to SP mode, the lettered registers are copied again. Everything else
stays where and as it was, if you used ≤ 44 DP registers – just each SP RRA points to only
one half of a former DP register; and the memory released by the shrinking special
registers allows for adding (or returning) twelve numbered registers on top, each now
containing zero.
With >44 DP registers, the correspondence becomes more complicated – the number of
global registers will not, however, exceed 112.

WP 34S Owner’s Manual Edition 3.1 Page 187 of 211


For the following table, assume startup in BASE 10, WSIZE 32, REGS 16. Now see the
contents of J, K, and the lowest numbered registers, checked by recalling them to X:

J K R00 R01 R02 R03


d d d d d d
Starts with e.g. 3,504 14 54 66 543 126,441
DECM 3.43-395 1.40-397 5.40-397 6.60-397 5.43-396 1.23-393
DBLON80 3.43
-395
1.40
-397
1.03
-HIG
1.95
-HIG
n/a n/a
Recall by sRCL 1.40-397 1.28-23 5.40-397 6.60-397 5.43-396 1.23-393
… and by iRCL 14.00 0.00 54.00 66.00 543.00 126,441.00
-395 -397 -397 -397 -396 -393
DBLOFF 3.43 1.40 5.40 6.60 5.43 1.23
Recall by dRCL Out of range Out of range
n/a n/a
Error Error 0.00 0.00
DBLON 3.43-395 1.40-397 1.03-HIG 1.95-HIG n/a n/a
RCL J, STO J,
123E456 STO 00,
then recall by sRCL 1.40-397 1.28-23 1.23-396 5.12-30 5.43-396 1.23-393
… and by iRCL 14.00 0.00 123.00 0.00 543.00 126,441.00
-395 -397 -396 -30 -396
DBLOFF 3.43 1.40 1.23 5.12 5.43 1.23-393
Recall by dRCL Out of range Out of range +∞
0.00 n/a n/a
Error Error Error

Note iRCL and sRCL keep working as explained on p. 159.

In DP mode, shows the first (most significant) 16 digits of the 34-digit mantissa of x
and its four digit exponent, and displays the 18 trailing digits, both as temporary
messages. For example, returns here

, and .

80
DP mode reals are stored coarsely following decimal128 packed coding, though with some exceptions.
The lowest 110 bits take the rightmost 33 digits of the significand. Going left, a 12 bit exponent field follows,
then 5 bits used and coded exactly as in SP, and finally the sign bit. The greatest value of the stored
exponent is 10 1111 1111 11112 = 1228710 . For reasons like the ones explained for SP, 6176 must be
subtracted from this value to get the true exponent of the floating point number represented. Thus, DP
supports 34-digit numbers within . Coding works in full analogy to the way
described for SP in App. B.
Even smaller numbers may be entered using a decimal mantissa, but you will lose one digit per factor of 10.
–6143 –6176
The same happens if you divide 10 by 10 several times. At 10 , only one digit will be left, stored as
–6176
hexadecimal 1. Divide it by 1.999 999 999 99 and the result will remain 10 . Divide it by 2 instead and
the result will become zero.
Numbers beyond the interval will be displayed with -HIG or HIG in their
exponent, respectively. Only will show you the true exponent of them. This way
r00 = 1.032 000 000 000 000 054 E-6155 and r01 = 1.951 656 000 000 000 000 000 543 E-6152 are
found here.
Returning to SP with numbers in lettered registers exceeding the SP number range will cause 0 or
Infinity being displayed instead.
WP 34S Owner’s Manual Edition 3.1 Page 188 of 211
Remember not every return may be as precise as this one. And errors accumulate as
explained in the footnote at CONST.

ATTENTION: Rounding mode (see RM) may affect the results of DP calculations!

Further Commands Used in Library and XROM Routines

The operations listed below are used by the programmers of said routines. They ease their
work by allowing some more comfort in writing programs than the original function set.
These commands and pseudo-commands are explained here to foster understanding of
those routines, but they are not accessible through any catalog. The assembler (often
already its preprocessor) will translate most of them into proper program steps employing
the commands documented above. See the WP 34s Assembler Tool Suite User Guide for
additional information.

(Pseudo-) XROM Function


Command or LIB

A..D XROM saves a, b, c, and d at a volatile temporary storage. Volatile means it


will vanish some 500ms after last keystroke.

GSB longlabel Both This pseudo-command allows for


calling long alphabetic labels. The ATTENTION: Note that
assembler translates GSB into STOP and PROMPT will
BSRB or BSRF (see p. 191) – or not work in XROM on your
into XEQ (via a jump table in XROM WP 34S. And there are no
or via a label in LIB) if the distance LBL nor END statements in
is too far. XROM code.
JMP longlabel Both This pseudo-command allows for Long alphabetic labels
branching to long alphabetic labels. (looking like longlabel:: )
The assembler translates JMP into may show up in code
BACK or SKIP – or GTO (via a jump written for the assembler
table in XROM or via a label in LIB) but will be resolved.
if the distance is too far.

Num sn XROM with sn = 0, π, √2π, … inserts the respective constant like recalling
it from CONST would do.

POPUSR XROM See XEQ or XEQUSR.

XEQ label XROM calls the user routine carrying the global label specified (there are no
LBL statements in XROM).
XEQ’xyz’ in XROM must be followed by POPUSR immediately to
restore the XROM execution state (registers, flags, return addresses)
correctly. See also xIN.

WP 34S Owner’s Manual Edition 3.1 Page 189 of 211


(Pseudo-) XROM Function
Command or LIB

XEQUSR label XROM calls the user routine containing the function to be solved, integrated,
summed, or multiplied, respectively. The label of this function is
transmitted as a parameter of the respective user command
described above (such as e.g. SLV).
XEQUSR must be followed by POPUSR immediately to restore the
XROM execution state (registers, flags, return addresses) correctly.

xIN type XROM with type = NILADIC, MONADIC, DYADIC, TRIADIC,


or ..._COMPLEX defines how many stack levels are used for
parameter input to the function under consideration. Furthermore it
does some initialization work (e.g. SSIZE8 and DBLON).
xIN is the recommended way to start an XROM routine. Thereafter,
SSIZE4 is legal but DBLOFF is not. Note xIN cannot nest and XROM
routines using xIN cannot call user code.

XLBL”label” XROM defines the eXternal LaBeL of this routine. XLBL doesn't generate
any code, it provides an entry point that the C function tables can take
advantage of. Typically, label equals the respective command name.

xOUT way XROM typically, way = xOUT_NORMAL . xOUT cleans and reverts the
settings of xIN, taking care of a proper return including the correct
setting of l and the stack. Generally, xOUT shall be the last command
of an XROM routine.

_INT n XROM 1, 2, …, 128, but also calculated constants are possible. _INT inserts
the respective constant like # does.

“text” Both The double quotes allow for convenient entry of text strings. The
assembler translates the string into the amount of α-statements
required.

A..D XROM recalls a, b, c, and d from their temporary storage (very short range,
see above).

Furthermore, there are several ‘alias’ spellings to ease input of some commands or even
of individual nonstandard characters on an English computer keyboard. E.g. the Greek
letter ‘α’ is often replaced by a Latin ‘a’, the print character ⎙ by ‘P.’ etc. Again, the
assembler will take care of translating the aliases.
See the file 8queens_alias.wp34s (or other files of type …_alias.wp34s in ‘library’)
for examples. Command-Aliases.pdf in ‘doc’ contains a list of all aliases used (some
50 pages).

WP 34S Owner’s Manual Edition 3.1 Page 190 of 211


Assembler Output

The Assembler takes your .wp34s text file as input and generates optimized code for your
WP 34S. It does not generate optimized code for your further editing nor maintaining your
routines. “Optimized” means it will use as little of the limited resources of your WP 34S in
the programs as possible, less labels in particular. Thus expect heavy use of BACK and
SKIP instead of GTO in assembled programs since all required information for this
conversion is available at assembly time.

Furthermore, there are two special commands frequently replacing GSB / XEQ in
assembled code:
1. BSRB n calls a subroutine starting n steps backwards with 0 ≤ n ≤ 255. It pushes
the program counter on the subroutine return stack and executes BACK n then.
2. BSRF n calls a subroutine starting n steps forwards. It pushes the program counter
on the subroutine return stack and executes SKIP n then.

The subroutines called this way do not require a starting label. So, BSRB and BSRF are
useful if you are short on local labels – as BACK, SKIP, and CASE are. On the other hand,
if you edit a section of your routine that is crossed by one or more BSRB, BSRF, BACK,
SKIP, or CASE jumps, this may well result in a need to manually maintain all those
statements individually. We recommend editing unassembled code and leaving the
generation of label-less branches to the assembler for sake of readability, maintainability,
and reliability of your routines.

Basic Electrical Hardware Specifications of the HP-20b and HP-30b

Power supply: 3V (2 CR2032 coin cells)

Processor: Atmel AT91SAM7L128


– ARM 7 core with thumb mode (32kHz to 40MHz)
– RAM: 2kB with backup, 4kB volatile
– ROM: 128kB flash, SAM-BA boot loader
– I/O: parallel ports, serial port for flashing (see picture below for pinout)
– LCD controller
– Elaborate power management
– Multiple clock sources (RC, quartz, PLL), real time clock

Liquid crystal display with 400 segments:


– 12 x 9 segments (mantissa), 3 x 7 segments (exponent), 2 signs,
– 6 x 43 pixels in a dot matrix,
– 10 annunciators, and
– 1 additional fixed symbol (see p. 35).

Keyboard with 37 keys in 5 to 6 columns by 7 rows

WP 34S Owner’s Manual Edition 3.1 Page 191 of 211


The six pad serial port of the calculator is located
under the battery door, below the RESET hole. For
those who want to connect to it directly, here is the
pinout. It is identical for both business calculator
models, the HP-20b and the HP-30b. The pitch is
2mm. The picture is of the printed circuit board un-
covered, the RESET button is north of the six pads.
There is an additional JTAG connector at the
southern end of the main board (not accessible
from outside), partially covered by a tag here.

Tuning the Hardware of Your WP 34S

As mentioned at the very beginning of this manual, there are three optional hardware
modifications you may benefit from. Please read the respective descriptions and decide
what you are interested in. Then read the complete instructions for these cases before
actually starting the modification – this is to prevent you will end with your WP 34S half
usable. These are the opportunities:

1. Create an accurate real-time clock by adding a quartz crystal and two capacitors on
the main board of your WP 34S.
This modification requires some fine soldering skills due to the size of the
capacitors – they are SMD components and significantly smaller than a grain of rice!
Besides fixing these two properly at the correct location, this modification is an easy
job. See Alexander Oestert’s very nice instruction file at
wp34s.svn.sourceforge.net/viewvc/wp34s/doc/How_to_install_crystal_and_IR_diode.pdf .

2. Establish an IR line for printing on an HP-82240A/B by adding an IR diode and a


resistor to the package after modification 1. Electrically, this part is significantly
easier than the job above. There is some additional mechanical work: drilling a little
hole in the plastic case, requiring a slow hand and a correct drilling bit. The
complete job is also described in the file mentioned above.

3. You may combine the setup of both the IR line to the printer and an USB
connection for data exchange and power supply at once by installing a little custom
board Harald Pott developed 81 (it was already mentioned in Appendix A).
Here you will need some sol-
dering skills as well (not as
fine as for modification 1
above, but close). And
mechanical skills are required,
too, since an extra gap in the
calculator case for the micro
81
Harald developed more than this custom board for your WP 34S. We chose what he calls PCB version 2.
We rate it most helpful since it combines the features mentioned above and is easy to install as described
below. See also http://wp34s.svn.sourceforge.net/viewvc/wp34s/doc/WP_34S_USB_installation.pdf for
information about his other boards and Harald’s address.
WP 34S Owner’s Manual Edition 3.1 Page 192 of 211
USB socket is needed in addition to the little hole for the IR diode.

In the following, the job is described for the board as pictured above and an HP-30b,
based on material provided by Alexander and Harald:

a. Open your calculator (the easy part).


Tools needed: Small Philips screwdriver, wooden toothpick.
Procedure:
Remove the battery cover by sliding it open.
Remove the two coin cells using the toothpick.
Remove the three screws you see. Two more are hidden under each end of the
bottom rubber foot – remove these screws, too.

b. Open your calculator (the hard part).


Tool needed: Swiss Army knife.
Helpful information: The shiny front part of the HP-30b is clipped into the grey frame.
There are three clips at either side, two at the top, one at the bottom. See the picture
of an opened frame here:

Procedure:
Carefully insert the big blade of the knife between front part and frame near the
center of one side. Force the shiny part up.
Repeat on the opposite side.
Move the blade toward the calculator top. Gently force the shiny part up on both
sides. Think of the LCD, so be careful here!
Insert the blade between front part and frame at calculator top. Gently force the
shiny part up. The top of the shiny part should be loose now.
Finally, get the bottom part loose the same way.
Turn the shiny part over (printed circuit board up) and lay it on your desk. Put the
grey frame aside.

WP 34S Owner’s Manual Edition 3.1 Page 193 of 211


c. Prepare the mount for the USB board.
Tool needed: Small knife with a sharp thin blade.
Procedure:
Cut away the plastic supports
of the top right screw post
and of the LCD frame in
that area. The bottom pic-
ture shows how it shall
look when you are done.

d. Mount the USB board.


Material needed: Arbitrary
plastic glue, e.g. UHU.
Procedure:
Fix the board around the
screw post, left to “h.pott”,
and below the terminal end. Then lay the assembly aside for drying.

e. Make the opening for the


USB plug.
Tool needed: Small square file.
Procedure:
Take the grey frame and lay
it on the desk bottom up,
rubber foot pointing to you.
Then you need to extend
the opening top right.
Take measures looking at the
mounted USB board. The
pictures show how the
frame shall look when you
are done. Remember the
battery door covers the
upper part.

WP 34S Owner’s Manual Edition 3.1 Page 194 of 211


f. Connect the USB board to the IR diode.
Tool needed: Soldering iron (tip shall not significantly exceed the size of the smallest
terminal you want to solder, < 30W recommended).
Material needed: Solder.
Procedure:
Heat up the soldering iron.
Look for a decent place where the diode may peek through the grey frame. Stay
away from the clips.
Turn the diode so its longer leg will be at right. Bend that leg so it will reach terminal
1 (see the picture). Bend the
shorter leg so it will reach
terminal 2. Cut the legs at
proper length. Both legs must
not contact each other.
Apply a tiny bit of solder on
terminal 1. Do the same for
terminal 2 and for the ends of
both legs of the diode.
Solder the formerly longer diode leg to terminal 1. Solder the other one to terminal 2.

g. Connect the USB board to the main board.


Tool needed: Soldering iron as for step f.
Material needed: Isolated wire (diameter about 0.3mm to be easily laid out), solder.
Procedure:
There are six connections between both boards and one more on the main board.

WP 34S Owner’s Manual Edition 3.1 Page 195 of 211


First, however, two components must be removed from the main board:
Remove LB1. Since you are going to discard it anyway, put the tip of the heated
soldering iron with a bit of solder on both terminals of component LB1 and heat it
up. After a while, the solder below LB1 shall melt, so LB1 can be moved easily. It
may even stick to the tip of the soldering iron and may be lifted this way.
Remove LB2 the same way.
Connect the terminals as listed, i.e. number
3 (IR printing) to upper terminal of R18,
4 (Bat+) to the big upper battery terminal,
5 (Vcc) to the right terminal of LB1,
6 (GND) to the top end of C21,
7 (RxD) to either end of JP5,82
8 (TxD) to either end of JP6,
and connect both big upper battery terminals.
For each connection, cut a piece of wire (some 150mm). Skin its first 2mm, put
some solder on it, put some solder on the respective terminals on either board,
and solder one end to the terminal on the USB board. Then determine the
necessary length for an easy connection, cut the wire to proper length, skin its
last 2mm, and put some solder on it. Solder this end to the terminal on the main
board. Finally, check the proper connection using an Ohmmeter (test voltage =
some 1.5V).83

h. Optional: Install the additional components for an accurate real-time clock.


Tools needed: Soldering iron as for step f, wooden toothpick for holding components.
Material needed: Two 18pF 0603 capacitors, one 32.768kHz quartz crystal, solder.
Procedure:
Check top right of the main board. If the
quartz crystal Y1 and the two capacitors
C3 and C4 next to it are not installed yet,
you can do that now. It is not mandatory,
but will help in communication (cf. p.
166). It is inevitably required, however,
for using the full set of print, time, and
date commands including the stopwatch.
See the picture here for comparison and
the file mentioned at point 1 above for
parts specifications and work instructions.
Soldering these capacitors is a challenge for the common man since they are really
(really!) small. Alternatively, you may use wired capacitors: the result does not look
half as professional but it is far easier achieved and does the job as well.
Remember to activate the crystal when the tuning is completed (see p. 185).

82
JP5 and JP6 are jumpers, i.e. 0Ω resistors.
83
Since the electronic components are small and densely packed on these boards, there is a fair chance of
producing solder bridges – thoroughly check and remove where you produced them. Also take care that
you do not heat any mounted component to an extent where it will start moving away. Soldering on the
main board requires a slow hand, good eyesight, and a bit of experience.
WP 34S Owner’s Manual Edition 3.1 Page 196 of 211
i. Drill the hole for the IR diode.
Tool needed: Drill bit matching the size of the diode, but also any other tool making
decent round holes of suitable size may do (even a Swiss Army knife), if you
proceed slowly.
Procedure:
Hold the grey frame over the mounted diode to learn the location where to drill the
hole. Its center shall be about 5mm away from the edge of the frame. Drill the
hole there.

j. Reassemble.
Tool needed: small Philips screwdriver.
Procedure:
Check your shiny setup will properly fit. Then insert it in the grey frame with the
diode pointing in its hole. Press to click into place (best do top and bottom first –
do not press on the LCD window nor apply exaggerated force!).
Turn in the five screws.
Return the coin cells.
Close the battery door – your calculator is done!

k. Connect your calculator to a USB port of your computer.


Material needed: micro USB cable.
Procedure:
Plug in the cable in your calculator. Depending on the shape of the plug, a little
cutting may be required on the plastic part of the plug to allow for proper insertion.
Turn on your calculator.
Connect it to your computer. If you did everything right, the operating system of your
PC will automatically start searching for a driver on connection. Most probably it
will not find any. Get the newest drivers for your operating system from FTDI for
free – look here: http://www.ftdichip.com/Drivers/VCP.htm . After driver installa-
tion, your calculator will show up in the Device Manager as a ‘FT230X USB Half
UART’ when connected, and a COM port will be allocated. You will need this
COM port for flashing.
Now you are prepared for easy flashing and updating as described in Appendix A.

WP 34S Owner’s Manual Edition 3.1 Page 197 of 211


APPENDIX I: ADVANCED MATHEMATICAL FUNCTIONS

Your WP 34S contains several operations covering advanced mathematics. They are all
implemented for the first time on an RPN calculator; and all work in DECM. Find those
functions collected here and described in more detail than in the IOP, together with a few
traditional pocket calculator functions matching the topic.

For reasons explained in the first section of this manual, we assume you are able to read
and understand mathematical formulas for real domain functions. Wherever complex
numbers may be valid input or output, be sure you understand the respective basic
mathematical concepts (at minimum as outlined on p. 29). Else leave these functions
aside. By experience, it is not beneficial to use something you do neither overview nor
know the background of – it may even become dangerous for you and your fellow men.

Numbers

The following are all one-number functions unless specified otherwise.

Name Remarks (see p. 74 for general information)

n 1
Bn Bn returns the Bernoulli number for an integer n > 0 given in X: Bn 1 n 1 n
.

* 2 2n !
Bn* works with the old definition instead: Bn 2n
2n .
Bn* 2
See p. 208 for ζ(x).

y y!
COMB (2) The number of combinations is C y , x .
x x! y x !

In integer modes, returns the Fibonacci number fn with n = x . These numbers are
defined as f0 = 0 , f1 = 1 , and fn = fn–1 + fn–2 for n ≥ 2 . With UNSIGN, f93 is the
maximum before an overflow occurs.

In DECM, returns the extended Fibonacci number


FIB
x x
1 1 5 2 1
Fx cos x x x
cos x
5 2 1 5 5

for an arbitrary real number x, with Φ denoting the golden ratio.

y!
PERM (2) The number of permutations is Py , x x! C y , x .
y x!

WP 34S Owner’s Manual Edition 3.1 Page 198 of 211


Statistical Distributions

Stack-wise, the following are all one-number functions. They are all stored in PROB. In the
table below, the three discrete distributions are covered first, the continuous ones
thereafter. Typical plots are shown for the pmfs or the pdfs, respectively.

Name Remarks (see p. 74 for general information)

Binomial distribution with the number of


successes g in X, the gross probability
of a success p0 in J and the sample size
n in K.

BinomP returns
n n g
p B ( g ; n; p0 ) p0g 1 p0
g
n g
Cn, g p0g 1 p0
(see COMB above).
Binom Binom returns
m
84
FB (m; n; p0 ) p B ( g ; n; p0 ) with the maximum number of successes m in X.
g 0

The binomial distribution is fundamental for error statistics in industrial sampling, e.g. for
designing test plans. If you want to know, for example, the probability for finding no
faulty items in a sample of fifteen items drawn from a batch of 300 wherein you expect
3% defective items overall, this will tell you:
0.03 15 0 Binom returns 0.633 – so the odds are almost two
out of three that you will not detect any defect in your sample! 85
Read here for more information:
http://www.itl.nist.gov/div898/handbook/eda/section3/eda366i.htm .

Geometric distribution:
n
GeomP returns f Ge n p0 1 p0 ,
m 1
Geom returns FGe m 1 1 p0 ,
being the probability for a first success
after m = x Bernoulli experiments. The
Geom
probability p0 for a success in each such
experiment must be specified in J.
Start reading here for more:
http://en.wikipedia.org/wiki/Geometric_dist
ribution .

84
BinomP equals BINOMDIST(g; n; p0; 0) and Binom equals BINOMDIST(m; n; p0; 1) in MS Excel.
85
The exact result for said probability is 0.626, calculated using a mathematically more elaborated model of
such a test. These results show nicely that two significant digits are a typical accuracy of statistical
statements – frequently the (simplified) statistical model used matches reality no better than that.
WP 34S Owner’s Manual Edition 3.1 Page 199 of 211
Name Remarks (see p. 74 for general information)

Poisson distribution with the number of


successes g in X and either the Poisson
parameter λ in J (for Poisλ) or the gross
probability of a success p0 in J and the
sample size n in K. (for Poiss). In the
latter case, n p0 is calculated auto-
matically.
Poiss, PoisλP 86
computes
g
e ,
PP ( g ; )
Poisλ g!
and Poisλ returns the cdf for the
maximum number of successes m in X.
The Poisson distribution provides the
mathematically easiest statistical model for industrial sampling tests. Repeating the
example we made with Binom above, we get 0.638 here.
Read here for more information:
http://www.itl.nist.gov/div898/handbook/eda/section3/eda366j.htm .

Cauchy-Lorentz distribution (also known


as Lorentz or Breit-Wigner distribution)
with the location x0 specified in J and the
shape in K :

CauchP returns
1
2
x x0
f Ca x 1 ,

Cauch
Cauch returns
1 1 x x0
FCa x arctan ,
2

1
Cauch-1 returns FCa1 p x0 tan p .
2
This distribution is quite popular in physics. It is a special case of Student’s t-distribution.
Start reading here for more: http://en.wikipedia.org/wiki/Cauchy_distribution .

86
PoisλP corresponds to POISSON(g; ; 0) and Poisλ to POISSON(g; ; 1) in MS Excel.
WP 34S Owner’s Manual Edition 3.1 Page 200 of 211
Name Remarks (see p. 74 for general information)

Exponential distribution with the rate in


J.
x
ExponP87 returns f Ex x e .
See some curves plotted here.

Expon Expon returns FEx x 1 e x


.

Read here for more information:


http://www.itl.nist.gov/div898/handbook/ed
a/section3/eda3667.htm .

Fisher’s F-distribution with the degrees of


freedom in J and K. It is used e.g. for the
analysis of variance (ANOVA).

The picture shows the pdf plotted for


different degrees of freedom m and n
F(x)
corresponding to j and k.

Read here for more information:


http://www.itl.nist.gov/div898/handbook/ed
a/section3/eda3665.htm .

Log-normal distribution with ln x g


specified in J and ln in K.

LgNrmP returns
2
ln x
1 2 2
f Ln x e .
x 2
Some curves are plotted here.
LgNrm
ln x
LgNrm returns FLn x

with Φ(z) denoting the standard


normal cdf as presented at the
bottom of this table.

The picture shows plots of log-normal pdfs


with the parameters specified. Read here for more information:
http://www.itl.nist.gov/div898/handbook/eda/section3/eda3669.htm .

87
The pdf corresponds to EXPONDIST(x; ; 0) and the cdf to EXPONDIST(x; ; 1) in MS Excel.
WP 34S Owner’s Manual Edition 3.1 Page 201 of 211
Name Remarks (see p. 74 for general information)

Logistic distribution with an arbitrary mean


μ given in J and a scale parameter s in K.
Substituting ,

e
LogisP returns f Lg x 2
s 1 e
Logis
1
and Logis returns FLg x .
1 e
p
Logis-1 returns FLg1 p s ln .
1 p
Start reading here for more:
http://en.wikipedia.org/wiki/Logistic_distrib
ution .

Normal (or Gaussian) distribution


with an arbitrary mean µ given in
J and an arbitrary standard
deviation σ in K.
NormlP88 returns
2
x
1 2 2
fN x e and
Norml 2
x
Norml returns FN x

with Φ(z) denoting the


standard normal cdf as presented at the bottom of this table.
Read here for more information:
http://www.itl.nist.gov/div898/handbook/eda/section3/eda3661.htm .

Standardized Student’s t-distribution


with its degrees of freedom in J as
used for hypothesis testing and
calculating confidence intervals e.g. for
means. The picture shows the pdf
plotted for different degrees of
freedom. For df → ∞ , the shoulders of
t(x) the function shrink and the pdf
approaches the one of the standard
normal distribution (compare the red
curve at NORML).
Read here for more information:
http://www.itl.nist.gov/div898/handbook/
eda/section3/eda3664.htm.

88 -1
NormlP corresponds to NORMDIST(x; µ; σ; 0) in MS Excel, Norml to NORMDIST(x; µ; σ; 1) and Norml
to NORMINV(FN; µ; σ) .
WP 34S Owner’s Manual Edition 3.1 Page 202 of 211
Name Remarks (see p. 74 for general information)
T = 1 , b = 0.5
Weibull distribution with its shape T = 1 , b = 1
T = 1 , b = 1.5
parameter b in J and its characteristic T = 1 , b = 5
lifetime T in K.
WeiblP89 returns
b 1 t b
b t
fW (t ) e T
for t ≥ 0 (else it
T T
Weibl returns 0). This is a very flexible function
– see the curves plotted here.
t b

Weibl returns FW (t ) 1 e T
.

Read here for more information:


http://www.itl.nist.gov/div898/handbook/e
da/section3/eda3668.htm . Also see
http://en.wikipedia.org/wiki/Weibull_distribution#Uses for some more application fields.

1 x2
2
x e is the standard normal pdf (the famous bell curve, see the red curve at
2
φ(x), x
Φ(x) Norml), while x d is the corresponding cdf (compare the error function as

shown on p. 208). See the link at NORML above for more information.

The chisquare distribution is used for


calculating confidence intervals for
standard deviations and the like. The
2
picture shows the pdf plotted for
different degrees of freedom.
Read here for more information:
http://www.itl.nist.gov/div898/handbook
/eda/section3/eda3666.htm .

89
The pdf equals WEIBULL(x; b; T; 0) and the cdf WEIBULL(x; b; T; 1) in MS Excel.
WP 34S Owner’s Manual Edition 3.1 Page 203 of 211
More Statistical Formulas

Note that complete results of measured samples must include both an information about
the expected value and about its uncertainty.
For samples drawn out of a Gaussian (additive) process, the expected value is the
arithmetic mean and its uncertainty is given by its standard error (see x and SERR).
For samples drawn out of a log-normal (multiplicative) process, the expected value is
the geometric mean and its uncertainty is given by its scattering factor (see xg , m).
For samples drawn out of other processes other measures apply.

Be assured not everything is Gaussian in real world! 90 Process features can be detected
(and should be checked well in advance of calculating e.g. means) using suitable tests –
turn to applicable statistical reference literature.

The following functions as named in the left column are all found in STAT (except x and s,
which are added to this table for comparison only).

Name Remarks (see p. 74 for general information)

s xy
(1) For a linear fit model, the correlation coefficient is r . See sXY and s below.
sx s y
2
2
R xi yi
For an arbitrary fit model R(x) , the value r 1 2
is the coefficient of
y yi
determination; it indicates the fraction of the variation of the dependent data y deter-
CORR
mined by the variation of the independent data x . For r 2 1 , y is fully determined by
x ; for r 2 0 , y is completely independent of x ; and for e.g. r 2 0.93 , 93% of the
total variation of y is due to x .

n 2
A regression is significant if r t n 12 0.99 , with the right side being the
1 r2
inverse of the t-distribution.

(1) For a linear fit model, the population covariance is


COV COVxy n xi y i xi yi n 2 . Compare sXY below.

90
Generally, the statistical model shall be chosen that matches observations best. In many real life cases,
however, dramatic deviations from the model distribution are found – then you cannot expect the calculated
consequences matching the reality any better.
By the way: Since the pdf of the Gaussian distribution will never reach zero, this statistical model tells you
to expect individual items far, far away from the mean value when your sample becomes large enough.
This, however, does not match reality. So we must conclude nothing at all is really Gaussian in real world.
Nevertheless, the Gaussian distribution is a very successful model for describing a lot of real world
observations. Just never forget the limits of such models.
WP 34S Owner’s Manual Edition 3.1 Page 204 of 211
Name Remarks (see p. 74 for general information)

xi2 yi xi xi y i
(2) For a linear fit model, the line parameters are a0
n n 1 s x2
s xy sy
and a1 r (see CORR above, sXY and s below). Their standard errors can
s x2 sx
L.R. sy 1 r 2 n 1 2
be calculated using s E (a1 ) and s E (a0 ) s E (a1 ) sx x2 .
sx n 2 n

ai
Generally, a regression parameter is significant if t n 12 0.995 , with the right
s v ai
side being the inverse of the t-distribution.

(1) For a linear fit model, the sample covariance is


sXY
s xy n xi y i xi yi n (n 1) . Compare COV above.

2
2
n xi2 xi xi2 n x 2
(2) The sample variance is s x .
n (n 1) n 1
s,
SERR The sample standard deviation (SD) is s x s x2 .
sx
And the standard error (i.e. the SD of the mean x ) is s Ex .
n

(1) The sample SD for weighted data (where the weight yi of each data point xi was
2
yi yi xi2 y i xi
entered via ) is s w .
yi 1 yi
sW ,
SERRW And the respective standard error (the SD of the mean xw ) is
2
1 yi yi xi2 y i xi
s Ew .
yi yi 1

1
x (2) The arithmetic mean (or average) is calculated as x xi .
n
1
ln x
xg (2) The geometric mean is calculated as xg n x e n
.

xi y i
xw (1) The arithmetic mean for weighted data (see sW) is calculated as x w .
yi

(2) The scattering factor εx for a sample of log-normally distributed data is calculated via:

ln 2 xi 2n ln x g
ln( x ) . Compare s.
n 1

WP 34S Owner’s Manual Edition 3.1 Page 205 of 211


Name Remarks (see p. 74 for general information)

1
m (1) The scattering factor of the geometric mean is m
n
. Compare SERR.

(2) The scattering factor εp for a population of log-normally distributed data is calculated
via:
p n 1
ln( p ) ln( x ) . Compare .
n

1 2 n 1
(2) The SD of the population is x xi x sx .
n n

2
y i xi xw
W (1) The SD of the population for weighted data (see sW) is w .
yi

WP 34S Owner’s Manual Edition 3.1 Page 206 of 211


Orthogonal Polynomials

These polynomials are all found in X.FCN.

Name Remarks (see p. 74 for general information)

n x2 dn x2
(2) Hermite polynomials for probability: Hn x 1 e 2
e 2
with n in Y,
Hn
dx n
solving the differential equation
f" x 2x f ' x 2n f x 0.

n x2 dn x2
Hnp (2) Hermite polynomials for physics: H np x 1 e e with n in Y, solving
dx n
the same differential equation.

ex d n
(2) Laguerre polynomials (compare Ln below): Ln x x ne x
L(n0) x with n
n! dx n
Ln in Y, solving the differential equation
x f" x 1 x f' x n f x 0.

(3) Laguerre’s generalized polynomials (compare Ln above):


Ln x ex d n n
L(n ) x x e x
with n in Y and in Z.
n! dx n

1 dn n
(1) Legendre polynomials: Pn x x2 1 with n in Y, solving the
2 n! dx
n n

Pn differential equation
d d
1 x2 f x nn 1f x 0.
dx dx

(2) Chebychev (a. k. a. Čebyšev, Tschebyschow, Tschebyscheff) polynomials of first kind


Tn Tn(x) with n in Y, solving the differential equation
1 x2 f " x x f' x n2 f x 0.

(2) Chebychev polynomials of second kind Un(x) with n in Y, solving the differential
Un equation
1 x2 f " x 3x f ' x nn 2 f x 0.

WP 34S Owner’s Manual Edition 3.1 Page 207 of 211


Even More Mathematical Functions

Also these are all found in X.FCN. Some of them are for pure mathematics only, but were
useful at some stage of the WP 34S project, so we made them accessible for the public.

Name Remarks (see p. 74 for general information)

(2) Returns the arithmetic-geometric mean. Find more about it here:


AGM
http://mathworld.wolfram.com/Arithmetic-GeometricMean.html .

(1) Returns the error function


x
2 2
erf x e d .
0

erf
x
Note that erf 2 x 1 with
2
Φ(x) representing the standard normal
cdf as described on p. 203.

2 2
(1) Returns the complementary error function erfc x 1 erf ( x) e d . This
erfc
x
function is related to the error probability of the standard normal distribution.

(1) Returns the Gudermann function


x
d
gd x . It links hyperbolic and
0
cosh
trigonometric functions. See the plot for its
gd
real values. The inverse Gudermann
x
1 d
function is g d x . Start reading
0
cos
here for more: http://en.wikipedia.org/wiki/Gudermannian_function .

x, y , z
(3) Returns the regularized (incomplete) beta function x
with (y,z) being
y, z
Iβ x
z 1
Euler’s Beta (see there) and x x, y, z ty 1 1 t dt being the incomplete beta
0
function.

x, y
(2) Returns the regularized (incomplete) gamma function P x, y . For x, y
IΓp x
see γXY below, for x see there.

x, y
(2) Returns the regularized (incomplete) gamma function Q x, y u
. For
IΓq x
u x, y see ΓXY below, for x see there.

WP 34S Owner’s Manual Edition 3.1 Page 208 of 211


Name Remarks (see p. 74 for general information)

(1) Returns Lambert’s W with its principal branch


(called W p here) and its negative branch (called Wm
for minus). The connecting point is (-1/e , -1). The
diagram shows the real values of both branches.
Start reading here for more information:
Wp, http://en.wikipedia.org/wiki/Lambert_W_function .
Learn more here:
Wm http://mathworld.wolfram.com/LambertW-
Function.html .

(2) Returns the lower incomplete gamma function x, y t x 1e t dt . Required for IΓp
γXY
0
above.

(2) Returns the upper incomplete gamma function x, y t x 1e t dt . Required for


ΓXY u
y

IΓq above.

1
(1) Returns Riemann’s Zeta for real arguments, with x x
for x > 1 ,
n 1 n

ζ x 1
and its analytical continuation for x < 1 : x 2x sin x 1 x 1 x .
2
Read here for more: http://mathworld.wolfram.com/RiemannZetaFunction.html .

You will find lots of information about the special functions implemented in your WP 34S in
the internet in addition. Generally, Wikipedia is a good starter – check the articles in
different languages since they may well contain different material and use different
approaches. Mathworld may contain more details than you ever wanted to know. And for
applied statistics, the NIST Sematech online handbook quoted above is a competent
source. Further references are found at these sites.

WP 34S Owner’s Manual Edition 3.1 Page 209 of 211


APPENDIX J: RELEASE NOTES

Date Release notes

1 9.12.08 Start
1.1 15.12.08 Added the table of indicators; added NAND, NOR, XNOR, RCLWS, STOWS, //, N, SERR, SIGMA,
< and >; deleted HR, INPUT, 2 flag commands, and 2 conversions; extended explanations for addressing and
COMPLEX & …; put XOR on the keyboard; corrected errors.

1.2 4.1.09 Added ASRN, CBC?, CBS?, CCB, SCB, FLOAT, MIRROR, SLN, SRN, >BIN, >DEC, >HEX, >OCT, BETA, D>R,
DATE, DDAYS, D.MY, M.DY, Y.MD, CEIL, FLOOR, DSZ, ISZ, D>R, R>D, EMGAM, GSB, LNBETA, LNGAMMA,
MAX, MIN, NOP, REAL, RJ, W and WINV, ZETA, %+ and %-; renamed the top left keys B, C, and D, and bottom
left EXIT.

1.3 17.1.09 Added AIP, ALENG, ARCL, AROT, ASHF, ASTO, ATOX, XTOA, AVIEW, CLA, PROMPT (all taken from 42S),
CAPP, FC?C, FS?C, SGMNT, and the …# commands; renamed NBITS to BITS and STOWS to WSIZE; specified
the bit commands closer; deleted the 4 carry bit operations.

1.4 10.2.09 Added CONST and a table of constants provided, D>J and J>D, LEAP?, %T, RCL and STO and , and 2
forgotten statistics registers; deleted CHS, EMGAM, GSB, REAL and ZETA; purged and renamed the bit
operations; renamed many commands.

1.5 5.3.09 Added RNDINT, CONV and its table, a memory table, the description of XEQ B, C, D to the operation index, and a
and ge to the table of constants; put CLSTK on a key, moved CLΣ and FILL, changed the % and log labels on the
keyboard, put CLALL in X.FCN; checked and cleaned alpha mode keyboard and added a temporary alpha
keyboard; rearranged the alphabet to put Greek after Latin, symbols after Greek consistently; separated the input
and non-programmable commands; cleaned the addressing tables.

1.6 12.8.09 Added BASE, DAYS+, DROP, DROPY, E3OFF, E3ON, FC?F, FC?S, FIB, FS?F, FS?S, GCD, LCM, SETDAT,
SETTIM, SET24, SINC, TIME, VERS, αDAY, αMONTH, αRC#; %Σ, as well as F-, t-, and 2-distributions and their
inverses; reassigned DATE, modified DENMAX, FLOAT, αROT, and αSHIFT; deleted BASE arithmetic, BIN, DEC,
HEX, and OCT; updated the alpha keyboards; added flags in the memory table; included indirect addressing for
comparisons; added a paragraph about the display; updated the table of indicators; corrected errors.

1.7 9.9.09 Added P.FCN and STAT catalogs, 4 more conversions, 3 more flags, Greek character access, CLFLAG, DECOMP,
DENANY, DENFAC, DENFIX, Iβ, IΓ, αDATE, αRL, αRR, αSL, αSR, αTIME, 12h, 24h, fraction mode limits, normal
distribution and its inverse for arbitrary µ and , and Boolean operations working within FLOAT; deleted αROT,
αSHIFT, the timer, and forced radians after inverse hyperbolics; renamed WINV to W –1, and beta and gamma
commands to Greek; added tables of catalog contents; modified label addressing; relabeled PRGM to P/R and
PAUSE to PSE; swapped SHOW and PSE as well as % and % on the keyboard; relabeled Q; corrected CEIL and
FLOOR; updated X.FCN and alpha commands; updated the virtual alpha keyboard.

1.8 29.10.09 Added R-CLR, R-COPY, R-SORT, R-SWAP, RCLM, STOM, alpha catalogs, 1 more constant and some more
conversions, a table of error messages, as well as the binomial, Poisson, geometric, Weibull and exponential
distributions and their inverses; renamed some commands; put √‾ instead of on hotkey D.

1.9 14.12.09 Added two complex comparisons; swapped and changed labels in the top three rows of keys, dropped CLST;
completed function descriptions in the index.

1.10 19.1.10 Added IMPFRC, PROFRC, CENTER, αBEG, αEND, and an addressing table for items in catalogs; updated
temporary alpha mode, display and indicators, RCLM and STOM, alpha-commands and the message table;
renamed the exponential distribution; wrote the introduction.

1.11 21.9.10 Changed keyboard layout to bring Π and Σ to the front, relabeled binary log, swapped the locations of π, CLPR, and
STATUS, as well as SF and FS?; created a menu TEST for the comparisons removed and the other programmable
tests from P.FCN; added %MG, %+MG, %MRR, RESET, SSIZE4, SSIZE8, SSIZE?, CDROP, CFILL, CR , CR ,
registers J and K, a table of contents and tables for stack mechanics and addressing in complex operations;
updated memory and real number addressing tables, DECOMP, αOFF, αON, Π, and Σ; renamed ROUNDI,
WSIZE?, β(x,y), Γ(x) and the constant p0 ; deleted DROPY (use x y, DROP instead), αAPP, αBEG, αEND, and the
“too long error” message; deleted Josephson and von Klitzing constants (they are just the inverses of other
constants already included in CONST); brought more symbols on the alpha keyboard.

1.12 22.12.10 Modified keyboard layout; added catalogs MODE and PROB; changed mode word, catalog contents and handling
(XEQ instead of ENTER), as well as some non-programmable info commands; expanded IMPFRC and PROFRC;
added a paragraph about the fonts provided and explained alpha catalogs in detail; added PRIME? and some
conversions; deleted FRACT, OFF and ON.

1.13 3.2.11 Modified keyboard layout; modified αTIME, radix setting, H.MS+ and H.MS-; added EVEN?, FP?, INT?, LZOFF,
LZON, ODD?, RCLS, STOS, returned FRACT; added and renamed some conversions; updated the paragraph
about display; added appendices A and B; baptized the device WP 34S.

WP 34S Owner’s Manual Edition 3.1 Page 210 of 211


1.14 18.3.11 Started the Windows emulator.
Added DEC and INC, renamed FLOAT to DECM; redefined αTIME and H.MS mode; updated appendix A;
documented the annunciators BEG and = as well as underflows and overflows in H.MS; corrected some errors
showing up with the emulator.

1.15 21.3.11 Modified FIX, removed ALL from MODE, updated CONV.

1.16 27.3.11 Added LBL?, f’(x), and f”(x); modified PSE; upgraded catalog searching.

1.17 9.5.11 Modified keyboard layout for adding a fourth hotkey; added AGM, BATT, Bn, Bn*, Cauch, Lgnrm, Logis and their
inverses, all the pdf, COV, CUBE, CUBERT, DEG , ENGOVR, ENTRY?, erfc, GRAD , GTO . hotkey, KEY?,
RAD , SCIOVR, SERRw, SLVQ, sw, sxy, TICKS, TVM, xg, , m, p, , w, (-1)X, the polynomials, four angular
conversions, four Planck constants, the regional settings, global alpha labels, and three messages; renamed most
cdf; changed DEG, RAD, GRAD to leaving angular mode as set; altered PSE for early termination by key-
stroke; made D.MY default instead of Y.MD; moved degrees to radians conversions to CONV; removed CCLx,
H.MS mode, %+ and %-; corrected errors.

1.18 5.6.11 Expanded program memory; modified label addressing (A ≠ ‘A’) and fraction mode limits, changed ANGLE to work
in real and complex domains, renamed MOD to RMDR, changed the keyboard layout; put BACK, ERR, SKIP, and
SPEC? to the main index; added CAT and the I/O commands for flash memory, expanded R-COPY; corrected
x α.

2.0 21.7.11 Entered beta test phase.


Added DAY, MONTH, YEAR, FAST, SLOW, S.L, S.R, VWα+, flag A, ON + and –, some constants, and a
paragraph about I/O; renamed old DAY to WDAY, RRCL to RCFRG, SRCL to RCFST; added an inverse
conversion shortcut, stones kg, and changed Pa mbar to Pa bar; modified the VIEW commands, ALL, DISP,
MODE, RCLM, STOM, and X.FCN; repaired hyperlinks; corrected some errors; included flash.txt; updated the first
chapters, explained stack mechanics in more detail.

2.1 3.10.11 Added serial I/O commands, DELP, DSL, EXPT, IBASE?, INTM?, ISE, KTY?, MANT, NEXTP, PUTK, REALM?,
RM, RM?, SMODE?, TOP?, x√y, signed tests for zero, some constants, and the paragraph about interactive
programming; updated the values in CONST to CODATA 2010, also updated SLVQ, SHOW, Σ, Π, and the
paragraphs about statistics, predefined alpha labels and memory; corrected some errors; deleted complex ANGLE,
BIN, DEC, HEX, and OCT; redistributed the contents of X.FCN and P.FCN; renamed S.L and S.R to SDL
and SDR; put ‘?’ on the alpha keyboard and moved £ to P to make room for ; expanded Appendix A; reorganized
the structure of the document; added first aid to the front page; rewrote the keyboard chapter.

2.2 1.11.11 Added MSG, y , z , and matrix operations, a paragraph about them and two new error messages for them, plus a
footnote for DELP; updated the introduction to statistics.
This version is the last one working with the old overlays. It is maintained to
incorporate the latest common bug fixes – last v2.2 build available is 2739 so far.
3.0 21.4.12 Added CLPALL, CROSS, DOT, iRCL, sRCL, END, FLASH?, gd, gd–1, GTO. and , LOAD…, LocR, LocR?,
MEM?, NEIGHB, PopLR, RDP, REGS, REGS?, RSD, SEPOFF, SEPON, SETJPN, STOPW, t , ULP, , #, as well
as SUMS and MATRIX catalogs and four conversions; renamed CLFLAG to CFALL, CUBE to x3 and CUBERT to
3
√x, KTY? to KTP?, and αVIEW to VIEWα; split Lambert’s W into W p and Wm; returned BIN, HEX, and OCT;
made PSTO and SAVE nonprogrammable; redefined SHOW, corrected CLREG, deleted DELP; changed keyboard
layout to bring MATRIX, CLP, SF, and CF to the front and to swap OFF and SHOW, removed x α from the key
plate and π from CONST; modified the virtual alpha keyboard, some characters and the respective catalogs;
redistributed commands in the catalogs; updated and rearranged large parts of the text; added information about
complex calculations, bitwise integer operations, browsers, local data, memory management, and the character
sets; implemented five new ttf-fonts.

3.1 30.11.12 Added the print commands ⎙…, graphic commands g…, error probabilities, ½,
IDIV, IΓq, MOD, XTAL?, γXY, ΓXY, ×MOD, ^MOD, four conversions, two messages,
an optional back label, and some references to external documents; modified GCD
and LCM, DATE, SETDAT, SETTIM, and TIME; expanded STOPW; renamed
IBASE? to BASE? and IΓ to IΓp; purged TEST moving non-tests into P.FCN;
abandoned the internal expert’s catalog; updated and rearranged large parts of the
introductory text and the appendices, also explaining the floating point formats,
carry and overflow conditions in integer modes, the emulator, troubleshooting,
corresponding operations of HP-42S and HP-16C, and some hardware
modifications; moved most mathematical explanations into Appendix I; modified
many internal page number references to ease working with a printed manual.

WP 34S Owner’s Manual Edition 3.1 Page 211 of 211

You might also like