Programming in Z80 Assembly Language

Also from Macmillan

Advanced Graphics with the BBC Model B Microcomputer

Ian O. Angell and Brian J. Jones
Advanced Graphics with the Sinclair ZX Spectrum
Ian O. Angell and Brian J . Jones
Assembly Language Programming for the Acorn Electron
Ian Birnbaum
Assembly Language Programming for the BBC Microcomputer,
second edition
Ian Birnbaum
Advanced Programming for the 16 K ZX8l
Mike Costello
Using Your Home Computer
Garth W.P . Davies
Beginning BASIC
Peter Gosling
Continuing BASIC
Peter Gosling
Practical BASIC Programming
Peter Gosling
Program Your Microcomputer in BASIC
Peter Gosling
Codes for Computers and Microprocessors
P. Gosling and Q. Laarhoven
Microprocessors and Microcomputers - their use and programming
Eric Huggins
The Sinclair ZX81 - Programming for Real Applications
Randle Hurley
More Real Applications for the ZX8l and ZX Spectrum
Randle Hurley
Digital Techniques
Noel Morris
Microprocessor and Microcomputer Technology
Noel Morris
The Alien, Numbereater, and Other Programs for Personal
Computers - with notes on how they were written
John Race
Understanding Microprocessors
B.S. Walker
Assembly Language Assembled - for the Sinclair ZX81
Anthony Woods
 Programming in
Z80 Assembly Language

Roger Hutty

M
MACMILLAN
© Roger Hutty 1984

All rights reserved. No part of this

publication may be reproduced or
transmitted, in any form or by any
means, without permission.

First published 1984 by

Higher and Further Education Division
MACMILLAN PUBLISHERS LTD
London and Basingstoke
Companies and representatives
throughout the world

ISBN 978-1-349-06959-0 ISBN 978-1-349-06957-6 (eBook)

DOI 10.1007/978-1-349-06957-6
Contents

Pre f a c e ix

Cha pt er The Z80 a rch i te c tu re

1.1 Mic ro p r o ce s sor sy s t e ms 1

1. 2 Th e Z80 c e ntra l p roc ess i ng uni t 2
1.3 Memo ry 3
1.4 Ins tructions 4
1.5 As se mbly l a ngua ge 4

Chapt er 2 Ac cumulat or and r egis t e r instructions 6

2.1 Load A r e gist er with a v a l ue 6

2. 2 Add a n d subtra ct a va l u e 6
2. 3 Ad d an d subtract a re gist er 7
2 .4 Load one r e g ister wi th a n o t he r 7
2. 5 Incremen t a nd de c rement a re g i ster 7
2 .6 Nega t e the accumu la tor 8
2. 7 Add res si ng mode s - immed i at e and e xt e nd ed 8
2 .8 Labels 9
2.9 Program 10

Cha pt er 3 Subr ou t ines a n d di spla y output 11

3.1 Subr o utine conc e p ts 11

3. 2 Th e CALL a nd RET i ns t r uc tion s 12
3.3 Disp la y ou t p ut 13
3. 4 Ps e ud o o pe rations 14
3.5 Progra m 15

Chap te r 4 Uncondi tiona l jumps a nd k ey board i n pu t 16

4. 1 Unco n d i t i ona l jumps 16

4.2 Keyboard input 17
4.3 Cha r a c t e r code s and values 18
4 .4 The EQU ps eudo o pe r a t o r 19
4 .5 Program 20

Cha pt er 5 Fla gs, con d i t iona l jump s and the CP instruct i on 21

5. 1 Th e f l ag r e g i st er 21
5 .2 Con di ti on al j ump i nstructions 22
5.3 The c o mpare in s t r u c t ion 23
5 .4 Con di t i onal loo p te r mi na tion 24
5. 5 Pro gram 25

v
vi Contents

Chapter 6 Counting loops and the stack 26

6.1 Counting loops 26

6.2 Number input 28
6.3 The stack 29
6.4 The stack instructions 30
6.5 Saving and restoring registers 30
6.6 Program 31

Chapter 7 Nested loops and addressing modes 33

7.1 Nested loops 33

7.2 Immediate extended and register indirect
addressing modes 33
7.3 The DEFM pseudo operator 36
7.4 Text output 37
7.5 The subroutine mechanism 37
7.6 Program 38

Chapter 8 Carry and overflow 40

8.1 Carry 40
8.2 The carry flag 40
8. 3 Over flow 41
8.4 The overflow flag 42
8 .5 Conditional CALLs and RETs 42
8.6 Program 43

Chapter 9 Bit operations and the index registers 45

9.1 The BIT test instruction 45

9.2 The SET and RES instructions 46
9.3 The DEFS pseudo operator 46
9.4 The index registers 47
9.5 Expressions 47
9.6 Jump tables 48
9.7 Program 50

Chapter 10 Shift instructions, multiply and divide 51

10.1 The SRL instruction 51

10.2 The SRA instruction 52
10.3 The SLA instruction 52
10.4 8-bit multiplication and division 53
10.5 Program 55

Chapter 11 Logical operations and macros 56

11.1 Logic operators 56

11.2 Logical i ns t r uc t i ons 56
11.3 Masking 57
11.4 Macros 57
11.5 Conditional pseudo operations 60
11.6 Program 62
 Contents VII

Chapter 12 Rotate instructions and parity 63

12.1 Accumulator rotate instructions 63

12.2 Register and memory byte rotate instructions 65
12.3 Packing and unpacking 65
12.4 Parity 67
12.5 The parity flag 67
12.6 Program 68

Chapter 13 16-bit and multiple byte arithmetic 69

13.1 The DEFW pseudo operator 69

13.2 The 16-bit ADD, ADC and SBC instructions 70
13.3 Extended loops 72
13.4 Multiple byte arithmetic 74
13.5 Program 75

Chapter 14 Block transfer and search instructions 76

14.1 Block transfer instructions 76

14.2 Block search instructions 79
14.3 Program 81

Chapter 15 Decimal arithmetic 83

15.1 BCD representation 83

15.2 BCD arithmetic 84
15.3 The DAA instruction 86
15.4 The digit rotate instructions 87
15.5 Program 89

Chapter 16 Miscellaneous instructions 90

16.1 The NOP instruction 90

16.2 The auxiliary registers 90
16.3 Input and output instructions 91
16.4 Interrupt instructions 91

Appendix A Binary and hexadecimal number systems 93

A.l Binary and hexadecimal numbers 93

A.2 Binary and hexadecimal arithmetic 95
A.3 Decimal to hexadecimal conversion 95
A.4 Hexadecimal to decimal conversion 95
A.5 Binary-hexadecimal conversion 96
A.6 Decimal-binary conversions 97
A.7 Bytes 97
A.8 Signed (2's complement) numbers 97
Exercise answers 99

Appendix B Hexadecimal-decimal conversion tables 101

viii Contents

Appendix C Summary of 280 instructions 103

Table C.lSummary of flag operations 104

Table C.28-bit load group 105
Table C.316-bit load group 106
Table C.4Exchange group and block transfer 107
and search group
Table C.S 8-bit arithmetic and logical group 108
Table C.6 General purpose arithmetic and
CPU control groups 109
Table C.7 16-bit arithmetic group 110
Table C.8 Rotate and shift group 111
Table C.9 Bit set, reset and test group 112
Table C.IO Jump group 113
Table C.11 Call and return group 114
Table C.12 Input and output group 115

Appendix D Display and keyboard character codes 116

Appendix E Expression operators 117

Index 118
Preface

Many microcompute r systems, home and business o ne s , are based on

Zilog's Z80 microprocessor, one of the mos t commonly used
microprocessors.

All microcomputer systems have the facility to run programs

written i n one or more high-level languages such as BASIC,
PASCAL, COBOL, FORTRAN and FORTH. However, s ome t i me s it is
necessary (because of speed or memory restrictions) or des irable
to run programs written in the assembly language of a
microcomputer system. If the microcomputer s y s t e m is based on a
Z80 microprocessor then the Z80 assembly language must be used.

This book covers all of the facilities provided by the Z80

assembly language, starting with the simple fa cilities and
building on them step-by-step to gradually include the more
complex facilities.

Learning a programming language is like learning a foreign

language - you need plenty of practical e xperience. At the end of
each chapter there i s a practical program for you to do. Programs
from some of the chapters are used in later practicals so that
you can develop quite substantia l programs, rela tively easily.
The practicals have been carefully chosen to be i n t e r e s t i n g and
visually rewarding - all practicals except the first produce some
form of output on the displa y.

ACKNOWLEDGEMENTS

My thanks to Leicester Polytechnic for allowing me to use their

computer facilities for the development of programs and
production of the book, and to Mark Folker for his thorough
checking of the text and many suggested improvements. Also, I
wish to thank my wife, Susan, for her support in many ways,
particularly in the typing of the text using a word processor.

Roger Hutty

ix
1 The zao Architecture

1.1 MICROPROCESSOR SYSTEMS

Computing, like many other processes, has three main parts

I input H pr.ocessi.og H output

In a computer system, data (numbers and words) is input by an

input device, the processing is performed by a central processing
unit and data is output by an output device.

A computer system can have many different types of device but as

far as a microprocessor system is concerned the most common
system is shown in Figure 1.1.

I central processing unit I

l and memory
I

-.
display

keyboard
/
Figure 1.1

A keyboard is used to input programs, and data for the programs.

A program is a list of instructions to tell the computer what to
do dur ing the processing stage. A display is used to output
information, such as results from a program. The memory is used
to store programs and data.

So for a microprocessor system, a keyboard is used as the input

device and a display is used as the output device.
2 Programming in Z80 Assembly Language

If you are not familiar with binary and hexadecimal number

systems and signed (2's complement) and unsigned numbers you
should work through Appendix A before continuing with the text.

1.2 THE Z80 CENTRAL PROCESSING UNIT

There are several different microprocessor central processing

units available. This book is concerned only with the Zilog 80
(Z80) microprocessor.

* Zilog and Z80 are trademarks of Zilog, Inc., with whom the
publisher is not associated. *

The components of the Z80 which are of most importance to a

programmer are the registers shown in Figure 1.2.

Accumula"or A F Flag register

B C

D E

H L

SP Stack pointer

PC Program counter

IX X index register

IY Y index register

I Interrupt vector
register

Figure 1.2

The accumulator is an 8-bit register used for arithmetic and

logical operations. For example, to add two numbers in the Z80
microprocessor the first number must be in the accumulator and
the second number is added to the accumulator, thereby leaving
the sum in the accumulator.

The flag register is used to hold information about the results

of some operations. For example, the flag register indicates
whether the result of adding a number to the accumulator is
positive, zero or negative.

The B, C, D, E, Hand L registers are often referred to as

secondary registers and are used mainly to store data
temporarily. They can be used as single 8-bit registers or as 16-
bit registers when they are referred to as BC, DE and HL. These
secondary registers tend to be used, by convention, in particular
 The Z80 Architecture 3

ways as you will see throughout the book. In particular, the HL

register pair is usually used to point to data in memory.

The 16-bit stack pointer is used to provide a stack facility

this will be explained later.

The 16-bit program counter is used by the central processing unit

to keep a track of the place in memory where the next instruction
to be obeyed is located.

The us e of the IX and IY index registers and the interrupt vector

register will be explained later.

1.3 MEMORY

The memory of a Z80 microprocessor system consists of locations,

usually called bytes, which are 8 bits long. Look at Figure 1.3.

o 01010010

01110111

2 10110000

3 01111110

/
address contents locations

Figure 1.3

The number of bytes in a memory varies from one system to another

but will normally be 4K, 8K, 16K, 32K, 48K or 64K which are round
binary values (equivalent to decimal tens and thousands, etc.).
lK is equivalent to decimal 10 24 (binary 10000000000). For
example, a 64K memory has decimal 65536 byt es or hexadecimal
10000 bytes.

The bytes of a memory are numbered sequentially starting at zero

- the number of a byte is referred to as its address. Each byte
contains an 8-bit pattern which is referred to as the contents of
the byte. The 8-bit pattern may represent anyone of several
quantities such as an instruction, a number or a character.

Considering the memory contents as shown i n Figure 1.3, the

contents o f the byte whose address i s 2 is hexadecimal BO (binary
1011 0000) •

A shorthand form of writing 'the content of a byte whose address

is' is to enclose the address in parentheses (curved brackets) so
that, for example, (2) is hexadecimal BO.
4 Programm ing in Z80 Assembly Language

In the remainder of this text binary numbers are postfixed with

the letter B and hexadecimal numbers with the letter H. Numbers
with no postfix letter may be assumed to be decimal.

1.4 INSTRUCTIONS

Have a quick glance at Appendix C where you will see the complete
zao instruction set.

In general, an instruction consists of an operation and an

operand. The operation indicates what has to be done and the
operand indicates what is to be used in the operation. For
example, the instruction

10010010

has an operation code (op-code) of 10010B which specifies that

the operand has to be 'subtracted from the accumulator' and an
operand of 010B which specifies the 'contents of the register D'
are to be used in the subtraction.

An instruction may occupy 1, 2, 3 or 4 bytes depending mainly on

how the operand is specified. Look at the first five
instructions listed in Table C.S of Appendix C. All five
instructions have an op-code of 'add to the accumulator'. Look
down the column headed No. of Bytes and you will see that the
number of bytes occupied by the instructions varies from one to
three. This is because there are differences in the
specification of the operand's address. The different ways in
which an operand may be addressed are called 'addressing modes' •

1.5 ASSEMBLY LANGUAGE

In the computer itself instructions are held in binary. We could

write programs in binary but this would be very tedious and
error-prone. We could take advantage of the hexadecimal number
system and use, for example, 92H in place of the instruction
10010010B. This would be an improvement but still tedious and
error-prone for any reasonably-sized program.

A more convenient way of writing programs is to use an assembly

language. An assembly language has many facilities to make
programming easier.

To start with, mnemonics may be used in place of operation codes.

Mnemonics are usually chosen to help a programmer by indicating
what the operation is. For example, the ins truc tion referred to
in the previous section, 'subtract register D from the
accumulator' may be written using the SUB mnemonic, as

SUB D

which is easier to remember than 10010010B, or even 92H. Notice

also that the operand part of the instruction, in this case
 The Z80 Architecture 5

register D, may be specified as the letter D in place of the code

010B.

There are many more facilities provided by the Z80 assembly

language which will be introduced to you throughout the text.

Assembly language programs cannot be executed directly by a

computer - they have to be converted to their equivalent binary
codes. This conversion is performed by a program called an
assembler. The assembler inputs an assembly language program,
called the source program, assembles the program and produces a
machine code program, called the object program, which can then
be executed.

All Z80 assemblers provide the same assembly language facilities

(with occasional odd exceptions). You will need to find out how
to assemble and execute programs on the Z80 computer system which
you intend to use for the practical programs.
2 Accumulator and Register
Instructions

In th is chapter we shall consider the one-byte operat ions of

loading registers, adding to the accumulator, subtracting from
the accumulator, incrementing and decrementing registers and
negating the accumulator. All these instructions allow 8-bit
arithmetic to be performed.

2.1 LOAD A REGISTER WITH A VALUE

Any of the s i n gl e registers may be loaded d irectly with a value

using the two-byte instruction

LD r,n

where r is any of the reg isters A, B, C, D, E, H o r Land n is an

uns igned number in the r ange 0 to 255 or a signed number in the
range -128 to +127. For example, LD B,99 will load the B
register with 99.

2. 2 ADD AND SUBTRACT A VALUE

Quant ities can be added to, and subtracted from, the accumulator
by means of the

ADD A,n and SUB n

instructions, respectively. For example, the sequence of

instructions

LD A,15
ADD A,46
SUB 22

will cause the accumulator to conta in 15 after the LD

instruction, 61 after the ADD instruction and 39 after the SUB
instruction. Hence, the instructions compute 15 + 46 - 22.

6
 Ac cumulator and Register Instruction s 7

Notice that the ADD instruction requires the accumulator to be

specifi ed explicitly, whereas the accumulat or is implied in the
SUB instruction.

2.3 ADD AND SUBTRACT A REGISTER

The contents of any of the registers can be added to, or

subtracted from, the accumulator using the

ADD A,r and SUB r

instructions, respectively. This allows, for example,

intermediate results which have been saved temporarily in one of
the registers to be added to, or subtracted from, the
accumulator.

2.4 LOAD ONE REGISTER WITH ANOTHER

A register may be loaded with the contents of another register

with the instruction

LD rl,r2

which loads register rl with the contents of register r2. This

instruction is often used to save the accumulator temporarily
whilst it is needed for another purpose, during which time the
saved value may be used. The contents of r2 are not changed
all LD instructions perform a 'copy' rather than 'load' function.

The s equence of instruct ions

LD A,56
SUB 22
LD B,A
ADD A,B
ADD A,B

computes 3 x (56 - 22). It does the 3 times multiplication by

adding 56 - 22 twice to itself. The result of computing 56 - 22
is stored in register B so that it can be added to the
accumulator by the last two instructions.

2.5 INCREMENT AND DECREMENT A REGISTER

The single-byte instructions

INC r and DEC r

increment and decrement, respectively, the register r by one.

These instructions are used mainly for looping as will be
explained later. However, they can also be used to add one or
subtract one during an arithmetic computation. In particular, INC
A and DEC A execute faster than ADD A,l and ADD A,-l or SUB 1 and
occupy one byte rather than two bytes of memory.
8 Programming in Z80 Assembly Language

2.6 NEGATE THE ACCUMULATOR

Another single-byte instruction

NEG

negates the accumulator. If, for example, the accumulator

contains 78 and a NEG instruction is executed the accumulator
will then contain -78.

The following table shows the contents of the accumulator, in

decimal and hexadecimal, after the execution of each of the three
instructions.
accumulator

LD A,27 27 (lBH)
NEG -27 (ESH)
INC A -26 (E6H)

2.7 ADDRESSING MODES - IMMEDIATE AND EXTENDED

The three instructions LD r,n, ADD A,n and SUB n are using the
immediate addressing mode - so called because the value of the
operand is included in the instruction (in the second byte for
these particular instructions).

Another addressing mode - extended addressing - allows the

address of the operand to be specified. For example,

LD A,027AH)

specifies that the contents of byte 127AH are to be loaded into

the accumulator. Conversely, the contents of the accumulator can
be stored in a byte using the instruction

LD (nn),A

where nn is the address of the byte in which the contents of the

accumulator are to be stored.

These two instructions occupy three bytes - the address of the

operand nn occupies the last two bytes of the instruction.

To demonstrate the difference between the immediate and extended

addressing modes consider the contents of the accumulator after
the execution of the two instructions LD A,3SH and LD A,(3SH)
assuming that byte 3SH contains 79 . The accumulator will contain
3SH after execution of the instruction LD A,3SH and 79 after
execution of the instruction LD A,(3SH).

By looking at Table C.2 in Appendix C you will see that, of the

single registers, only the accumulator may be loaded and stored
using the extended addressing mode.
 Accumulator and Register Instructions 9

2.8 LABELS

Keeping track of addresses in a program can be quite difficult.

The assembler overcomes this difficulty by allowing bytes in
memory to be given labels - like giving a house a name. A
labelled byte may then be referred to by its label rather than
its address number. Only those bytes to which a reference is
made are labelled, as shown in Program 2.1 which adds 10 to a
number.

Program 2.1 adds 10 to a number

0000 3AOO09 LD A, (NUMBER)

0003 C60A ADD A,10
0005 32000A LD (RESULT) ,A result number + 10
0008 76 HALT
0009 4A NUMBER: DEFB 74
OOOA 00 RESULT: DEFB 0

Labels are chosen by a programmer within the constraints that

they must consist of no more than six letters and digits, the
first of which must be a letter. A label is terminated by the
colon character. Careful thought in choosing labels can produce
intelligible programs as shown in Program 2.1. Lack of thought
can produce really 'bad' programs.

Program 2.1 shows the listing produced by the assembler after the
source program has been assembled. The first column of the
assembler listing gives the byte address of an instruction or
data byte and the second column is the object program, in
hexadecimal.

The listing shows how the assembler has assigned values - their
byte address number - to the labels. For example, NUMBER has the
value 0009H.

The HALT instruction causes the program to stop executing.

The DEFB pseudo operator requests the assembler to allocate a

byte and set it to the specified number. Pseudo operators will be
dealt with in more detail later in the text.

Comments should always be included in assembly language programs

to make their function understandable to a human reader. As shown
in Program 2.1 comments start with a semi-colon. The assembler
ignores all characters after a semi-colon in a line during
assembly but includes them in the listing. Comments may start at
any position in a line, including the beginning of a line. It is
difficult to define the level of comments - certainly, the level
should be higher than a comment beside each instruction. Looking
at the comments in the programs throughout the text will give you
a good idea of the level required.
10 Programming in Z80 Assembly Language

2.9 PROGRAM

Write a program which computes

RESULT = N1 - 3(N2 + 1) - 1

Set up Nl and N2 and RESULT as data at the end of your program.

When supplying values for the variables ensure that all results,
intermediate and final, will be in the range -128 to 127.

Enter the program in the computer and have it assembled and

executed. Check the computation by looking at the contents of
the byte RESULT. Find out from your computer reference manual
how to look at the contents of bytes.

To see what is happening during execution of your program,

single-step through the program looking at the contents of the
registers and relevant bytes after the execution of each
instruction. Find out from your computer reference manual how to
single-step through a program.

Change the values of N1 and N2 and repeat the assembly,

execution, checking and single-stepping of your program.
3 Subroutines and Display Output

3.1 SUBROUTINE CONCEPTS

Subroutines are a very important feature of programming which,

un fortunately, are usually left until the end of a programming
course. At this stage in the book you will learn why subroutines
are used and how to use them, but not how they work. The
subroutine mechanism will be covered in a Chapter 7.

Let us first look at the reason for using subroutines. Suppose a

program contains two or more groups of statements which are
functionally identical, as indicated by the shaded areas on the
left of Figure 3.1.

program >main
program

1
/ / / / JI
su broutine

Figure 3.1

11
12 Programming in Z80 Assembly Language

It is wasteful to repeat the statements, so instead we make the

group of statements into a subroutine as shown on the right of
Figure 3.1. The program now consists of a main program followed
by the subroutine. In the main program, when the group of
statements is required to be executed, a reference is made to the
subroutine, the group of statements is executed and return is
made to the main program. Execution then continues normally in
the main program until a further reference is made to the
subroutine.

An important point to notice is that references from the main

program are made to the same place (the start of a subroutine)
but returns to the main program are made to different places in
the main program (following the reference). The mechanism for
returning to the correct place in the main program will be dealt
with later. When a main program makes a reference to a
subroutine we say that it is calling the subroutine.

3 .2 THE CALL AND RET INSTRUCTIONS

Subroutines are called in the Z80 microprocessor by using the

CALL instruction followed by a label which is the label of the
first instruction of the subroutine. Program 3.1 shows a
subroutine being called twice from a main program.

Program 3.1 multiplies two numbers by 4

LD A,(N!) start of main program

CALL QUAD compute N1 x 4
LD (Rl) ,A Rl = Nl x 4

LD A,(N2)
CALL QUAD
LD (R2) ,A R2 N2 x 4
HALT
N1 : DEFB 31
N2: DEFB 25
Rl: DEFB 0
R2: DEFB 0
,
; Subroutine - multiplies the accumulator by 4

QUAD: ADD A,A Ax 2

ADD A,A (A x 2) x 2 A x 4
RET

The subroutine, referred to by its label as QUAD, multiplies the

number in the accumulator by four.

Going through the program as it would be executed, first the

accumulator is loaded with N1 and then the subroutine QUAD is
called with the CALL QUAD instruction. This causes the
subroutine 's first instruction ADD A,A to be executed, followed
by the second instruction ADD A,A. The last instruction in the
 Subroutines and Display Output 13

subroutine, RET, is used to return to the main program - to the

LD (Rl),A instruction in this case.

The main program then continues its execution with the LD A,(N2)
instruction which loads the accumulator with N2 before calling
the subroutine again with the second CALL QUAD instruction. The
subroutine's two ADD instructions will be executed, followed by
the RET instruction which causes a return to the main program
to the LD (R2),A instruction this time. Finally, the program
terminates with the HALT instruction.

The accumulator is being used by the QUAD subroutine to accept

data - the number to be quadrupled - from the main program and to
return data - the number quadrupled - to the main program. We
say that the accumulator is being used to 'pass parameters' to
and from the subroutine, the parameters in this case being the
number and the number quadrupled. Other registers, memory bytes
and the stack may also be used to pass parameters as we shall see
later.

The following is another example of a simple subroutine called

SUMREG which sums the contents of the single registers B, C, D,
E, Hand L and leaves the result in the accumulator.

; subroutine to s um the single registers

SUMREG: LD A,a
ADD A,B
ADD A,C
ADD A,D
ADD A,E
ADD A,H
ADD A,L
RET

A main program may need to use more than one subroutine. In that
case, all the required subroutines are located at the end of the
main program, one after the other.

3.3 DISPLAY OUTPUT

One advantage of subroutines is that a programmer can use someone

else's subroutine without necessarily knowing how the subroutine
performs its function.

You would usually have to go further than the third chapter in an

assembly language programming book to learn how to output to the
display. However, if you are given a subroutine which performs
that function and told how to use it, it does not matter that you
do not understand what is going on inside the subroutine.

A subroutine to output a character to the display is given in

Program 3.2.
14 Programmi ng in Z80 Assembly Language

Program 3.2

Subroutine to ou t pu t a character t o the display

-on entry the acc umulator contains the code of the
character t o be output
-all re gisters preserve d

COUT : PUSH AF ; save all registers

PUSH BC
PUSH DE
PUSH HL
LD E,A cha r a c t e r int o E
LD C,02H
CALL 0005H output character
POP HL restore all registers
POP DE
POP BC
POP AF
RET ; return

The subr outine is ca ~led by the instruction CALL COUT after

ensuring that the code of the character to be output is in the
ac cumulator. Incidentally, the cha r a c t e r code is still in the
accumulator on return from the subroutine.

The ASCII chara cter codes are given in Append i x D. A character's

ASCII code i s the binary patt ern which is used when transmitting
a cha r a c t e r fro m the cpu to a devi ce and vice versa. Look ing at
the ASCII table, you can see that the ASCII code for G is 47H and
the ch aracter wi th an ASCII co de of 28H is + .

The d isplay output subroutine above is t he one used with Z80

microcomputer s ystems which ope r a t e under the CP/ M (Control
Progra m/Moni tor) op erat ing s ystem. You must ch eck with y ou r
computer referen ce manual o r computer supplier that the display
output subroutine is co r r e c t in every detail for y ou r part icular
Z80 computer s yst em. There may be small di f fer ences or, i n fact,
the who le subroutine may be quit e different.

A character code can be specified in a program by writing either

the numeric equ ivalent of the code or the character between
a pos t r ophe s . For example , LD A,2AH and LD A,'*' can both be used
to load the accumulator with the ASCII code of the * character.
The latter form is preferable because i t is more understandable -
the assembler does the work of converting it into its numeric
equivalent.

3.4 PSEUDO OPERATIONS

Pseudo operat ions are so-called because, although they appear in

an assembl y language program in the same place as th e operator of
an e xecutable instruct ions, they are not e xecuted - not even
assembled. In fact, pseudo operations are informat ion provided by
the programmer for the assembler. We have already used one pseudo
 Subroutines and Display Output 15

operator - the DEFB pseudo operator which informs the assembler

that a byte and its contents are being defined.

There are several other pseudo operators which will be i n t r o d u c e d

throughout the text. For now, we will look at two little-used
pseudo operators.

The ORG nn pseudo operator informs the assembler that the

following instruction is to be assembled so that it starts at
address nn. The mnemonic ORG, which is short for ORiGin, was
chosen because this pseudo operator is mainly used to indicate
the address of the start of a program, although it can be used
anywhere within a program. The ORG pseudo operator allows a
programmer to specify the absolute address of the place in memory
where a program is to reside. However, most microprocessors are
used under the control of an operating system, such as CP/M
(Control Program / Monitor), which either controls the allocation
of memory to programs or allows a program to be located at an
absolute place after the assembly of the program - usually during
the linking phase. Hence, it is now standard practice for an
assembler always to assemble a program relative to byte zero, and
the ORG pseudo operator has gone out of use.

The END pseudo operator has also now gone out of general use. Its
function is to inform the assembler that there is no more program
to assemble, that is, to indicate the end of a program . This is
necessary in some environments, such as the old paper-tape
systems, where it is not clear that the end of the program has
been reached. However, programs are now more normally held in
files, of one kind or another, and the assembler automatically
finds the end of a program by detecting the end of the file
containing the program. Purists would say that a program should
always finish with the END pseudo operator but since it is
usually obvious where the end of a program is, it is now rarely
used. Exceptionally, some assembler systems require programs to
be terminated with an END pseudo operator.

3.5 PROGRAM

Using the COUT subroutine write a program which outputs your

initials followed by a carriage-return and line-feed (referred to
as CR and LF in the ASCII character codes in Appendix D).
4 Unconditional Jumps and
Keyboard Input

4.1 UNCONDITIONAL JUMPS

Execution of Z80 programs we have looked at so far proceeds

sequentially through the instructions, one after another.
However, it is possible and, in fact, usual to change the
sequence of execution for one reason or another. Jump
instructions allow the execution of a program to take different
paths through the i n s t r uc t i o n s . There are two types of jump
i ns t r uc t i o n s - unconditional and conditional. We will look at
conditional jumps in the next chapter.

An unconditional jump has the form

JP nn

where nn is the address of a memory byte . The instruction will

cause the instruction at address nn to be executed after the JP
instruction. The address is normally specified as the label of
the instruction to be jumped to, rather than its actual address.

Without conditional jumps the use of the JP instruction is

limited to providing indefinite loops, as shown in Program 4.1.

Program 4.1 an indefinite loop

initial part of program

LOOP: *
* instructions to be repeated

JP LOOP
*jump to LOOP

The initial part of the program will be executed first, followed

by the instructions to be repeated. Then the JP LOOP instruction
causes a jump to the first instruction of those to be repeated.
The remainder of the instructions in the loop will be executed
until the JP LOOP is executed again, and so on .

The programming of indefinite loops is not recommended because

the only way to stop the looping is to ' c r a s h ' the system by
pressing the RESET button or worse, switching off the computer.
,J s e of conditional jumps prevents the need to code indefinite
~ o o p s , as we shall see in the next chapter.

16
 Unconditional Jumps and Keyboard Input 17

The following set of instructions repeatedly output an asterisk

character to the display.

LD A,'*'
NEXT: CALL COUT
JP NEXT

The JP instruction allows a jump to be made to any location in a

memory of 64K bytes. However, because most jumps are to
instructions within a few bytes from the jump instruction a
shorter version, the JR instruction, is also available. The JR
mnemonic stands for 'Jump Relative', so-called because the
address part of the instruction specifies the number of bytes
between the JR instruction and the instruction to which the jump
is to be made. The JP instruction occupies 3 bytes whereas the JR
instruction occupies only 2 bytes. Also, the JR instruction has a
faster execution time than the JP instruction.

4.2 KEYBOARD INPUT

In the same way that you were given a subroutine to output to the
display you are now given a subroutine called CIN which accept's a
character from the keyboard. The subroutine is shown in Program
4.2.

Program 4.2

Subroutine to input and echo a character from the keyboad

-on return the accumulator contains the code of the
character

CIN: PUSH BC save registers

PUSH DE
PUSH HL
LD C,OlH
CALL OOOSH input character
POP HL restore registers
POP DE
POP BC
RET

On return from the subroutine, the accumulator contains the ASCII

code of the character which was input from the keyboard. The
subroutine is called by the instruction CALL CIN.

A character which is input from the keyboard is not always

automatically displayed' on the display unit. If the input
subroutine supplied for your computer system does not
automatically echo the keyed character on the display you should
include a CALL COUT at the end (but before the RET) of the CIN
subroutine.
18 Programm ing in Z80 Assembly Language

You s h ould check with y ou r computer reference man ual or comp uter
s u p p l i er that the input subroutine in Program 4 .2 is correct, in
every detail , for your particular compute r system. You will als o
need to check if an input character is automatically echoed, or
not . The subroutine given is the one used with Z80 microcomputer
syst~ms operating under the CP/M operating s ys tem .

Groups of sta temen ts which are re quired more t h e n once in a

program or which are likely to be required by more than one
program should be made i n t o a su broutine. The p r o g r a m you had to
write in Chapter 3 required a carriage- return and li ne-feed to be
output to tne display. This is obviously going to be a common
r eq u i r e ment in many programs which affect the display . I t is
prude nt, t h ere f ore, to make the g roup of statements which perfo rm
that func tion into a su brout ine ca l led CRLF, which can then be
cal led from a prog ram by t he ins t ruction CALL CRLF wheneve r a
carriage-re turn a nd a l i ne - f e e d is required to be o u tpu t . Th e
su broutine CRLF wou ld be coded a s fo llows

; su broutine t o o u tput a carriage-return a n d line -feed

CRLF: LD A,13 output carriage-return

CALL COUT
LD A,l O output line -f eed
CALL COUT
RET

This approach t o programm in g is referred t o as modular

p ro g r a mm i n g , or more precisely modular desi gn. Although normall y
assoc iated with hi gh -level l a ng u a g e p rog ramm i n g , it is a
desirable featur e of asse mbl y language programming b e c ause it
s a v e s ti me when a program is originat ed and a l s o reduces
subs equ ent ma i n t e na nc e time s a s s oc i a t e d with a program.

4 .3 CHARACTER CODES AND VALUES

You must be careful when working with digits that y ou do not

confuse a digit's character code with its value. The two a re
quite diffe rent.

The values of t h e digits 0 to 9 are r e p r e s e n t e d in registers and

memory by t e s in the co mputer by the numbers OOH to 09H.

The character codes of the digi ts 0 to 9 are represented by t he

numbers 30H t o 39H (see Appendix D) .

The values of digits are used when doing arithmetic in the

computer and the character codes of digits ar e used during input
and output of digital c ha r a c t e r s . The c h a r a c t e r code of a digit
which is input and needs to be used in arithmetic must be
converted to the va lue of the digit, and vi ce v ers a .

An exa mp l e showing the differ ence between a digit's value and

its character code i s gi ven by the two instructions LD A,'7' and
 Unconditional Jumps and Keyboard Input 19

LD A,S. The LD A,'7' instruction loads the accumulator with the

number 37H and the LD A,S instruction loads the accumulator with
OSH.

When a digit is input it must have 30H subtracted from it before

the digit can be used in any arithmetic computation. Similarly, a
single digit result of an arithmetic computation must have 30H
added to it before it is output. The 30H addition and subtraction
are necessary to convert from a digit's character code to its
value, and vice versa.

4.4 THE EQU PSEUDO OPERATOR

The EQU pseudo operator is a little-used pseudo operator which

should be used more often. It allows a programmer to give a
constant a name by assigning a value to a label. Its main
advantages are that it makes a program more readable and easier
to change for different environments. For example, when loading
the accumulator with the carriage-return character code, it is
more obvious what is being done if
CR: EQU 13

LD A,CR

is used. The EQU (short for EQUate) pseudo operator normally

occurs at the beginning of a program. A label may only be equated
once, but may be used as many times as necessary once it has been
defined.

Another example of the EQU pseudo operator being put to good use
is the naming of the address in the input and output. The outline
of a program using the display and keyboard subroutines given in
Programs 3.2 and 4.2 would look like

OPSYS: EQU OOOSH

CIN:
CALL CPM

COUT:
CALL CPM

Now, if it is necessary to change the OPSYS address, only the EQU

pseudo operator at the beginning of the program need be amended
and the program re-assembled.

In general, all numbers in a program which represent a code or an

address should be equated. This does not include numbers which
are to be used for the value which they represent.
20 Programm ing in Z80 Assembly Language

4.5 PROGRAM

Write a program which repeatedly inputs (and echoes) two decima

digits from the keyboard and outputs their sum. It may b
assumed that the sum of the two digits does not exceed 9.
typical run of the program would leave the d isplay containing

2+2=4
5+3=8
4+1=5

It is the program's responsibility to output the + and

characters.
5 Flags, Conditional Jumps
and the CP Instruction

5.1 THE FLAG REGISTER

The Z80 has an 8-bit flag register which is used to contain

i nf o rma t i on regarding the result produced by the last executed
instruction. In fact, only six of the eight bits are used as
follows

S Sign flag
Z Zero flag
H Half-carry flag
p/v - Parity/Overflow flag
N Add/Subtract flag
C Carry flag
X Unused bits

The SIGN FLAG is set to 1 if the result produced by an

instruction is negative, otherwise the flag is reset to O. For
example, after execution of the instructions

LD A,23
SUB 56

the accumulator will contain -33 and the sign flag will be set to
1.

The ZERO FLAG is set to 1 i f the result produced by an

instruction is zero, otherwise the flag is reset to O. The zero
flag would be reset to 0 after the execution of the two
instructions above.

The operation of the Sand Z flags will suffice for now - the H,
P/V, Nand C flags will be dealt with later in the book.

21
22 Programming in Z80 Assembly Language

Not all instructions affect all of the flags. For example, none
of the LD instructions affects any flags . You can discover which
instruction s affect which flags by looking in the third column of
the t a b l e s in Appendix C. Glancing through the tables you will
see that the sign and zero flags are mainly affected by
arithmetic, shift, rotate and bit ins tructions .

The contents of t h e accumulator and the S a nd Z flags after

exec ution of each of the f ollowing six instructions are

A S Z

LD A,120 120 ?
SUB 122 -2 1 0
LD B,A -2 1 0
SUB B 0 0 1
ADD A, 70 70 0 0
NEG -7 0 1 0

We shall no w look at conditional jump instruc tio ns which use the

flag bits.

5. 2 CONDITIONAL JUMP I NSTRUCTI ONS

Conditional jump instructions allow a program either to continue

executing t h e instruct i ons following the instruction or to
exe c u t e a set of instructions e lsewhere in t he program , depending
on the state of one of the flag bits . This allows a program t o
go one way or another depending on a condition. A conditional
j ump instruction is used i n Pr ogra~ 5 .1 .

Program 5.1

LD A, (X)
SUB 10 compute X-I0
JP Z,EQUAL
LD A,1 X I- 10
JP CaNTIN
EQUAL: LD A,O X 10
CONTIN: -

X: DEFB 25

The program resets t h e a ccumulator to 0 if X is equal t o 10 or

se ts it to 1 if X is no t e qual t o 10. The SUB instruc tion
computes X - 10 and sets the sign and zero flags accor ding to t h e
result of that computation. With X having a value of 25, both
the sign and zero flags will be reset to O.

The JP Z,EQUAL instruction causes a jump to the instruction

labelled EQUAL if the zero flag is 1, or execution of the
following instruc tion, LD A,I, if the zero flag is O. With X
having a value of 25 , no jump will be made , and the LD A,1
instruction will be executed followed by the unconditional jump
 Flags. Conditional Jump s and the CP Instruction 23

i ns t r uc t i on JP CONTIN.

I f t he value o f X had been defined as 10 then the o r d e r of

e xe cut ion of instructions would have been LO A,( X) SUB 10
JP Z, EQUAL a nd LO A,O.

Four of the conditions whic h can be tested by a conditional jump

instruction are

Zero j ump if zero flag is 1 ego JP Z,E QUAL

Non Zero jump if zero flag is 0 ego JP NZ,OIFF
Minus jump if sign flag is 1 ego JP M,NEG
Positive jump if s ign flag is 0 ego JP P,POS

The following prograc segment outputs the letter N if the sum of

the two numbers which are in the Band C registers, is negative,
the letter Z if the sum is zero or the letter P if the sum is
.>ositive, but not zero.

LO A,B
ADD A,C B+ C
JP M,NEG
JP Z, ZERO
LO A, ' p' +ve
JP DONE
NEG: 10 A,' N' -ve
JP DONE
ZERO: LO A, ' z' 0
DONE : CALL COUT

The JP Z and JP NZ ha ve equival ent relative jump i nstructions JR

Z and JR NZ in the same way as the unconditional JP instruction
has an equivalent JR i ns t r uc t i on . The JP M and JP P instructions
d o not have relativ e equi va l en ts .

5.3 THE COMPARE INSTRUCTION

The Z80 compare instruction i s very useful because i t allows the

a c c umul a t or t o be compared with another value without affecting
the contents of the a ccumul a t or as, for exampl e, do t he SUB and
ADD instruct ions. The compare instruction computes the value of
the operand subtract ed f rom the accumulator and s ets the flags
ac c or d i ng to the result of the computation . The result does NOT
replace the contents of the accumulator - the result is
discarded. The instruction has a mnemonic of CP a nd can have an
o pe ra nd of a r egister or an 8-bit value.

The most straightforward us e of the CP instruction is determining

if t he accumulator con t a i ns a spec i f ic value or not. For example,
the i nstructions

CP 50
JR Z,FIFTY

will caus e the Z fl ag to be se t to 1 i f the accumulator contains

24 Programming in Z80 Assembly Language

50 and reset to a otherwise, followed by a jump to the

instruction labelled FIFTY if, in fact, the accumulator does
contain 50.

The following program segment causes a jump to an instruction

labelled LESS if the value of the variable COUNT is less than
100, to an instruction labelled EQUAL if COUNT is equal to 100 or
to an instruction labelled GREAT if COUNT is greater than 100.

LO A,(COUNT)
CP 100
JP M,LESS COUNT < 100
JP Z,EQUAL COUNT 100
JP GREAT COUNT > 100
The CP a instruction is useful for setting the flag register
after an instruction which does not affect the flags. For
example, the LO instruction in the sequence

LO A,(TEMP)
CP a
does not affect the flags and a CP a instruction is, therefore ,
necessary if the status of the acc umul a t o r is required.

5.4 CONDITIONAL LOOP TERMINATION

The loops which were considered in the previous chapter were

referred to as indefinite loops because there were no
instructions to stop the looping.

We shall be considering several different methods of terminating

loops in subsequent chapters, but for now we shall consider just
one method in which the loop is terminated when a specified
condition occurs. Program 5.2 contains a loop which inputs
characters until a blank character is i nput .

; Program 5.2
,
NEXTCH: CALL CIN
CP
JP NZ,NEXTCH
BLANK:

The three instructions in the program will be repeatedly executed

until the character input is the blank character . When this
occurs, the CP instruction will produce a zero result causing the
instruction following the JP NZ,NEXTCH instruction to be executed
next rather than the instruction labelled NEXTCH.
 Flags. Conditional Jumps and the CP Instr uction 25

5.5 PROGRAM

First write a subroutine to categorise a character code. On

entry to the subroutine the accumulator contains the character
code. The subroutine returns with the accumulator unchanged and
register B containing

-1 if the character is a decimal digit, 0 to 9

o if the character is a letter of the alphabet (small
and capital), a to z or A to Z
or if the character is a special character (neither a
decimal digit nor a letter of the alphabet)

You will need to consult the table of character codes in Appendix

D. Use the Compare instruction to determine first if the
character code is less than '0', in which case the character is a
special character, second if the character code is less than : ,
in which case the character is a digit, and so on.

Use the subroutine to write a main program which repeatedly

inputs a character from the keyboard and responds, after
outputting a space, by displaying the letter D, A or N indicating
that the character just input was a decimal digit, an alphabetic
character or neither, respectively. The letter should be
followed by a carriage-return and line-feed.

Pressing the RETURN key should terminate the program.

6 Counting Loops and the Stack

There are many different constructions used in looping. We have

already used an 'indefinite loop' construct (not recommended) and
a 'conditional terminated loop' construct. We shall consider
several more types of loop construct throughout the book,
starting now with a 'counting loop' construct.

6.1 COUNTING LOOPS

One of the simplest loop constructions allows a sequence of

instructions to be executed n times. An example of this is shown
in Program 6.1 which inputs 10 digits and sums them.

Program 6.1 input and sum 10 numbers

LD C,O C contains running sum

LD B,10 initialise loop counter
NEXT: CALL CIN input digit
SUB 30R convert digit to its value
ADD A,C add digit to running s um
LD C,A
DJNZ NEXT
CONTIN: -

The sequence of instructions between, and including, CALL CIN and

LD C,A will be executed 10 times. The loop counter, register B
in this example, is initially loaded with the number of times
that the loop is to be executed using the LD B,10 instruction.
The DJNZ (Decrement and Jump if Not Zero) instruction causes
register B to be decremented and a jump made to the specified
label NEXT if B is not zero. When B does become zero, after the
loop instructions have been executed 10 times, execution will
continue with the instruction following the DJNZ instruction,
that is, the one labelled CONTIN.

The maximum loop count that can be used by the loop construction
in Program 6.1 is 255.

It is not good programming practice to write specialised programs

like, for example, Program 6.1 which has been written for a fixed
loop count of 10. Normally, programs are written to be of

26
 Counting Loops and the Stack 27

general use, although not so general that it is difficult to use

a program for a particular case. For counting loops the initial
value of the loop counter can be a parameter of a subroutine if
the loop is contained in a subroutine, a value input from the
keyboard, or a value computed prior to the loop.

An example of a counting loop which obtains its initial value

from the keyboard is shown in the following program which inputs
a digit n from the keyboard and displays n asterisks.

program to output n asterisks

CALL CIN input digit

SUB 30H convert to value n
LD B,A
LD A , '*'
NEXTAS: CALL COUT output *
DJNZ NEXTAS
HALT

Other registers, and memory bytes, can be used as the loop

counter as well as the B register. The reason why the B register
is a first choice is because of the DJNZ instruction. There is no
equivalent instruction for the other registers. A counting loop
using, for example, the accumulator looks like

LD A,initial-value
NEXT:

DEC A
JR NZ,NEXT

which requires two separate instructions to decrement and test

the accumulator. Hence, DEC B followed by JR NZ,label is
equivalent to DJNZ label.

It is sometimes convenient to use the loop counter within the

sequence of instructions to be repeated. Suppose, for example,
we wanted to sum the numbers from 1 to n.

Program 6.2 does just that by adding in the loop counter, which
goes from n to 1, each time through the loop.

; Program 6.2 sum the numbers 1 to n

N: EQU 11
LD
LD
A,O
B,N
set sum to °
NEXADD: ADD A,B add n to sum
DJNZ NEXADD
28 Programming in Z80 Assembly Language

The loop counter is being used to count the number of times that
the sequence of instructions is executed. It is also being used
as the value to be added in each time round through the loop.
After looping is complete the accumulator will contain the sum.
Notice that the accumulator must be set to zero to start with.

When the loop counter is being used within a repeated sequence of

instructions it is not always convenient to count down to O. If
the lower value is not zero then a compare instruct ion must be
used to detect the end of looping. As the compare instruction
operates only on the accumulator, this technique implies the use
of the accumulator as the loop counter. An example of such a
program is the following program which outputs the decimal digits
9 to 1, in descending order.

program to output the digits 9 to 1

LD A, '9' ; code for 9 (39H)

NEXDIG: CALL COUT
DEC A
CP '0' code for 0 (OOH)
JR NZ,NEXDIG
HALT

6.2 NUMBER INPUT

So far we have only considered the input of a single digit number

from the keyboard. Inputting a number with more than one digit
is not so straightforward.

If, for example, we wanted to input the number 123 it would have
to be input as three separate digits - the digit 1 followed by
the digit 2 followed by the digit 3. This would be done with
three CALLs to the CIN subroutine so the program would then have
the character codes for the three digits 1, 2 and 3. How then do
we convert these codes to form the 8-bit value 123 i n a register
or memory byte ?

First of all each digit's character code must be converted to the

digit's value. Then the first digit's value is multiplied by
100, the second digit's value is multiplied by 10 and these two
products are added to the third digit's value to give the total
value of the number. So that for the number 123 the computat ion

1 x 100 + 2 x 10 + 3

would be performed.

An equivalent, but more efficient, algorithm for computing the

value of a 3-digit number is

«first digit x 10) + second digit) x 10 + last digit

which can easily be extended to an algorithm which inputs an n-

digit number.
 Counting Loops and the Stack 29

6.3 THE STACK

The zao stack is an area of memory in which a stack facility

exists. The main characteristic of a stack is that items may
only be added to, and removed from, the top of the stack. A
stack is also referred to as a LIFO (Last-In-First-Out) queue.
Figure 6.1 shows how a zao stack is organised.

A stack has a top and a bottom. The bottom of the stack is where
the first item was added to the stack and the top of the stack is
where the last item was added. (A ZaD stack item is two bytes.)
The stack pointer (SP) register always points to the top of the
stack. Initially, when the stack is empty, the stack pointer
points to just below the bottom of the stack.

low address

- _stack top
I -=======r-
stack pointer(SP)
register _stack bottom

high address

.....
Figure 6.1

You can see from Figure 6.1 that the stack bottom is a high
address and the top is a low address, so usually the stack bottom
is a memory byte with one of the highest addresses. The address
of the stack bottom can be set by program instructions and there
may be more than one stack. However, a system will usually have
only one stack and the address of that stack bottom will be set
by the operating system.

Items may be added to the stack, or pushed on the stack, as we

say, when the SP register will be decremented by two to point to
the new top of the stack. Also, items may be removed from the
stack, or popped off, as we say, when the SP register will be
incremented by two to point to the new top of the stack.

The stack is used mainly for the temporary storage of data and
addresses, an example of which we shall look at in the next
section and also in the subroutine mechanism which will be
discussed in the next chapter.
30 Programming in Z80 Assembly Language

6.4 THE STACK INSTRUCTIONS

The stack bottom can be initialised using one of several

instructions which load the SP register with a value.

The two most commonly used of these instructions are LD SP,value

and LD SP,(label). So, for example, LD SP,32767 will set the
bottom of the stack to the byte with address 32767.

Having initialised the stack, items may be pushed on the stack

and popped off the stack. Only the contents of the register
pairs AF, BC, DE, HL, IX and IY may be pushed on and popped off
the stack, so for any operation on the stack, its size increases
or decreases by two bytes. The two instruct ions for popping and
pushing are POP rp and PUSH rp, where rp is anyone of the
register pairs l isted above. When a register pair has been
PUSHed on to the stack the contents of the SP register are
decremented by two to point to the new top of the stack, and when
a register pair has been POPped off the stack the contents of the
SP register are incremented by two.

The operation of the PUSH and POP instructions is shown in the

following table in which the contents of the registers A, B, C,
D, E and SP are shown after execution of each of the
instructions.

A B C D E SP

LD SP,16383 DAH DBH DCH DDH DEH 16383

PUSH AF OAH OBH OCH ODH OEH 16381
PUSH BC DAH DBH DCH DDH DEH 16379
PUSH DE OAH pBH OCH ODH OEH 16377
POP BC DAH DDH DEH DDH DEH 16379
POP DE DAH DDH OEH DBH DCH 16381

Another useful stack instruction is EX (SP),HL which exchanges

the top of the stack with the contents of the HL register pair.
The exc ha nge may take place also with the IX and IY registers
using the EX (SP),IX and EX (SP),IY instructions, respectively.

6.5 SAVING AND RESTORING REGISTERS

A subroutine should normally leave all registers in the state

that it finds them, except those which are used to pass
p~rameters back to the main program or calling subroutine. The
technique for doing this is referred to as 'saving and restoring
registers'. On entry to the subroutine, the subroutine saves any
registers that it is going to use and then restores those
registers just before returning. The simplest method of saving
and restoring registers uses the stack. For example, if a
subroutine uses registers B, C and D, then the start and end of
 Count ing Loop s and th e Stack 31

the subroutine would look like

SUBIN: PUSH BC
PUSH DE

POP DE
POP BC
RET

When using the stack in this way care must be taken to ensure
that registers placed on the stack are removed before the RET
instruction and also that the registers are POPped off in reverse
order.
We are now able to code the CRLF (carriage-return line-feed)
subroutine in a better way - using EQU and saving and restoring
the accumulator which would otherwise be corrupted by the
subroutine. The improved version is as follows

; subroutine to output carriage-return and line-feed

CR: EQU 13
LF: EQU 10
CRLF: PUSH AF
LD A,CR output carriage-return
CALL COUT
LD A,LF output line-feed
CALL COUT
POP AF
RET

When a subroutine does not save used registers, there should be a

comment at the beginning of the subroutine indicating which
registers are affected by the subroutine.

6.6 PROGRAM

Repeated addition ~In be used as a crude method of multiplying

two numbers since, for example, 3 x 4 is equivalent to 3 + 3 + 3
+ 3. So, to multiply p by q, P is added to itself q - 1 times.

Write a subroutine called MULT which multiplies the B register by

the C register and leaves the product in the accumulator. The
subroutine may assume that the resulting product is not too large
to be contained in the accumulator.

Write a subroutine called NUMIN which inputs an .unsigned decimal

number consisting of any number of digits. The number is
terminated by any non-digit character and may have leading spaces
which are ignored by the subroutine. On return from the
subroutine the accumulator should contain the number, the B
register should contain the number of digits in the number and
the C register should contain the non-digit terminating character.
32 Programming in Z80 Assembly Language

Use the MULT and NUMIN subroutines, and subroutines from previous
programs, to write a main program which repeatedly inputs (with
echo) a decimal number terminated by a space and outputs the
number of digits (I, 2 or 3) in the number, a space, the word OK
if the number is in the range 50 to 100, inclusive, or otherwise
the word NO, and a carriage-return and a line-feed.

One advantage of modular programming is that a module - a

subroutine in assembly language programm ing - can be replaced by
another module without affecting the rest of the program provided
that the new module performs exactly the same function as the old
module and that parameters are passed in the same way. For
example, when you are able to write a 'better' multiply
subroutine, the crude one you have used in this program can
simply be replaced, without affecting any program which uses the
subroutine.
7 Nested Loops and Addressing
Modes

7.1 NESTED LOOPS

A natural extension of the loops we have used so f a r is a

construction which has a loop within a l oop, with in a lo op, and
so on. This type of constructi on i s referred to a s ne sted loops,
becaus e each loop nest s inside another loop. The terms outer
loop and inner l oop are us ed to describe a loop cont ain ing
ano t h e r loop, and a loop whi ch is i nsi de a no t he r loop ,
respect i vely. Program 7.1 shows a program with t wo loops.

Program 7.1 ou tpu t s 4 lines of 6 *'s

LD C,4 i nit ia lis e line cou nt

NE XTC: CALL CRLF line l oop start
LD A, ' * '
LD B,6 initial ise * co unt
NE XTB: CALL COUT * l oop start
DJNZ NEXTB * loop e nd
DEC C
JP NZ,NEXTC line l oop en d
HALT

The program outputs fo ur line s of si x aster i s ks. The oute r l oo p

uses the C register t o co unt the number of lines a nd the i nne r
loop uses the B regist er t o count the number o f a s t e r i sks .

Although Program 7.1 sh ows two standard count ing l oops, there may
be more than two loops nested a nd a ny one of the nested loops may
be any type of loop which is used in any way a nd terminated in
any way.

7.2 IMMEDIATE EXTENDED AND REGISTER INDIRECT ADDRESSING MODES

The Z80 microprocessor has ten addressing modes, that is, ten
different ways of speci fying the operand o f an instruct ion. We
have already looked a t two a ddres si ng mode s expl i ci t ly (immediate
and e xtended), but have us ed another three (re g ister, implied and
rela tive) without speci al menti on.

33
34 Programming in Z80 Assembly Language

Immediate addressing mode has the value of the operand included

in the instruction - in the second byte. The instruction CP 27
has an immediate operand of 27.

Extended addressing mode has the address of the operand included

in the instruction - in the second and third bytes. The
instruction LD A,(1234H) has an extended operand of 1234H.

Register addressing mode has one of the registers as the operand.

The register B is specified as the register operand in the
instruction DEC B.

Implied addressing mode has the operand implied, that is, not
explicitly stated in the instruction. The accumulator is the
implied operand in the instruction NEG.

Relative addressing mode is the one in which the instruction

contains a relative displacement between a jump instruction and
the instruction to which the jump is to be made, such as the JR
conditional and unconditional jump instructions.

Figure 7.1 compares the immediate and extended addressing modes.

§8
opcode DA opcode

~ operand 78 operand
address
56

SUB 73 LD A, (5678H)

Immediate addressing Extended addressing

Figure 7.1

We shall now consider two more addressing modes - register

indirect and immediate extended.

Immediate extended addressing mode is, as its name suggests, an

extension of the immediate addressing mode. Immediate addressing
refers to 8-bit values whereas immediate extended refers to 16-
bit values.

In the immediate extended addressing mode the operand is the 16-

bit value in the last two bytes of the instruction. For example,
the instruction LD BC,4985 would cause the register pair BC to be
loaded with the value 4985. The three register pairs BC, DE and
HL may all be loaded with a value, although you will find that
the register pair HL is the most used in this way because that
pair is used to point to data following the end of a program. To
use the data in the program it is first necessary for HL to
 Nested Loops and Addressing Modes 35

contain the address of the data as follows

LD HL,NUMBER

NUMBER: DEFB -25

The instruction LD HL,NUMBER causes the address of the data byte

labelled NUMBER to be loaded into the HL register pair.

In register indirect addressing mode, the address of the operand

is contained in a register pair, so that, for example, the
instruction LD A,(HL) causes the accumulator to be loaded with
the contents of the memory byte whose address is in HL.

Figure 7.2 compares the immediate extended and register indirect

addressing modes.

I LD I
~
D HL opcode A, (HL) opcode

34

6Ej
operand
12 H operand
address
L 2B

LD HL,1234H LD A, (HL)

Immediate extended addressing Register indirect addressing

Figure 7.2

The register indirect addressing mode can be used in addition

ADD A,(HL) - and subtraction - SUB (HL). Also, a memory byte may
be loaded from the accumulator using this addressing mode - LD
(HL) ,A.

Program 7.2 shows the use of the immediate extended and register
indirect addressing modes.

The program computes the sum and the difference of two numbers Nl
and N2 which are located in a data area at the end of the
program.

The first instruction in the program sets HL to point to the data

byte N2. With Nl in the accumulator, the SUB (HL) instruction
causes the contents of the memory byte pointed to by HL, to be
subtracted from the accumulator. Similarly, the ADD A,(HL)
instruction causes the contents of N2 to be added to the
accumulator.
36 Programming in Z80 Assembly Language

Program 7.2 sum and difference of 2 numbers

LD HL,N2 ; HL points to N2

LD A,(Nl)
SUB (HL) DIFF Nl - N2
LD (DIFF),A
LD A, (Nl)
ADD A, (HL) SUM Nl + N2
LD (SUM) ,A
HALT
;
Nl: DEFB 14H
N2: DEFB -23H
DIFF: DEFB 0
SUM: DEFB 0

7.3 THE DEFM PSEUDO OPERATOR

A string of characters can be specified in the data area of a

program by using several DEFB pseudo operators, as follows

TEXT: DEFB 'M'

DEFB 'E'
DEFB 's'
DEFB 's'
DEFB 'A'
DEFB 'G'
DEFB 'E'

However, that is rather tedious and unnecessary because

TEXT: DEFM 'MESSAGE'

is exactly equivalent to the seven DEFBs above but is quicker to

write and easier to read and understand. On detecting a DEFM
pseudo operator the assembler puts the first character, M in this
case, in the memory byte labelled TEXT and the remaining
characters in order in the next six bytes.

The pseudo operator DEFM actually stands for DEFine Message.

The DEFM and DEFB pseudo operators can be mixed to produce

strings of characters which contain control codes. For example,
the character string

MESS12: DEFM 'FIRST LINE'

DEFB ODH CR code
DEFB OAR LF code
DEFM 'SECOND LINE'
 Nested Loops and Addressing Modes 37

if output as one string would cause

FIRST LINE
SECOND LINE

to be dis played.

7.4 TEXT OUTPUT

When only one, two or three characters are being output to the
display it is normal to output them with separate sets of
statements, each of which loads the accumulator with a character
and calls COUTo However, for three or more characters this would
be a very inefficient way to output them to the display.

A more efficient method uses a loop to output the message.

Assuming that the message is defined by a DEFM pseudo operator at
the end of the program, the program code performs as follows. The
HL register pair is set to point to the first of the characters,
the character is loaded into the accumulator and output to the
display. The HL register pair is then incremented by one, so
that it is pointing to the next character which is loaded into
the accumulator and output to the display, and so on. Program
7.3 outputs a string of eleven characters to the display.

Program 7.3 output text to the display

LD HL,TEXT ; HL points to first character

LD B,11

NEXCH: LD A,(HL) get next character

CALL COUT
INC HL HL points to next character
DJNZ NEXCH
HALT
,
TEXT : DEFM 'ABCDEFGHIJK'

The INC HL instruction in Program 7.3 is a new instruction for

you to note. In fact all the register pairs, BC, DE, HL, SP, IX
and IY, can be incremented by one with an INC rp instruction and
also decremented by one with a DEC rp. One important difference
between incrementing and decrementing register pairs and
incrementing and decrementing single registers is that none of
the flags is affected by the register pair INC and DEC
instructions. Check this fact for yourself by looking at Table
C.7 in Appendix C.

7.5 THE SUBROUTINE MECHANISM

Having used subroutines we are now going to see how they work,
that is, what happens when a call is made from a main program to
a subroutine and a return is made from that subroutine. The
following description refers to Figure 7.3.
38 Programming in Z80 Assembl y Language

The stack is used by the subroutine mechanism to save temporarily

the return address. When the main program CALL instruction is
executed, the address of the instruction following the CALL
instruction (NEXTI) is automatically pushed on to the stack.
Execution of the subroutine is then started by loading the
program counter (PC) register with the address specified in the
CALL instruction (SUBR) .

,.

CALL SUBR
NEXT!
.... - ,.

.... ....
NEXTr

SUBR

,-
- ....
RET
,. -
program and subroutine stack

PC SUBR after execution of CALL SUBR

PC NEXTI after execution of RET

Figure 7.3

When the subroutine RET instruction is executed, the top of the

stack (NEXTI) is automatically popped off into the PC register
and the instruction at the address pointed to by the PC register
(that is, the one following the CALL instruction - NEXTI) is
executed next.

If a subroutine uses the stack for any reason, then it must

ensure that it has popped off all that it has pushed on, so that
when the RET instruction is executed the top of the stack
contains the return address.

7.6 PROGRAM

Write a subroutine called CHSOUT which outputs a character n

times on m lines on the display. On entry to the subroutine the
 Nested Loops and Addressing Modes 39

accumulator contains the character and m and n are contained in

the Band C registers, respectively.

Write a subroutine called TEXOUT which outputs a string of

characters. On entry to the subroutine HL contains the address
of the first character in the string. The string must be
terminated by a byte containing a zero value.

Using the subroutines CHSOUT and TEXOUT and subroutines from

previous programs, write a program which outputs the following on
the display

******************************
******************************
******************************

your name
your address first line
second line
- third line

******************************
******************************
******************************
As an extra challenge you could output after the last line the
following pattern

****************************
**************************
************************
**********************
********************
******************
****************
**************
************
**********
********
******
****
**
You should use nested loops to provide this additional output,
not just fourteen calls to the TEXOUT subroutine with fourteen
different fixed-character strings.
8 Carry and Overflow

Carry and overflow are two conditions which can occur during
addition and subtraction. The conditions affect bits in the flag
register which can then be tested by conditional jump
instructions.

8.1 CARRY

Carry normally refers to the carry out of the most significant

bit during the addition of two numbers. For example, the sum

00110011 (+51)
+ 00011100 (+28)
--------
01001111 (+79)
--------
does not produce a carry, but the sum

11111110 (-2)
+ 11111111 (-1)
--------
[ 1] 11111101 (-3)
--------
has produced a carry which would be ignored as far as the result
of the 8-bit addition is concerned.

8.2 THE CARRY FLAG

When an ADD instruction is executed the CARRY FLAG will be set to

1 if carry occurs, otherwise it will be reset to O. The carry
flag can then be tested by using one of the conditional jump
instructions

JP C,label
JP NC,label
JR C,label
or JR NC,label

where C stands for Carry (the carry flag is set to 1) and NC

stands for No Carry (the carry flag is reset to 0).

40
 Carry and Overflow 41

The carry flag is also used to indicate that a SUB instruction

needed to borrow a 1 during subtraction of the two most
significant bits.

When the following program segment

SUB 8
JP NC,NCARRY
CARRY:

is executed with 7 in the accumulator, for example, the

instruction labelled CARRY will be executed after the JP
instruction rather than the instruction labelled NCARRY because a
carry would have been produced by the subtraction.

The carry flag is involved in the execution of shift, rotate and

decimal adjust instructions, which are dealt with later in the
book.

There are two instructions which may be used to change the value
of the carry flag. The instructions are SCF which Sets the Carry
Flag to I, and CCF which Complements the Carry Flag. These two
instructions are specified in Table C.6 in Appendix C.

The sequence of instructions

SCF set carry flag to 1

CCF complement i t - now reset to 0

can be used to reset the carry flag to O.

8.3 OVERFLOW

Overflow occurs when the r esult of an operation is outside an

arithmetic range. For 8-bit registers and bytes the range is
-128 to +127.

For addition, two numbers with different signs will never cause
overflow. However, when adding two posit ive numbers or two
negative numbers overflow mayor may not occur. For example, the
sum

01100100 (+100)
+ 00110001 (+49)

10010101 (-107)

does not produce the correct arithmet ic result because the real
sum of the two numbers, +149, is greater than +127 and is
therefore outside the arithmetic range.

For subtraction , overflow can only occur if the two numbers have
different signs. For example, the subtraction
42 Programm ing in Z80 Assembly Language

01111110 (+126)
- 11000000 (-64)

0) 10111110 (-66)

does not produce the correct arithmetic result. The real result,
+190, is outside the range. Additionally, in this particular
example, carry has occurred.

Both carry and overflow can occur together when adding two
numbers. For example, the sum

10010101 (-107)
+ 10010101 ( -107)

[1) 00101010 (+42)

produces overflow and carry.

8.4 THE OVERFLOW FLAG

The parity/overflow flag is used to indicate overflow or parity

depending on the instruction. The flag indicates overflow for
arithmetic operations when it is referred to as the overflow
flag.

The OVERFLOW FLAG is set to 1 if overflow occurs and reset to 0

if overflow does not occur, f o r the ADD, SUB, INC, DEC, NEG and
CP instructions. The flag can be tested using one of the
conditional jump instructions

JP PO,label
or JP PE,label

where PE stands for Overflow (the overflow flag i s set to 1) and

PO stands for No Overflow (the overflow flag is reset to 0).
Unfortunately, PO and PE refer to the usage of the flag as a
parity indicator and, therefore, are rather misleading when
referring to overflow. Notice that there are no JR equivalent
instructions for overflow.

8.5 CONDITIONAL CALLS AND RETS

In addition to the simple CALL subroutine instruction, there are

conditional CALL instructions which operate in a similar manner
to the conditional JP instructions. For example, the program
segment

JP NZ,OVER
CALL SUBEX
OVER:

could be recoded, using the single instruction

 Carry and Overflow 43

CALL Z,SUBEX

thereby saving one instruction.

The conditions which can be tested by the conditional CALL

instructions are the same as those which can be tested by the
conditional JP instructions , that is, Z and NZ, M and P, C and NC
and PE and PO.

There are equivalent conditional RET instructions to return

conditionally from a subroutine. However, these are not normally
used because a subroutine should have only one exit place, that
is, only one unconditional RET instruction. Hence, the
subroutine structure

JP NC,EXIT

JP EXIT

JP Z,EXIT

EXIT: RET

would be used in preference to

RET NC

RET

RET Z

RET
A single exit subroutine provides a cleaner interface which
ensures that all that has to be done before returning is done and
extensions to the subroutines can easily be made without the
possibility of forgetting that exits were made other than at the
end of the subroutine.

The ' ex i t ' label is normally associated with the first of several
instructions ending with the RET i ns t r uc t i on . If, for example,
registers have to be restored then the end of the subroutine may
look like

GOBACK: POP BC
POP DE
RET

8.6 PROGRAM

Amend the subroutine NUMIN so that it inputs a signed, rather

than unsigned, decimal number. The subroutine should then accept
numbers such as -121, +84 and 53 (implied positive number).
44 Programming in Z80 Assembly Language

Write a program which first requests two numbers to be input as

follows

INPUT FIRST NUMBER nl

INPUT SECOND NUMBER n2

and then outputs

SUM IS znorp

where znorp is the word ZERO, NEGATIVE or POSITIVE, followed, if

appropriate, on the next line by

PRODUCED OVERFLOW

followed, if appropriate, on the next line by

PRODUCED CARRY

The two numbers nl and n2 are signed 2-digit decimal numbers, for
example, 63 and -08.

The program should repeatedly deal with pairs of numbers until

E is input as nl.
9 Bit Operations and the Index
Registers

The Z80 has an extensive range of bit instructions . The bit

instructions allow i ndi vi d ua l bits of a register or memory byte
to be tested (for 0 or 1), set to 1 and reset to O. The number
of the bit in the register or memory byte to be used in the
operation has to be specified in the instruction and, for this
purpose, the bits are numbered right to left starting at 0, as
follows

ITIIIIIJJ
7 6 5 4 320

All bit instructions operate on any of one of the single

registers or a memory byte pointed to by HL, IX or IY.

9 .1 THE BIT TEST INSTRUCTION

The BIT instruction tests a spec ified bit of a register or memory

byte and sets the zero flag accordingly - to 1 if the bit is 0,
or to 0 if the bit i s 1. For example, the instruction

BIT 6,E

tests the bit numbered 6 of reg ister E and would set the Z flag
to 0 if register E contained OlOOOlOOB.

Look back at Program 3.2 where you will see a BIT instruction
used to test bit 0 of the accumulator.

It is probably easier to remember that the BIT instruction causes

the Z flag to be set to the complement of the specified bit.

A BIT instruction is usually followed by a Jump on zero' or

'jump on non-zero' instruction, in which case the Zero and Non-
46 Programming in Z80 Assembly Language

Zero conditions refer, as you would expect, to the specified bit

being a or 1, respectively. So there is normally no need to
remember how the Z flag is set by a BIT instruction.

9.2 THE SET AND RES INSTRUCTIONS

The SET instruction allows a specified bit of a register or

memory byte to be set to 1. For example, the instruction

SET 2,C

sets bit 2 of register C to 1. The other bits of C remain

unchanged.

The RES instruction allows a specified bit of a register or

memory byte to be reset to a. For example, the instruction

RES 5, (HL)

resets bit 5 of the memory byte pointed to by HL to O. The other

bits of the memory byte remain unchanged.

The following program segment checks if the number in the

accumulator is odd or even and then sets bit 7 of the B register
to 1 if the number is odd or to a if the number is even, without
affecting the other bits in the B register.

BIT a,A is number odd or even?

JR Z,EVEN
ODD: SET 7,B odd - set bit 7 of B to
JR CONTIN
EVEN: RES 7,B even - reset bit 7 of B to a
CONTIN: -
9.3 THE DEFS PSEUDO OPERATOR

The DEFS pseudo operator allows memory bytes to be reserved,

usually in the data area at the end of the program. Unlike the
DEFB and the DEFM pseudo operators, the DEFS pseudo operator does
not initialise the bytes to a specified value. For example, in
the following program segment

NUM: DEFS 1
BUFFER: DEFS 96
NAME: DEFM 'FRED'

the first DEFS pseudo operator reserves 1 byte which is labelled

NUM, and the second DEFS reserves 96 bytes - the first of which
is labelled BUFFER.

Examples of the use of the DEFS pseudo operator appear in the

remainder of this chapter.
 Bit Operations and the Index Registers 47

9.4 THE INDEX REGISTERS

The 280 microprocessor has two index registers, IX and IY, so-
called because they allow a program to access a particular byte
in a block of bytes, by an index which is relative to the start
of the block.

The following program skeleton shows how an index reg ister can be
used.

LD IX,START ;IX points to start of block

LD A,(IX+2) A = 3rd byte of block

SET 3,(IX+5) Bit 3 of 5th byte = 0

DEC (IX+9) Decrement last byte

START: DEFS 10

A block of ten bytes is defined at the end of the program, using

the DEFS pseudo operator. The first of the bytes is labelled
START. Before any bytes in the block may be accessed the index
register is set to point to this first byte using the LD IX,START
instruction. Thereafter, a particular byte in the block is
referenced by specifying the index register plus the relative
displacement of the byte from the start of the block. For
example, the fourth byte of the block is referenced by (IX+3).

The IY index register can be used in exactly the same way as the
IX index register.

The index registers are useful for referencing bytes within

blocks in which the data is distinct but related; for example, in
a structure in which a file or table contains records of data
items.

However, in the main you will find the index registers being used
with a zero displacement, in which case they are used in much the
same way as the HL register pair.

9.5 EXPRESSIONS

Operands that we have used so far have consisted of a single

label. In fact, the 280 assembler allows a wide range of
expressions to be written as an operand - a single label being
the simplest example of an expression. For example, the
instruction

LD A, (NUM+1)

would cause the contents of the byte following the byte labelled
NUM to be loaded into the accumulator.
48 Programming in Z80 Assembl y Language

When an expression is used it is normally a simple one, as in the

LD instruction above. However, an expression may be quite complex
as a result of using the many operators which can be used in an
expression. The full list of the twenty available operators is
shown in Appendix E.

All expressions are evaluated left to right, except that unary

operators are performed first, exponentiation next,
multiplication, division, modulo and shifts next, followed by
addition and subtraction, and then logical operations and
comparisons. 16-bit signed arithmetic is used throughout the
evaluation of an expression.

The following program skeleton shows one use of expressions

ITEMS: EQU m
LENTH: EQU n

TABLE: DEFS ITEMS*LENTH

The characteristics (number of items in the table and the number

of bytes in each item) of the TABLE need to be varied from one
version of the program to another. Thus the TABLE is defined in
terms of its characteristics which are given values, m and n, at
the beginning of the program (using the EQU pseudo operator) for
a particular version of the program.

9.6 JUMP TABLES

A jump table is a convenient way of executing one of several

program segments depending on the value of a variable which has
sequential values 1 to N.
The basis of the technique is as follows

rp contains JPTAB + 3 x (N - 1)
JP ( rp)

JPTAB: JP N1CODE start of jump table

JP N2CODE

JP NNCODE end of jump table

Before the jump table can be used, one of·the IX, IY or HL

register pairs must contain the address of one of the JP
instructions in the jump table. A JP (rp) instruction is then
executed which causes one of the JP instructions in the jump
table to be executed.

Program 9.1 shows the use of a jump table in displaying the name
of a day whose number is input from the keyboard.
 Bit Operations and the Index Registers 49

Program 9.1 input day number, output day name

CALL CIN ; i nput day number

SUB 30H

DEC A day number - 1

LD C,A
ADD A,C
ADD A,C 3 x (day number - 1)
LD C,A
LD B,O j um p table off s e t now in BC

LD HL, DAYJ P HL co nt a ins start of jump t ab le

ADD HL,BC + offs et
JP (HL)

DAYJP: JP MON ;jump t able

JP TUE
JP WED
JP THU
JP FRI
JP SAT
JP SUN

MON: LD HL,MONDAY
CALL TEXOUT
JP FINI

SUN: LD HL,SUNDAY
CALL TEXOUT
JP FINI

FINI: HALT

MONDAY: DEFM ' MONDAY'

DEFB 0

SUNDAY : DEFM 'SUNDAY'

DEFB 0

The day number, which i s i nput f r om the keyboard, is i n the range

1 to 7, repres enting t he days Monday to Sunday. The a ddres s i n
the jump table of the corresponding JP instructions is ca l c ul a t ed
i n the HL register pair, and execution of that JP instruction
follows execution of the JP (HL) instruction. Hence, a j um p is
made to the program segment which outputs the name of the day
corresponding to the number which was input.

Becaus e t he seven program segments are the same length (that is ,

the y occupy the same number of byt es), an algorithm cou ld hav e
50 Programming in Z80 Assembly Language

been used to compute the start address of the corresponding

program segment, rather than using a jump table. However, the
jump table technique makes the code more intelligible and, in
fact, has to be used when the program segments are of unequal
lengths.

9.7 PROGRAM

Write a program which repeatedly inputs a number in the range

to 12 and outputs the corresponding name of the month.

Instead of using a 'jump table' technique to solve this problem,

use a 'look-up table' technique as follows: set up two data areas
at the end of your program, one in which the names of the months
(terminated by a byte containing 0) are stored end-on in a
contiguous data area (occupying 86 bytes), and a second 12-byte
data area in which the nth byte contains the position of the
start of the nth month's name in the months' names data area
(that is, the first of the 12 bytes will contain 0, the second 8,
the third 17, and so on); to obtain the start address of a
month's name , look-up the corresponding entry in the 12-byte data
area and add it to the start address of the months' names data
area.
10 Shift Instructions, Multiply and
Divide

Shift instructions allow the bits of a register or memory byte to

be shifted one bit place to the left or to the right. There are
two types of shift instructions - logical and arithmetic.
Logical shifts consider the contents of the register or memory
byte to be just a bit pattern when the shift is made. Arithmetic
shifts consider the contents of the register or memory byte to be
a signed number so that when the shift is made the number is
arithmetically multiplied by two (left shift) or divided by two
(right shift). The Z80 microprocessor has one logical
instruction and two arithmetic instructions. Descriptions of the
shift instructions are included in Table C.8 in Appendix C.

10.1 THE SRL INSTRUCTION

The Shift Right Logical instruction shifts a register or memory

byte one bit place to the right. Bit 7 of the register or memory
byte is reset to 0 and the original bit 0 goes into the carry
flag, as follows

o---OIIIIID-D
1 6 5 4 3 2 o carry
flag
register or memory byte

The form of the instruction is

SRL m

where m is any of the single registers or a memory byte pointed

to by HL, IX or IY.

As an example, suppose that the contents of the accumulator and

carry flag are

accumulator 1 0 100 1 1 1 carry flag o

then after the instruction SRL A is executed they will contain

accumulator 010 100 1 1 carry flag

51
52 Programming in Z80 Assembly Language

In all shift instructions, the bit which is moved out of the

register or memory byte, be it to the left or right, is placed in
the carry flag. This can be useful because the value of a bit
which has been moved out can be checked by any o f the carry
conditional jumps, such as JP C, l a be l and JR NC,label.

10.2 THE SRA INSTRUCTION

The Shift Right Arithmetic instruction is the same as the SRL

instruction except that, i ns t ea d of bit 7 being reset to 0 it is
s et to what it was before the shift. In other words, the sign
bit remains unchanged so that a positive value would remain
positive and a negative value would remain neg ative. The SRA
instruction, in fact, divides the register or memory byte by two,
a nd leaves the rema ind er in the carry flag .

The form of th~ instruc tion i s

SRA m

where m is any of t he single registers or a memory byte point ed

to by HL , I X or I Y.

The following program segment shows the effect of the SRA

ins t r uc t i on on both posit ive and negative numbers

B C Car r y

LD B , ll 000 01011B (+11) ?

SRA B 00000101B (+5) ? 1
LD C,-8 " 1111l000B (-8) 1
SRA C " 1111 1100B ( - 4) 0

10.3 THE SLA INSTRUCTION

The Shift Left Arithmetic instruction shifts a r egister or memory

byte one bit place to the left. In doing so, bit 0 of the
r egister or memory byte i s reset to 0 and bit 7 is placed in the
carry f l a g , as follows

D- DIDITIJ-o
ca rry 765 4 3 2 o
flag
register or memory byte

The ~ ffect of the instruction is to multiply the cont ents of the

register or memory byte by two.
 Shih Instr uct ions. M ultiply and Divide 53

The usual d i fference b e t we e n an arithmetic left shift and a

logical left shif t is tha t the overflow flag would be set
accordingly for an arithmetic left shift but not for a logical
left shi ft . Overflow would occur, for e xample, during an
arithmet ic l eft sh i ft when the register contained 127 since, in
signed number a r i t hme t i c , multiplying 127 by 2 would c a u s e
o ve rflow in an 8-bi t register .

As the shift left a rithmetic instruction does not set the

overflow flag , it should more precisely be called a shift left
logical instruction.

If the con tents of the register or memory by te to be shifted left

are considered to be a n unsigned number in the r ange a to 255,
t hen the c arry f l a g would indicat e overfl ow.

The form o f t he instr uct ion is

SLA m

where m is anyone of the singl e r egisters, or a memory byte

pointed t o by HL, IX or 1Y.

The f ollowing program segment shows the use of the SLA

instruction in multiplying the contents of t h e accumulator by ten

SLA A 2 x A
LD B,A
SLA A 2 x (2 x A) 4 x A
SLA A 2 x (4 x A) 8 x A
ADD A,B 8 x A + 2 x A 10 x A

10 .4 8-BIT MULTI PLI CATI ON AND DIVISION

The multiplication method of repeat ed a d d i t i on which we have been

using is very inef ficien t for mult ipliers grea t er than fiv e, or
so. A more efficient metho d for larger multiplier s is called
'shift and add'. To und erstand the reason for using ~his method
work throug h t he fol lowing multiplication

00111 mul tiplicand

x 01010 multiplier

00000 x a
00 111 x 10
00000 x 000
0011 1 x 1000
000 00 x 00000

001 00011 0 product

Each bit in the multiplier causes a one bit shift to t h e l eft of

54 Programming in Z80 Assembly Language

the multiplicand which is then added into the product. Also, it

can be seen that the value to be added in each time is either the
multiplicand or zero.

Hence, the algorithm for multiplication by 'shift and add' is:

for each bit of the multiplier, working from right to left, the
multiplicand is added to the partial product if the multiplier
bit is 1, otherwise nothing is done : The multiplicand is then
shifted left one bit place before the next bit of the multiplier
is considered.

Program 10.1 shows a program segment which multiplies the

contents of the Band C registers, using the 'shift and add'
method. The product is accumulated in the accumulator.

Program 10.1 computes A =B x C

LD A,O
LD D,l
NEXBIT: SRL C test next bit of multiplier
JP NC,NOADD
ADD A,B add in multiplicand
JR PO,OVEKFL
NOADD: DEC D
JR Z,DONE
SLA B shift multiplicand
JR NEXBIT

The program segment deals only with , pos i t i ve signed numbers . It

would have to be extended to deal with negative signed numbers.
One straightforward method of doing this is first to mult iply the
absolute values of the two numbers and then compute the sign of
the product according to the signs of the numbers.

The operation of Program 10.1 is as follows : the next multiplier

bit is tested for 0 or 1 by shifting the bit i nt o the carry flag;
using a carry flag conditional jump instruction the multiplicand
is added in, or not; the multiplicand is then shifted left one
bit place ready to be added in, or not, when the next multiplier
bit is tested. A check for an overflow cond ition is made when
the multiplicand is added in.

Division can be performed by 'repeated subtraction' (equivalent

to multiplication by 'repeated addition ') or by a method which is
similar to the standard pencil and paper approach, in which a
check is made to see if the divisor goes into the remaining
d ividend (equivalent, in some respects, to multiplication by
' s h i f t and add' and could, in fact, be referred to as the 'shift
and subtract' technique).
 Shift Instruction s. Multiply and Divide 55

10.5 PROGRAM

Write a new MULT subroutine which uses the 'shift and add' method
of multiplication and additionally deals with signed numbers.

Write a subroutine called DIV which divides a signed number in

register B by a signed number in register C and leaves the result
(quotient) in the accumulator and the remainder in regist er D.

Using the DIV subroutine to divide by ten, write a subroutine

called NUMOUT which outputs the contents of the accumulator as a
signed number in the range -128 to +127. Leading zeros should
not be output and positive numbers should be output without a +
character.

Using the NUMOUT, DIV and MULT subroutines and subroutines from
previous programs, write a main program which inputs two signed
numbers separated by either an * character or a / character and
followed by an = character. The program should then output
either the product of the number (if the * character was input),
or the quotient and remainder ( if the / character was input). A
typical dialogue would look like

-11*5=-55
125/10=12 REMAINDER 5
11 Logical Operations and Macros

11.1 LOGIC OPERATORS

There are several Z80 logical instructions which allow logical

operations to be performed between corresponding bits in the
accumulator and an 8-bit operand.

To understand the operation of the logical instructions, it is

necessary to know the rules of the basic logical operators, such
as AND, OR, XOR (Exclusive OR) and NOT. The rules are normally
displayed in tables as follows

o AND 0 o o OR 0 o o XOR 0 o NOT 0 1

o AND 1 o o OR 1 1 o XOR 1 1 NOT 1 o
1 AND 0 o 1 OR 0 1 1 XOR 0 1
1 AND 1 1 1 OR 1 1 1 XOR 1 o
Apart from the NOT operator, logical operators operate on two
one-bit values and produce a one-bit result. For example, the
result of the AND operator is one if, and only if, the two values
are one, otherwise, the result is zero.

Glance at Table C.5 in Appendix C where you will see the AND, OR
and XOR instructions specified. Notice that the Symbolic
Operation column of that table uses the Boolean Algebra symbols
of A for AND, v for OR and $ for XOR.

11.2 LOGICAL INSTRUCTIONS

All the Z80 logical instructions perform a logical operation

between the bits in the accumulator and their corresponding bits
in the operand, leaving the result in the accumulator. For
example, if the accumulator contains 00001010B and register B
contains llOOII11B then the instruction AND B would produce a
result of 00001010B in the accumulator, as follows

Contents of A OOOl1010B
Contents of B llOOllllB
Contents of A after AND B 00001010B

Corresponding bits in the accumulator and operand are logically

operated upon in isolation from the other bits.

56
 Logi cal Operations and Macro s 57

The operand of a logical instruction may be a register, an 8-bit

value or a memory byte pointed to by HL, IX or IY. The logical
instructions may be specified together as

AND r
n
OR (HL)
(IX)
XOR (IY)

Notice that the accumulator is implied in the instructions since

logical operations may be performed only on the accumulator.

The NOT logical operation is performed by the CPL instruction

which changes all the O's in the accumulator to l's and all the
l's to O's. The mnemonic CPL stands for the word 'ComPLement'.

The following program segment shows the effect of the four Z80
logical instructions on the accumulator and the S, Z and C flags

A S Z C

LD A,10110101B 10110101B ? ? ?
LD C,l1110000B 10110101B ? ?
AND 00011111B 00010101B 0 0 0
OR C 1111 0101B 1 0 0
XOR 11001100B 00111001B 0 0 0
CPL 11000110B 0 0 0

11.3 MASKING

One of the main uses of the AND logical instruction is in masking

when i t is required t o us e only some of the bits of an 8-bit
value and it is necessary to mask the requ ired bits. For example,
masking could be used instead of subtraction to obtain a decimal
digit's value from its character code, because the least
significant four bits of a decimal digit's character code are its
value and can be obtained by ANDing the character code with
00001111B, as follows

LD A, '7' accumulator containing 00110111B

AND OFH ANDed with 00001111B
gives a value of 00000111B

Conversely, the logical instruction OR 30H will convert a decimal

digit value in the accumulator to its character code.

11.4 MACROS

Macros provide a means for a programmer to define his own

opcodes. Suppose that there is a frequent need to reset the
carry flag to 0 in a program, then an opcode to do that could be
58 Programm ing in Z80 Assembly Language

defined at the beginning of the program with the following macro

definition

RSF: MACRO reset carry flag macro

SCF * macro body
CCF
ENDM
*end of macro definition

The MACRO pseudo operator informs the assembler that a macro

called RSF is to be defined. The code between MACRO and ENDM
(the macro-end pseudo operator) is called the macro body.

Now, at any place in the program the opcode RSF can be used - or,
as is usually said, 'the macro RSF can be called'. For example,
the RSF macro is called in the following program segment

LD (POINT) ,A
RSF reset carry flag macro
ADD A,B

During assembly the program segment will be expanded, according

to the macro definition, to

LD (POINT) ,A
SCF macro RSF
CCF expansion
ADD A,B

Another example of a macro is the MUL4 macro

MUL4: MACRO ; macro multiplies A by 4

SLA A
SLA A
ENDM

which multiplies the con t en t s of the accumulator by four.

A growing use of macros is i n the attempt to g ive assembly

language programs some structure (similar to the structuring
facilities of high-level languages) for constructs which are
often used. Program 11.1 shows how macros can be us ed to give
some structuring to loops.

Two macros are defined - one for the start of the loop, LOOPST,
and one for the end of the loop, LOOPEN. Unlike the previous RSF
macro, both of these macros have parameters. Any parameters used
in a macro definition must be listed after the MACRO pseudo
operator. The parameters may then by used anywhere in the macro
body. When a call is made to a macro with parameters, during
expansion of the macro the parameters are replaced by the actual
parameters in the macro call. The last macro call in Program
11.1 would be expanded to

DEC E
JP NZ,ROUND
 Logical Operation s and M acros 59

No tice that parameters in a macro definition are preceded by the

# character and parameters in a macro call are enclosed with
apostrophes .

; Prog ram 1 1. 1 structured loops using ma c r os

LOOPST: MACRO #R,#N ; start of loop macro

LD i/R, i/N
ENDM

LOOPEN : MACRO #R,# LABEL end of loop macro

DEC #R
JP NZ,i/LABEL
ENDM

LOOPST 'c ' ,' 99'

NEXTHG: -
LOOPEN 'c' , ' NEXTHG'

LOOPST 'E' " 18'

ROUND:

LOOPEN 'E ' ,'ROUND'

Macro definitions may not be nested, but a previously defined

macro may be called from another macro body .

In order to be able to use labels which are local to a macro body

a spec i al symbol g enerator fa ci li t y is provided. Th is f a cilit y
is best explained by way of an example . Conside r the macro
definition

TIMER: MACRO i/N

LD B, i/N
TMil $YM: DJNZ TMil$YM
ENDM

which provides a crude timing facility by counting down register

B from N to zero. A local label is required for the loop, but if
this label were to be chosen in the normal way as, say TMLOOP,
then a program which called the TIMER macro more than once would
caus e multiple definitions of the TMLOOP label to be produced
during macro expans ion of t he program. To overcome this problem,
local labels a r e defined with the last fo ur characters as #$YM
and dur ing the expansion of macros, #$YM is repl aced by a 4-digit
hexadecimal number s t a r t i ng at 0000 for the first macro call and
increment ing by one for ea c h subsequent macro call .
60 Programming in Z80 Assembly Language

For example, a program containing three calls on the TIMER

macro, as follows

TIMER '200'

TIMER '50'

TIMER '76'

would be expanded as

LD B,200
TMOOOO: DJNZ TMOOOO

LD B,50
TMOOOI : DJNZ TMOOOI

LD B,76
TMOO02: DJNZ TMOO02

The main difference between macros and subroutines is that

instructions in a macro body are repeated everywhere that a macro
is called, whereas subroutine instructions occur once only. In
general, macros should not be used when the macro body consists
of more than just a few instructions.

The format for defining and calling macros does, unfortunately,

vary from one zao assembly language system to another . You should
obtain the correct format for your system from the macro section
in your zao assembly language manual.

11.5 CONDITIONAL PSEUDO OPERATIONS

Conditional pseudo operations provide the capability to include

conditionally parts of a program during assembly time. The two
conditional pseudo operators are COND and ENDC, and they are used
as follows

COND CODE IN

ENDC

The instructions between the COND and ENDC pseudo operators will
be included during assembly if the value of CODEIN is non-zero,
but ignored by the assembler if CODEIN is zero.
 Logical Operations and Ma cros 61

The DEFL - DEFine Label - pseudo operator is normally used to set

the value of the COND pseudo operator label to zero or non-zero.
For example, if the statement
CODEIN: DEFL 1

had appeared before the COND pseudo operator, in the above

example, the instruction between the COND and ENDC pseudo
operators would be included in the assembly.

Although the condition usually depends on the value of a label,

an arithmetic or logical expression may be used.

Let us look at a typical example of conditional assembly. Suppose

a program needs to be assembled to produce output to a displ~y or
a printer, but not both. The program will have two subroutines
with the same name - one subroutine for outputting to the
display, the other for outputting to the printer. An outline of
the program would look like

DORP: DEFL n n is 1 for display output

n is a for printer output

COND DORP
aUTLIN: - display output subroutine

RET
ENDC

COND .NOT. DORP

OUTLIN: - printer output subroutine

RET
ENDC

When the program is to be used for display output only the n in

the DORP instruction will be set to 1, so that when the program
is assembled the display output subroutine will be included and
the printer output subroutine will not be included.

Alternatively, if n is set to a the display output subroutine

will not be included but the printer output subroutine will be
included.

The DEFL pseudo operator is similar to the EQU pseudo operator.

Both operators allow a label to be given a value but, whereas the
EQU pseudo operator can appear only once for any label in a
program, the DEFL pseudo operator can be repeated in the same
program to give different values to a label. This extra facility
of the DEFL pseudo operator allows more flexible conditional
assemblies to be constructed.

The conditional pseudo operations are very useful for tailoring a

general program for a particular requirement.
62 Programming in Z80 Assembly Language

11.6 PROGRAM

Two other common logical operations are NOR and NAND, which are
short for NOT OR and NOT AND, respectively. The rules of these
two logical operations are

o NOR 0 1 o NAND 0 1
o NOR 1 o o NAND 1 1
1 NOR 0 o 1 NAND 0 1
1 NOR 1 o 1 NAND 1 o
from which it can be seen that the result of the NOR operation is
the complement, or NOT, of the OR result and, similarly, the
result of the NAND operation is the NOT of the AND result.

Write a Logic Operation Trainer program which repeatedly allows

input of the form

b lop b =b
where b is 0 or 1
and lop is one of OR (space to be input after the R)
AND
XOR
NOR
or NAN (short for NAND)

and outputs further along the same line, TRUE if the input is
correct and otherwise FALSE.
A tra ining session is ended by inpu~ting the letter E.
12 Rotate Instructions and Parity

Rotate instructions are similar to shift instructions except that

the bit shifted out of one end is shifted into the other end
hence, the name rotate.

There are several r otate instructions availabl e i n the zao

microprocessor. Four of the rotate instructions involve the
accumulator; the other rotate instructions involve a register or
memory byte.

Some of the rotate instructions include the carry flag within the
rotation, whil e others, referred to as rotate circular
instructions, do not.

In line with the shift instructions, rotates may be left or

right and are for one bit place only. All rotates are a logical
type of shift.

The definition of all the rotate instructions is included in

Table c.a in Appendix C.

12.1 ACCUMUl.ATOR ROTATE INSTRUCTIONS

The accumulator can be rotated left and right, and for each
direction the carry flag may, or may not, be included within the
rotate, giving the four instructions

RLA rotate left accumulator,

RRA rotate right acccumulator,
RLCA rotate left circular accumulator,
and RRCA rotate right circular accumulator.

The Rotate Left Accumulator instruction includes the carry flag

within the rotate and operates as follows

co.
carry
flag
ITIIIIITJJ
7 6 5 4 321
accumulator
0

63
64 Programming in Z80 Assembly Language

The contents of the accumulator move to the left one bit place,
and in doing so, bit 7 of the accumulator moves into the carry
flag and the carry flag moves round into bit 0 of the
accumulator.

The Rotate Right Accumulator instruction also includes the carry

flag within the rotate and operates as follows

[ITIIJIIJJ-.Q]
165 4 3 2 o carry
accumulator flag

The contents of the accumulator move to the right one bit place,
and in doing so, bit 0 of the accumulator moves into the carry
flag and the carry flag moves round into bit 7 of the
accumulator.

The Rotate Left Circular Accumulator instruction is a circular

version of a rotate and does not, therefore, include the carry
flag within the rotate. However, the carry flag is still
involved as follows

carry 165 4 3 2 o
flag accumulator
The contents of the accumulator move to the left one bit place,
and in doing so, bit 7 of the accumulator is moved to the carry
flag and around into bit 0 of the accumulator. So after a RLCA
instruction the carry flag and bit 0 of the accumulator will
always be the same value - the value of bit 7 of the accumulator
prior to execution of the instruction .

The Rotate Right Circular Accumulator instruction is also a

circular type of rotate and operates as follows

LITIIJ;J;;-Lo
165 4 321
accumulator
0 carry
flag

The contents of the accumulator move to the right one bit place,
and in doing so, bit 0 of the accumulator is moved into the carry
flag and also around into bit 7 of the accumulator .
 Rotate Instructions and Parity 65

The effect of the four rotate instructions is shown in the

following sequence of instructions

A Carry

10101011B o
RLA 01010110B 1
RLCA 10101100B o
RRA 01010110B o
RRCA 00101011B o

12.2 REGISTER AND MEMORY BYTE ROTATE INSTRUCTIONS

The four types of rotate of the accumulator may be applied to any

of the single registers or a memory byte, using the four
instructions

RL m rotate l eft register or memory byte,

RR m rotate right register or memory byte,
RLC m rotate left circular register or memory byte,
and RRC m rotate right circular register or memory byte,

where m is anyone of the single registers or a memory byte

pointed to by HL, IX or IY.

The operation of the i ns t r uc t i ons is exactly the same as their

corresponding accumulator rotate instructions. However, by
looking at Table C.B in Appendix C you can see that the
accumulator rotate instructions each occupy one byte and set only
the carry flag, whereas the general register and memory byte
rotate instructions each occupy two or four bytes and set the
carry, zero and sign flags.

Hence, the RLCA instruction would be used in preference to the

RLC A instruction because it operates faster and uses fewer
bytes.

Because a rotate accumulator instruction occupies only one byte

whereas a shift accumulator instruction occupies two bytes, some
programmers use the rotate accumulator instruction to perform a
shift operation, instead of the shift accumulator instruction.
However, in saving a byte, and time of execution of the
instruction, the program becomes less intelligible because what
is actually a shift operation appears in the program as a rotate
operation.

12.3 PACKING AND UNPACKING

The term packing refers to having a register or memory byte

containing two or more distinct values. We say that the values
are packed into the register or memory byte. For example, a
66 Programm ing in Z80 Assembly Language

memory byte could be packed with the sex and age of a person, as
follows

ffiEEEEEJ
7 6 5 4 321 0

sex age

Bit 7 of the memory byte indicates the sex of the person, say 0
for female and 1 for male. Bits 6 to 0 of the memory byte
contain the person's age, giving a range for the age of 0 to 127.
The example above specifies a male aged 100 years.

Assuming that the sex and age of one thousand people was held in
memory then there is a saving of one thousand bytes by including
the sex bit in the byte containing the age. If the sex and age
were contained in separate bytes, two thousand bytes would be
needed to hold the same information.

To be able to use the packed values it is usually necessary to

unpack them. For example, to use the sex and age as packed above
it would be necessary to unpack the sex bit into one register and
the age bits into another register.

In general, packing is performed by ORing and shifting or

rotating, while unpacking is performed by ANDing and shifting or
rotating. For example, the program segment

LD A,(SEXAGE)
AND 10000000B
RLCA
LD B,A
LD A,(SEXAGE)
AND OlllllllB
LD C,A

unpacks a sex/age memory byte by placing the sex in the B

register and the age in the C register. There are many ways of
unpacking bytes, some using fewer instructions than others, and
some more obvious than others. The method of unpacking also
depends on the number, length and position of values that are
packed.

The following program segment packs into a memory byte labelled

SEXAGE, sex from the B register (bit 0) and age from the C
register

LD A,B get sex

RRCA
OR C add age
LD (SEXAGE),A
 Rotate Instructions and Parity 67

12.4 PARITY

Parity refers to the number of l's in a binary number. A binary

number is said to have even parity if the number of l's is even,
and odd parity if the number of l's is odd. For example,

01101000 has odd parity,

11111100 has even parity,
and 01110011 has odd parity.

Parity is often used in computer systems when transferring

information from one unit to another. A parity bit is added to
the information at source to make the number of 1 bits either odd
or even. At the destination, the parity of the information is
checked to make sure that it is still either odd or even. It is
not a foolproof check because the parity would remain correct if
an even number of bits changed their values during transfer.
However, a parity check is a great deal better than nothing.

Assuming an even parity system with bit 7 of a byte as the parity

bit, the hexadecimal value of a byte containing the ASCII
character code for X will be D8H, and for + will be 2BH.

12.5 THE PARITY FLAG

The parity/overflow flag is used to indicate the parity of a

result after most of the rotate instructions and the shift and
logical instructions.

If the number of l's in the register or memory byte is even after

any of these instructions has executed, the parity flag will be
set to I, but if the number of l's is odd the parity flag will be
reset to O.

The following program segment shows how the parity flag is

affected by the AND, SLA and RLA instructions

A Parity

10111001B ?
AND 11111110B 10111000B 1
SLA A 01110000B o
RLA 11100001B 1

The parity condition can be tested by any of the instructions

JP PE,label
JP PO,label
CALL PE,label
CALL PO,label
RET PE
and RET PO
68 Programmi ng in Z80 Assembly Language

The mnemonics PE and PO stand for Parity Even and Parity Odd,
respectively.

The following subroutine called CHKPAR checks the parity of the

accumulator. On entry to the subroutine register B contains
either 0 to indicate an even parity check, or 1 to indicate an
odd parity check. On exit from the subroutine the register C
should contain 0 if the parity was correct and 1 otherwise.

CHKPAR: BIT O,B parity check subroutine

JP Z,EVTEST
AND OFFH check for ODD parity
JP PO,OK
JP NOTOK
EVTEST: AND OFFH check for EVEN parity
JP PE,OK
NOTOK: LD C,l not OK
JP RETSUB
OK: LD C,O OK
RETSUB: RET

12.6 PROGRAM

Write a subroutine called BINOUT which outputs on the display the

contents of the accumulator as eight binary digits. The
subroutine will have to extract each bit separately from the
accumulator and output the character code for 1 or 0 depending on
the value of a bit.

Write a subroutine called PACK which packs the accumulator from

two consecutive memory bytes as follows

bit 0 of M to bit 7 of the accumulator,

and bits 6 to 0 of M+1 to bits 6 to 0 of the accumulator,

where M is the first memory byte.

The address of the first of the two bytes is contained in the HL

register pair on entry to the subroutine.

Write a subroutine called UNPACK which does the reverse of the

PACK subroutine.

Using the subroutines BINOUT, PACK and UNPACK and existing

subroutines, write a main program which inputs ten sets of two
numbers and packs each set as specified in the PACK subroutine
specification. The two numbers are input as decimal numbers 0 or
1 and 0 to 127, respectively. The ten sets of packed numbers
should be stored in ten consecutive bytes. After inputting the
ten sets of numbers, the program should output each of the packed
sets in binary for checking purposes.
13 16-Bit and Multiple Byte
Arithmetic

So far we have been concerned with 8-bit arithmetic - tha~ is, 8-

bit operands and 8-bit results. The range of numeric values
which can be manipulated by 8-bit arithmetic is small and so we
may need to use 16-bit ar ithmetic and more. The Z8G
microprocessor has instructions which allow 16-bit arithmetic to
be performed directly. These instructions can also be . used to
provide 32-bit arithmetic, 48-bit arithmetic and so on. The 16-
bit arithmetic instructions also allow additional loop facilities.

All the 16-bit arithmetic instructions are described in Table C.7

in Appendix C.

13.1 THE DEFW PSEUDO OPERATOR

When working with 8-bit quant ities we used the DEFB pseudo
operator when it was necessary to initialise memory bytes at the
end of ou r programs. When working with 16-bit quantities we will
need to initialise double bytes, or words as they are called. To
do this we use the DEFW (DEFine Word) pseudo operator. For
example,

DOUBLE: DEFW 56ABH

initialises DOUBLE to the value 56ABH. The label DOUBLE is

associated with the first of the two bytes and, in fact, ABH is
placed in the first of the two bytes and 56H in the second byte.
This unusual ordering is necessary to complement the instructions
which load register pairs with two consecutive bytes. For
example, the instruction LD HL,(LABEL) loads register L with the
first memory byte (LABEL) and register H with the s~cond memory
byte (LABEL+1) •

In fact, the DEFW pseudo operation saves us having to concern

ourselves with this reverse ordering. The LD HL,(LABEL)
instruction in

LD HL, (LABEL)

LABEL: DEFW 7B9Atl

will cause HL to be loaded with 7B9AH as you would expect.

69
70 Programming in Z80 Assembly Language

13.2 THE 16-BIT ADD, ADC AND SBC INSTRUCTIONS

The main 16-bit arithmetic instructions use the HL register pair

as the 'accumulator', For example, the 16-bit addition
instruction

ADD HL,DE

adds the contents of register pair DE to HL, leaving the result

in HL. The general form of the instruction is

ADD HL,ss

where ss is anyone of the register pairs BC, DE, HL or SP.

The program segment

LD BC,2054
LD HL,1362
ADD HL,BC

shows two 16-bit numbers being added together in the HL register

pair. After execution of the ADD HL,BC instruction HL will
contain the value 3416.

Remember that the range of signed numbers that can be dealt with
by 16-bit arithmetic is -32708 to +32767.

Another 16-bit add instruction adds the carry produced by a

previous operation. The general form of the 16-bit ADd with
Carry instruction is
ADC HL,ss

where ss is the same as for the 16-bit ADD instruction. The

contents of the register pair ss are added to HL along with the
carry flag and the result is placed in HL.

The ADC HL,ss instruction can be used to provide a simple 32-bit

arithmetic facility. Look at Program 13.1 which shows a program
segment which adds two 32-bit signed numbers.

First, the least significant 16 bits of the two 32-bit numbers

are added using the ADD HL,DE instruction, and stored in the
result bytes. During this addition any carry out of the addition
of the most significant bit will be recorded in the carry flag.

Second," the most significant 16 bits of the numbers are added

using the ADC HL,DE instruction, which will cause the carry flag
to be added also - that is, the carry from the addition of the
least significant 16 bits of the numbers.
 16-Bit and Multiple Byte Arithmetic 71

Program 13.1 addition of two 32-bit signed numbers

LD HL,(NiLS)
LD DE,(N2LS)
ADD HL,DE add least significant 16-bits
LD (RESLS),HL
LD HL,(N1MS)
LD DE,(N2MS)
ADC HL ,DE add most significant 16-bits
LD (RESMS),HL
JP PO,OVER}' overflow occurred?

N1MS: DEFW OSA1H first number

NiLS: DEFW 63B2H
N2MS: DEFW OOC6H second number
N2LS: DEFW OAS7EH
RESMS: DEFW 0 result
RESLS: DEFW 0

Hence, the overall affect of the program segment is to add two

32-bit signed numbers. Overflow is detected by checking the
overflow flag after execution of the ADC instruction.

After execution of the program, RESMS will contain 0668H and

RESLS will contain 0930H.

The technique i n Program 13.1 can be extended to add multiple 16-

bit numbers.
There is only one 16-bit subtract instruction - a 'subtract with
carry ' version. The general form of the 16-bit sustrac t with
Carry is

SBC HL,ss

where ss is anyone of the register pairs BC, DE, HL and SP - the

same as for the 16-bit add instructions.

The instruction causes the contents of the register pair ss and

the carry flag both to be subtracted from HL, the result being
left in the HL register pair.

The SBC HL,ss instruction can be used, in a similar way to the

ADC HL,ss instruction, to perform subtractions with multiple 16-
bit numbers.

To perform subtraction with s ingle 16-bit numbers, it is

necessary to set the carry flag to 0 just prior to execution of
the SBC HL,ss instruction to ensure that nothing other than zero
is subtracted from the true result.

The following program segment subtracts the contents of t he

72 Programming in Z80 Assembly Language

register pair BC from HL, assuming single 16-bit arithmetic.

SCF reset carry flag to 0

CCF
SBC HL,BC

One important point to note about the 16-bit ADD, ADC and SBC
instructions is the flag setting of these instructions. Look at
the specifications of these instructions in Table C.7 of Appendix
C and you will see that the 16-bit ADC and SBC instructions set
the carry, zero, overflow and sign flags as you would expect,
but the 16-bit ADD instruction causes only the carry flag to be
set. A 16-bit ADC instruction, preceded by an instruction to set
the carry flag to zero, can be used if the setting of the zero,
overflow and sign flags are required during a single 16-bit
addition.

13.3 EXTENDED LOOPS

The 16-bit load, increment/decrement and arithmetic instructions

which we have already considered can be used for loops in which
the loop index has a range of 0 to 65535. However, there are
special 16-bit instructions involving the IX and IY index
register pairs which can be used for such loops.

The IX index register pair is first loaded with an initial value,

either directly using an LD IX,nn instruction, or indirectly
using an LD IX,(LABEL) instruction. At the end of the set of
instructions to be repeated, the index register is incremented by
one using an INC IX instruction, or decremented by one using a
DEC IX instruction. A jump is then made back to the first of the
instructions to be repeated.

The loop must be terminated either by a condition occurring

within the loop or by IX becoming a specific value. However, it
must be remembered that the INC IX and DEC IX instructions do not
affect any of the flags.

The IY index register can be used wherever the IX index register

can be used.

There is a special ADD instruction relating to the index

registers which, amongst other things, allows the registers to be
incremented and decremented by a value other than one, when used
as the loop index register. The instructions are
 16-Bit and Mu ltiple Byte Arithme tic 73

ADD IX,pp

where pp is anyone of the regis ter pai rs BC s . DE, IX and SP and

ADD IY,rr

where rr is anyone of the register pairs BC, DE, IY and SP. The
instructions cause the contents of the register pair pp and rr to
be added to IX and IY, respectively.

The use of the ADD IY,rr instruction is shown in Program 13.2

which outputs the numbers 1000 to a in decrements of 5.

Program 13.2 outputs numbers 1000, 995, 990, 5, a

LD DE,-5 DE contains decrement
LD IY,1000 IY contains first number

NEXNUM: CALL IYOUT

CALL CRLF

ADD IY,DE decrement IY

check if IY is a
PUSH IY move IY to BC
POP BC
LD A,B
CP a B zero?
JP NZ,NEXNUM no
LD A,C
CP 0 C zero?
JP N2,NEXNUM no
HALT

The index register IY which is used as the loop index is

initialised to the first number, 1000, and the DE register pair
is set to the decrement (negat ive increment) value.

Each time through the loop a carriage-return line-feed is output

to the display followed by the value in IY - the IYOUT subroutine
outputs the contents of the IY register as an unsigned number in
the range a to 65535.

After the repeated instructions IY is decremented by 5, that is,

DE, containing -5, i s added to IY. It is then necessary to check
if IY is zero. This check is not as straightforward as you might
think. There are several methods of performing the check; none
of them is a neat method. In Program 13.2 the check is made by
splitting IY into the two registers Band C each of which is then
checked for zero.

You will often find that the IX and IY index registers are
awkward to use and that the other register pairs, particularly
HL, are more convenient to use because there is a wider range of
more flexible instructions involving these register pairs.
74 Programming in Z80 Assembly Language

13.4 MULTIPLE BYTE ARITHMETIC

There are two instructions which can be used for multiple byte
arithmetic, in the same way that we saw that the two ADC HL,ss
and SBC HL,ss instructions could be used for multiple 16-bit
arithmetic.

The equivalent add instruction is

ADC A,s

where s is either a value, a single register or a memory byte

pointed to by HL, IX or IY. The instruction causes s to be added
to the accumulator along with the carry flag.

The equivalent subtract instruct ion is

SBC A,s

where s is the same as for the ADC A,s instruction. This

instruction causes s and the carry flag to be subtracted from the
accumulator.

The principle of multibyte arithmetic is that the two least

significant bytes of the numbers to be added (subtracted) are
added (subtracted) using an ADD (SUB) instruction and the
remaining pairs of bytes, going from the next least significant
byte to the most significant byte, are added (subtracted) using
the ADC (SBC) instruction.
Depending on the number of bytes to be added, there are many
combinations of 8-bit arithmetic i ns t r uc t i ons and 16-bit
arithmetic instructions which can be used. However, for the
general case of any number of bytes (which may, therefore, be an
odd number), an appropriate number of 8-bit arithmet ic
instructions is most suitable.

Program 13.3 shows a program segment which adds mult iple byte
numbers.

Program 13.3 addition of multiple byte numbers

SCF reset carry flag

CCF

NEXBYT: LD A,( IX)

ADC A,(IY) add next pair of bytes
LD (HL) ,A and store result
DEC IX
DEC IY point to next more significant
DEC HL bytes
DJNZ NEXBYT
addition done
 16-Bit and Multiple Byte Arithmetic 75
Initially. the registers IX. IY and HL point to the least
significant bytes of the first number. the second number and the
sum. respectively. and the B register contains the number of
bytes to be added. The carry flag must be initialised to 0
before the loop is entered so that the first add instruction is
equivalent to an ADD instruction.

The program segment in Program 13.3 can be amended to make it

subtract the second number from the first number by replacing the
ADC opcode with a SBC opcode.

13.5 PROGRAM

Write a program which outputs on the display the numbers from m

to n in steps of k. one number per line.

The numbers m. nand k are uns igned hexadecimal numbers which the
program inputs from the keyboard.

The level of difficulty of this program may be varied by

restricting the values of m. nand k to fit into

8-bits (8-bit arithmetic).

16 bits (16-bit arithmetic or double 8-bit arithmetic).
24 bits (16-bit arithmetic for the least significant 16
bits and 8-bit arithmetic for the most
significant 8 bits).
32 bits (double 16-bit arithmetic).
nx16 bits (multiple 16-bit arithmetic).
or nx8 bits (multiple byte arithmetic).

Additionally. the program could be made to input and output

decimal. rather than hexadecimal numbers.
14 Block Transfer and Search
Instructions

The Z80 microprocessor has eight very powerful block instructions

which allow operations on blocks of consecutive memory bytes.
Four of the instructions are block transfer instructions which
allow the contents of one block of memory bytes to be transferred
to another block of memory; the other four instructions are block
search instructions which allow a block of memory bytes to be
searched for one of the bytes containing a specified value.

All the block instructions are included in Table C.4 in Appendix

C.

14.1 BLOCK TRANSFER INSTRUCTIONS

Suppose it was necessary to move the contents of a block of ten

memory bytes, starting at the byte labelled SOURCE, to a block
starting at a byte labelled DESTIN. Program 14.1 could be used
to perform that function.

Program 14.1 block transfer - the hard way

LD HL,SOURCE set up pointers

LD DE,DESTIN
LD B,10 and counter

NEXBYT: LD A,(HL) transfer byte

LD (DE),A

INC HL increment pointers

INC DE
DJNZ NEXBYT and counter
HALT

SOURCE: DEFM 'ABCDEFGHIJ'

DESTIN: DEFM '0000000000'

76
 Block Transfer and Search Instruct ions 77

The register pairs HL and DE are set to point to the first byte
of the source and destination blocks of memory ~ytes,
respectively. The register B is to be used as the counter and is
initialised to ten.

In the loop, a byte is transferred from the source block of

memory bytes to the destination block of memory bytes via the
accumulator. The register pairs HL and DE are both incremented
by one to point to the next bytes in the source and destination
blocks of memory bytes.

When the looping finishes and the HALT instruction is executed,

the string of the first ten letters of the alphabet will have
been moved, character by character, to the ten bytes starting at
DESTIN, thereby overwriting the zeros originally contained in
those bytes.

In order for the program to deal with blocks of thousands of

memory bytes instead of tens of memory bytes the LD B,10
instruction must be replaced by a LD BC,number-of-bytes
instruction and the DJNZ NEXBYT instruction must be replaced by

DEC BC
LD A,B
CP 0
JP NZ,NEXBYT
LD A,C
CP 0
JP NZ,NEXBYT

The Z80 microprocessor has an instruction which could replace the

instructions in the loop in Program 14.1 and deal with blocks of
thousands of memory bytes. It is the LDIR instruction - LoaD,
Increment and Repeat. Prior to execution of the LDIR
instruction, HL must contain the address of the first of the
source block memory bytes, DE must contain the first of the
destination block memory bytes, and BC must contain the number of
bytes to be transferred. Program 14.2 shows a version of Program
14.1 us ing the LDIR instruction.

Program 14.2 block transfer - the easy way

LD HL,SOURCE set up pointers

LD DE,DESTIN
LD BC,10 and counter

LDIR transfer block

HALT

SOURCE: DEFM ' ABCDEFGHIJ

DESTIN: DEFM '0000000000'

Program 14.2 is functionally the same as Program 14.1 except that

78 Programming in Z80 Assembly Language

Program 14.2 allows blocks of up to 64K bytes to be transferred,

since a register pair is being used as the counter instead of a
single register. For each byte transfer HL and DE are
incremented by one and BC is decremented by one - the transfer
continues until BC is equal to O.

The LDDR - LoaD, Decrement and Repeat instruction is the same as

the LDIR instruction except that, as its mnemonic suggests, HL
and DE are decremented instead of incremented.

The following program segment has the same effect as Program 14.2
but uses the LDDR instruction instead

LD HL,SOURCE+9 set pointers to the end

LD DE,DESTIN+9 of the blocks
LD BC,10
LDDR transfer block

Two other block transfer instructions LDI and LDD are similar to
the LDIR and LDDR instructions except that they do not
automatically go on to transfer the next byte.

The LDI - LoaD and Increment instruction increments up from the

beginning of the block of bytes, whereas the LDD - LoaD and
Decrement instruct ion decrements down from the end of the block
of bytes. The LDI instruction transfers a byte, increments both
HL and DE by one and decrements BC by one, whereas the LDD
instruction, after transferring a byte, decrements both HL and DE
by one and decrements BC by one.
It is important to know that the condition of BC becoming zero is
indicated by the P/V flag not the zero flag. The P/V flag is set
to 0 (PO mnemonic) if BC is zero, otherwise it is set to 1 (PE
mnemonic) •

Program 14.2 could be written using an LDI instruction instead of

the LDIR instruction, as follows

LD HL,SOURCE
LD DE,DESTIN
LD BC,10
NEXBYT : LDI transfer byte
JP PE,NEXBYT last byte? no, next byte
HALT yes

The LDIR and LDDR instructions can be used only when the number
of bytes to be transferred is known in advance. When the
criteria for the numbers of bytes to be transferred are not known
in advance, the LDI or LDD instructions must be used and the
program must write the instructions to transfer all the bytes.

The following program segment moves a block of characters from

one place to another.
 Block Transfer and Search Instruct ions 79

LD HL,HERE set up pointers

LD DE,THERE
LD BC,1000 and maximum counter
NEXBYT: LDI transfer byte
LD A,(HL)
CP 0 is byte zero?
JP NZ,NEXBYT no - go to transfer next byte

A maximum of one thousand characters in the block is catered for,

although the transfer stops when a byte containing zero is
encountered. The zero byte is not transferred.

Care must be taken during block transfers when the source and
destination blocks overlap. Take, for example, a situation where
it is necessary to move a block of bytes so many bytes through
memory as is done by the following program segment

LD HL,START
LD DE,START+100
LD BC,500
LDIR

The first one hundred bytes of the source block of bytes will be
copied into the first one hundred bytes of the destinat ion block
of bytes which also happens to be the second one hundred bytes of
the source block of bytes. So the last four hundred bytes of the
source block of bytes are overwritten before they can be
transferred to the destination block of bytes.

To make the block of bytes move down through memory correctly the
above program segment would have to be changed to

LD HL, START+4 99
LD DE,START+599
LD BC,500
LDDR

14.2 BLOCK SEARCH INSTRUCTIONS

There are four block search instructions which allow a block of

memory bytes to be searched for one of them containing the same
contents as the accumulator. The operation of all four
instructions requires that the accumulator shall contain the byte
contents to be searched for, HL contain the address of the first
byte of the block of bytes to be searched, and BC contain a count
of the number of bytes in the block, all to be set before the
instructions are executed.

As for the block transfer instructions, two of the block search

instructions automatically search through the block and the other
two block search instructions require extra instructions to move
on to the next byte in the block.

The two automatic block search instructions are CPIR - ComPare,

Increment and Repeat and CPDR - ComPare, Decrement and Repeat.
80 Programm ing in Z80 Assembly Language

Program 14.3 shows a use of the CPIR instruction in searching

for the value zero in a block of memory bytes.

Program 14.3 search a block for zero

LD HL,START set up pointer,

LD BC,10 counter
LD A,O and accumulator

CPIR search block for zero

LD A,C output value of counter

ADD A,30H
CALL COUT
HALT

START: DEFB 1
DEFB 2
DEFB 3
DEFB 0
DEFB 4
DEFB 5
DEFB 0
DEFB 6
DEFB 0
DEFB 7

First of all HL is loaded with the address of the first byte of

the block labelled START, the register pair Be is loaded with the
number of bytes in the block and the accumulator is loaded with
the value to be found.

The CPIR instruction then indexes through the block of bytes

until either a match with the contents of the accumulator is
found or the end of the block is reached, that is, BC becomes
zero. For each byte, the contents of the accumulator are
compared with the contents of the byte; if they are equal the
zero flag is set to 1; the HL register pair is incremented by one
and the BC register pair is decremented by one. Finally, if the
zero flag is set to 1, or BC is zero, the instruction is
finished; otherwise the next byte in the block is considered, and
so on.

In the same way that the P/V flag is used to indicate that BC
contains zero during execution of the LDI and LDD block transfer
instructions, so the P/V flag is used to indicate the contents of
BC on termination of the block search instructions.

As it stands Program 14.3 will output the number 6, ie. 10 - 4,

as the first zero byte is in the fourth byte of the block of
bytes. If the block of bytes did not contain a zero byte the
program would output the number zero.

The CPDR instruction can be used to search a block from the end
 Block Transfer and Search Instructions 81

back to the beginning. In this case, HL is initially set to

point to the last byte 'of the block of memory bytes and during
execution of the instruction HL is decremented, rather than
incremented, by one.

The CPI - ComPare and Increment and CPD - ComPare and Decrement
instructions are similar to the CPIR and CPDR except that they do
not automatically go on to the next byte. Extra instructions
have to be used to test for a match between the accumulator and
whether the byte has been found - zero flag is set to 1 - and to
detect if the whole of the block has been searched - BC is zero
and the P/V flag is set to O. These two instructions are used in
place of the CPIR and CPDR instructions when intermediate
processing is required and, for example, when more than one
occurrence of the value in the accumulator needs to be detected.

To make Program 14.3 output the value of the counter for every
occurrence of a zero byte the instructions CPIR to HALT would
have to be replaced by the following instructions

NEXBYT : CPI
JP PO,FINI end of block?
JR NZ,NEXBYT
LD A,C output counter
ADD A,30H
CALL COUT
LD A,O restore A
JR NEXBYT
,
FINI: HALT

It is sometimes useful to know the contents of HL and DE when the

block instructions LDIR, LDDR, CPIR and CPDR have finished
execution.

For the LDIR instruction, HL and DE will be pointing to the bytes

immediately following the ends of the blocks.

For the LDDR instruction, HL and DE will be pointing to the bytes

immediately preceding the beginning of the blocks.

For the CPIR instruction, HL will be pointing to the byte

immediately following the end of the block.

For the CPDR instruction, HL will be pointing to the byte

i mmedi a t e l y preceding the beginning of the block.

14.3 PROGRAM

This program is required to provide an internal filing system

which can hold up to nine records. Each record contains 20
characters.

The file should be defined at the end of the main program using
82 Programming in Z80 Assembly Language

the pseudo instruction

FILE: DEFS 180

So far as the program is concerned, the contents of the records

(the 20 characters) are immaterial. The program identifies each
record by a record number which is its numerical position in the
file; so the file has records numbered 1 to 9.

A user of the program should be able to input any of the

following

Dn to delete the record numbered n (n

i s 1 to 9) - following records move
up one record position,

I n 20-character-record to insert the specified record

after the record numbered n (n is
o to 8)- following records move
down one record position,

R n 20-character-record to replace the record numbered n by

the specified record (n is 1 to 9),

L to list the file on the display -

for each record, its number and its
contents,

F character-string to find and display the first

record in the file containing the
specified character-string, which
may be 1, 2 or 3 characters.
15 Decimal Arithmetic

Up to now, we have only considered the binary representation of

numbers in the Z80 microprocessor, and arithmetic has involved
signed and unsigned binary numbers . However, the Z80
microprocessor caters for another representation of numbers
Binary Coded Decimal, or BCD for short.

15.1 BCD REPRESENTATION

The Binary Coded Decimal representation of numbers requires that

each decimal digit be expressed as a 4-digit binary number, so
that 9 would be written as 1001B.

A 4-bit unit is called a nibble, so a nibble is half a byte and a

byte can hold a 2-digit BCD number, as follows

7 6 5 4 3 2 o

left nibble right nibble

84 Programming in Z80 Assembl y Language

The left nibble, bits 7 to 4, inclusive, contains the BCD number

7 and the right nibble, bits 3 to 0, inclusive, contains the BCD
number 4. The byte as a whole contains the BCD number 74.

A nibble can contain a BCD number in the range 0 to 9, but an

unsigned binary number in the range 0 to IS, so you can see that
the BCD representation is wasteful compared to the unsigned
binary representation. A byte can contain BCD numbers in the
range 0 to 99 and unsigned binary numbers in the range 0 to 255,
ie. more than double the range for BCD numbers.

Numbers are normally input from the keyboard and output to the
display as decimal digits. To enable arithmetic to be performed
on the numbers they have to be converted to binary, and then the
results converted from binary to decimal before being output. If
a computer has instructions which allow arithmetic to be
performed with numbers in their decimal digit form then the
conversion from decimal to binary, and vice versa, would be
unnecessary.

The Z80 microprocessor does have facilities for decimal

arithmetic (the Decimal Adjust Accumulator instruction) and
facilities for moving nibbles around (the Rotate Left Dig it and
Rotate Right Digit instructions), so sometimes it is more
convenient to use the BCD representation of numbers rather than
their binary representation.

15.2 BCD ARITHMETIC

As you would expect, BCD arithmetic is not as straightforward as

binary arithmetic on a computer which is binary orientated. The
standard arithmetic instructions cannot be used because they are
binary and, therefore, not suitable for BCD arithmetic. For
example, the sum of BCD 28 and BCD 39, would yield the incorrect
result of BCD 61, if a standard binary add operation were to be
used, as follows·

00101000 BCD 28
+ 00111001 BCD 39

01100001 BCD 61 - incorrect

The error occurs because, in BCD arithmetic, if the sum of the

right nibbles is greater than 9 a carry into the left nibble is
required. Binary arithmetic also produces nibbles containing
numbers in the range 1010B to 1111B for which there is no valid
BCD equivalent.

Having performed a binary add operation on two BCD numbers it is

possible to correct the erroneous result to give a corr ect BCD
result. The correction after the addition of two BCD digits is
as follows: if the nibble contains a value between 1010B and
1111B, inclusive, or a carry occurred from the most significant
 Decimal Arithmetic 85

bit of the nibble then 0110B must be added to the nibble,

otherwise nothing need be done. The two following examples show
the effect of the correction

0110 BCD 6
+ 0111 BCD 7

1101 -incorrect BCD result not a BCD digit

+ 0110
00010011 BCD 13 - correct BCD result

1000 BCD 8
+ 1001 BCD 9

00010001 BCD 11 - incorrect BCD result

+ 0110
--------
00010111 BCD 17 - correct BCD result
--------

An example of a 2-digit BCD addition is

00010111 BCD 17
+ 01101001 BCD 69
--------
10000000 BCD 80 - incorrect BCD result
+ 0110
--------
10000110 BCD 86 - correct BCD result
--------
A carry occurring from the left nibble of a 2-digit BCD addition
would indicate overflow, that is, a value greater than 99.

When subtracting one BCD number from another, 0110B must be

subtracted from the result if either a borrow has occurred into
the nibble or the nibble contains a value in the range 1010B to
1111B, inclusive.

The following shows the subtraction of BCD 56 from BCD 82

10000010 BCD 82
- 01010110 BCD 56

00101100 incorrect result - not a BCD digit

- 0110

00100110 BCD 26 - correct BCD result

86 Programming in Z80 Assembly Language

A borrow occurring into the left nibble of a 2-digit BCD

subtraction would indicate overflow.

15.3 THE DAA INSTRUCTION

In order to provide decimal arithmetic a computer must either

provide a separate set of decimal arithmetic instructions, such
as decimal Add, or provide a means of changing a binary
arithmetic result into a decimal arithmetic as shown in the
previous section. The Z80 microprocessor designers chose the
latter by including in the Z80 instruction set the DAA - Decimal
Adjust Accumulator - instruction.

Whenever an arithmetic instruction is executed the two flags in

the flag register which we have not considered so far - the half-
carry and subtract flags - are affected in the following way.

The half-carry flag is set to 1 if an add instruction produces a

carry out of the right nibble of a register, otherwise the flag
is reset to O. Also, the half-carry flag is set to 1 if a
subtract instruction requires a borrow into the right nibble of a
register, otherwise the flag is reset to O.

The subtract flag is set to 1 if the instruction is an add

instruction or to 0 if it is a subtract instruction.

Although the half-carry and subtract flags are used only by the
DAA instruction they are set accordingly after every arithmetic
instruction. Neither of these flags may be used by a programmer
because there are no instructions to set or test them directly.

Using the setting of the half-carry and subtract flags, the DAA
instruction corrects the contents of the accumulator, if
necessary, to give a result in the accumulator as if the previous
arithmetic instruction had been a BCD one. For example, the
program segment

~D A,43H
10 B,28H
ADD A,B
DAA

loads the accumulator and B register with BCD 43 and BCD 28,
respectively, sums them using a binary ADD instruction, and then
adjusts the result to BCD representation using the DAA
instruction. Notice that hexadecimal representation, being a 4-
bit representation, is very useful for writing BCD constants in
programs.

The DAA instruction is used during BCD arithmetic after the

 Decim al Arithmetic 87

instructions ADD A, SUB, INC A, DEC A, CP, NEG, ADC A, SBC and
the four block search instructions. Notice that the DAA
instructions operate on the accumulator only.

The two flags of most import ance to the programmer, which are set
by the DAA instruction, are the carry flag indicating BCD
arithmetic overflow and the zero flag indicating a zero BCD
value. The carry flag setting can also be used in multi-byte BCD
arithmetic, as we shall see later.

15.4 THE DIGIT ROTATE INSTRUCTIONS

Two Z80 instructions are available for rotating nibbles, or BCD

digits - one to rotate to the left and the other to rotate to the
right .

The rotation involves the right nibble of the accumulator and the
two nibbles of a memory byte pointed to by the HL register pair.
The Rotate Left Digit instruction, RLD, operates as follows

D_
t ---JI LJ

accumulator memory byte

The right nibble of the accumulator moves to the right nibble of

the memory byte. The right nibble of the memory byte moves to
the ieft nibble of the memory byte and the left byte of the
memory byte moves to the right nibble of the accumulator . The
left nibble of the accumulator is not affected by the rotation.

The Rotat e Right Digit instruction, RRD, operates as follows

accumulator memory byte

The RLD and RRD instructions are very useful for manipulating BCD
numbers. For example, look at Program 15.1 which inputs two 2-
digit BCD numbers, adds them, and outputs their sum.
88 Programming in Z80 Assembly Language

Program 15.1 add two 2-digit BCD numbers

LD HL,NUM
CALL CIN input 1st digit of first number
AND OFH
LD (HL) ,A
CALL CIN input 2nd digit of first number
AND OFH
RLD store first number in NUM
INC HL
CALL CIN input 1st digit of second number
SUB 30H
LD (HL),A
CALL CIN input 2nd digit of second number
SUB 30H
RLD store second number in NUM+1
LD HL,NUM
10 A,(HL)
INC HL
ADD A,(HL) add the two numbers
DAA decimal adjust for BCD
LD HL,SUM save result
LD (HL),A
LD A,O output result
RLD
ADD A,30H
CALL COUT - first digit
LD A,O
RLD
ADD A,30H
CALL COUT - second digit
HALT
,
NUM: DEFS 2
SUM: DEFS 1
Working through the program, the first number is input, digit by
digit, and stored in NUM using an RLD instruction, followed by
the input of the second number into NUM+1 using another RLD
instruction. TIle HL register pair must be used to point to the
memory bytes because the RLD instruction assumes that it does.

The two numbers are then added and immediately adjusted for BCD
arithmetic using the DAA instruction. The result is stored
before being output, digit by digit, using an RLD i ns t r uc t i on to
move each nibble in turn from SUM to the accumulator ready for
output. Notice that the accumulator has to be set to zero before
the last two RLD instructions are executed - this is to ensure
that the left nibble of the accumulator is zero. Although the
left nibble is not involved in the decimal rotate instruction,
and partly because it is not, you must ensure that the contents
of the left nibble of the accumulator are what you want them to
be.
 Decimal Arithmetic 89

15.5 PROGRAM

A program which inputs, adds, subtracts and outputs signed

multiple-digit BCD numbers is required.

A BCD number is assumed to be held internally as follows: an

initial byte specifies the sign of the number (bit 7 is 0 for a
positive number or 1 for a negative number) and the number of BCD
digits contained in the number (bits 6 to 0); subsequent bytes
contain the BCD digits packed two to a byte.

Write a subroutine called INBCD which inputs a signed BCD number

(no sign implies a positive number) from the keyboard and stores
it as specified above. On entry to the subroutine the HL
register pair points to the first of the bytes where the number
is to be stored.

Write a subroutine called OUTBCD which outputs a signed BCD

number (suppressed positive sign) to the display. On entry to
the subroutine the HL register pair points to the first byte of
the number to be output.

Write a subroutine called BCDADD which adds two equal-length

signed BCD numbers. On entry to the subroutine the IX and IY
register pairs point to the first bytes of the first and second
numbers, respectively, and the HL register pair points to the
first byte of the resulting number. On exit from the subroutine
the accumulator should contain a 1 if overflow occurred, or 0
otherwise.

Write a subroutine called BCDSUB which is similar to the above

add subroutine except that the subroutine subtracts the second
number from the first number.

Use the subroutines to write a main program which repeatedly

inputs two signed BCD numbers separated by either a + character
or a - character and followed by an equal character, and outputs
the result. The program should cater for BCD numbers containing
up to twenty BCD digits.
16 Miscellaneous Instructions

There are several instructions which have not yet been considered
because they are rarely used or because they are beyond the scope
of this book . However, for completeness, they are discussed
briefly in this final chapter.

16.1 THE NOP INSTRUCTION

The NOP instruction performs No OPeration but, paradoxically,

does have some uses. For example, the instruction can be used to
provide a delay in a sequence of code by inserting the
instructions
DELAY: NOP
DJNZ DELAY
which will cause a delay of (N - 1) x (4 + 13) + 8 clock periods.
(N is the value in the B register prior to the delay loop; the
NOP instruction takes 4 clock periods and the DJNZ instruction
takes 13 clock periods if B is not zero and 8 clock periods if B
is z ero , )

16.2 THE AUXILIARY REGISTERS

The Z80 microprocessor has another set of eight registers called

the auxiliary registers which are denoted by A', F', B', C', D',
E', H' and L'. These auxiliary reg isters can be used in exactly
the same way as their counterparts, but not at the same time.

To change over from using the standard registers to using the

auxiliary registers, the instruction

EXX ; exchange standard and auxiliary registers

must be executed, causing subsequent instructions to refer to the

auxiliary registers. To revert back to using the standard
registers another EXX instruction must be executed.

90
 Miscellaneous Instructions 91

Most programs need to use only the standard registers and to

exchange all eight registers would not be necessary. However,
for the more frequent requirement that a second accumulator is
sometimes necessary, the instruction

EX AF,AF', exchange AF and AF'

is available which interchanges the standard and auxiliary

accumulator and flag registers.

16.3 INPUT AND OUTPUT INSTRUCTIONS

There are twelve input and output instructions altogether and

these are specified in Table C.12 of Appendix C.
Input and output of data can be specified to be to and from any
of the single registers using the IN r,(C) and OUT (C), r
instructions, in which case the contents of the C register
identify the port to be used.

The remaining input and output instructions allow the input and
output of blocks of data. These instructions, and their
variations, are similar to the block search instructions except
that, instead of comparing a data byte, a data byte is input or
output. (Also, only the single B register is used as a counter -
not the BC register pair.)

These block input and output instructions appear very useful at

first sight, but it must be remembered that the repeat ones can
be used only with devices which operate at the same high speed as
instructions, which does not include, for example, displays or
keyboards.

16.4 INTERRUPT INSTRUCTIONS

An interrupt facility allows signals from outside to interrupt

the sequence of i ns t r uc t i ons in the central processing unit.

Five of the instructions associated with the zao interrupt

facility are specified in Table c.a in Appendix C. The zao
microprocessor has three modes of interrupt which can be set by
instruction using one of the three 1M 0, 1M 1 and 1M 2
instructions which set the interrupt mode to modes 0, 1 or 2,
respectively. The DI and EI instructions Disable Interrupts and
Enable Interrupts and, therefore, allow interrupts to be disabled
or enabled under program control.

Interrupts are dealt with by Interrupt Service Routines - normal

program segments with a specific function of dealing with a
particular interrupt. The last instruction of an interrupt
service routine is either a RETI (RETurn from Interrupt) or RETN
(RETurn from Non-maskable interrupt) instruction which causes a
return back to the sequence of instructions which was
interrupted. The RETI and RETN instructions are included in
Table C.11 in Appendix C along with the RST (ReSTart) instruction
92 Programm ing in Z80 Assembly Language

which is used to service a mode 0 interrupt .

Appendix A: Binary and
Hexadecimal Number Systems

In order to understand how information is stored in the memory of

a computer we need to know about binary and hexadecimal numbers.
In everyday l i f e we normally use decim al numbers. However,
computers store information in binary, and hexadecimal i s a
compact way of representing binary.

Put simply , decimal is counting in tens, binary i s counting in

twos and hexadecimal is counting in sixteens.

A.1 BINARY AND HEXADECIMAL NUMBERS

A decimal number, say 453 may be expressed in the following way

453 400 + 59 + 3 1)
(4 x 10 ) + (5 x 10 + (3 x 10°)

Similarly, a hexadecimal number, say 974, is expressed as

and the binary number 101 is expressed as

101B = 100B +2 00B + 1B 1

= (1 x 2 ) + (0 x 2 ) + (1 x 2°)

The H at the end of the hexadecimal number 974 is there to

indicate that that number is, in fact, hexadecimal rather than
decimal or binary. Sim ilarly, a binary number is postfixed by
the letter B.

Looking at the three numbers above you can see that de cimal
numbers are expressed in terms of the powers of tens, hexadecimal
numbers are expressed in terms of the powers of sixteens and
binary numbers are expressed in terms of the powers of two . The

93
94 Programming in Z80 Assembl y Language

ten, sixteen and two are said to be the base or radix, of the
numbers. Decimal numbers have a base of ten, hexadecimal numbers
a base of sixteen and binary numbers a base of two. Any number
can be used as a base, but in computing, and particularly for
microprocessors, the most common bases are sixteen and two.

Throughout this appendix there are several exercises for

you to do to give you practice in number systems. The
answers to the exercises are at the end of this appendix.

Exercise 1
By working out the expressions above, what are 974H and 101B
equivalent to as decimal numbers?
You know already that decimal numbers use the digits a to 9, that
is, zero through to one less than the base value.

Exercise 2
Which digits do binary numbers use?

Hexadecimal numbers need to use sixteen, that is a to something.

We can use the same digits as are used for decimal numbers up to
9 but for the remaining six digits we need single-character
symbols. The chosen symbols are the letters A, B, C, D, E and F,
so that hexadecimal A is equivalent to decimal 10 and hexadecimal
F is equivalent to decimal 15.

Look now at Figure A.l which shows the equivalent hexadecimal and
binary numbers of the decimal numbers a to 15.

decimal hexadecimal binary

a a 0000
1 1 0001
2 2 0010
3 3 0011
4 4 0100
5 5 0101
6 6 0110
7 7 0111
8 8 1000
9 9 1001
10 A 1010
11 B 1011
12 C 1100
13 D 1101
14 E 1110
15 F 1111

Figure A.l

You will need to know the hexadecimal and binary numbers in

Figure A.l by heart, so spend a short time making sure that you
know them without having to think about it - especially the
binary numbers.
 App endix A: Binary and Hexadecimal Number Systems 95

Exercise 3
What is the decimal equivalent of E8A5H?

A.2 BINARY AND HEXADECIMAL ARITHMETIC

Addition and subtraction can be done using any base. The

technique is the same as for decimal numbers except that any
reference to ten is replaced by a reference to the base. For
example, when adding two hexadecimal numbers, a carry is produced
when the addition of two of the digits results in a number
greater than FH (or decimal 15).

Examples of addition and subtraction using hexadecimal and binary

numbers are
3A7FH 10110110B
+ 10BBH - OlO11010B

4B3AH 01011100B

Exercise 4
Do the following arithmetic

C7BAH OllOllOlB
9FF8H + 01011110B

A.3 DECIMAL TO HEXADECIMAL CONVERSION

To convert a decimal number to a hexadecimal number, repeatedly

divide the decimal number by 16 until a 0 quotient is obtained.
The remainders from the divisions constitute the equivalent
hexadecimal number, the last remainder so obtained being the most
significant digit of the hexadecimal number. For example, the

I'::
conversion of 745 to hexadecimal looks like

16

16 r-emat nder 9

16 2 remainder E

o remainder 2
the equivalent hexadecimal number being 2E9H.

Exercise 5
Convert the decimal number 1582 to hexadecimal.

A.4 HEXADECIMAL TO DECIMAL CONVERSION

To convert a hexadecimal number to a decimal number just expand

96 Programm ing in Z80 Assemb ly Language

the hexadecimal number in powers of 16, and then add the terms.
For example, the conversion of 3AB2H to its equivalent decimal

number would look like

3AB2H

(3 x 4096) + (10 x 256) + (11 x 16) + (2 x 1)

12288 + 2560 + 176 + 2

15026

the equivalent decimal number being 15026.

A quicker way to convert a hexadecimal number to a decimal number

(and vice-versa) is to use a conversion table, assuming that one
is readily available.

Exercise 6
Using the converion tables in Appendix B, convert FBH and A3B2H
to decimal and 142 and 9467 to hexadecimal.

A.5 BINARY-HEXADECIMAL CONVERSION

Binary to hexadecimal conversion and hexadecimal to binary

conversion is based on the fact that one hexadecimal digit can be
replaced by four binary digits and vice-versa.

So to convert a hexadecimal number, s ay 6BH, to binary, just

replace each hexadecimal digit by its four dig it binary
equivalent - according to the values in Figure A.1. Hence,

6BH = 0110 1011

which is equal to 1101011B

with the leading zero removed and the two sets of binary digits
joined together.

To convert a binary number to a hexdecimal number, the binary

number is separated into groups of four binary digits from the
right. For example, the binary number 1111100111 would look like

11 1110 0111

Each group of bits is then converted to its equ ivalent

he xadec imal digit, so the binary number above would look like

3 E 7

Hence, 1111100111B is equivalent to 3E7H.

Exercise 7
Convert 9AB3H to binary and 110011101111B to he xadec imal.
 Appendi x A: Binary and Hexadecimal Numb er Systems 97

A.6 DECIMAL-BINARY CONVERSIONS

Conversions between decimal and binary numbers can be done in the

same way as we did decimal and hexadecimal number conversions,
except that 2 is used wherever we used 16.

However, those methods of conversion are rather tedious for

decimal/binary conversions so, either

use hexadecimal as an intermediary, so that,

for example, to convert from decimal to binary
first convert from decimal to hexadecimal and
then to binary,

or use a conversion table.

Exercise 8
Convert 1290 to binary and 101110111101B to decimal using both
suggested methods.

A.7 BYTES

The basic unit of data in the Z80 microprocessor is a byte, which

contqins eight binary digits (bits, for short) or two hexadecimal
digits. The contents (O's and l's) of a byte may represent any
one of several entities, such as

a character,
a number (unsigned),
or a signed number.

Representation of characters is dealt with in Chapter 3.

Representation of a number in a byte refers to the contents of a

byte being considered t o be the value of the binar y number
contained in the byte. For example, a byte conta in ing 01100110B
represents the number 11100110B, 66H or 102 decimal. The range
of numbers which can be contained in a byte is 0 to 11111111B
(FFH and 255 decimal). When it is necessary to do so, this
representation is distingu i shed from another representation by
referring to it as the unsigned number representation.

Exercise 9
What range of unsigned numbers can be represented in two bytes
(that is, 16 bits)?

A.8 SIGNED (2's COMPLEMENT) NUMBERS

Numbers which may have negative values as well as positive values

are held in computers in what is called '2's complement form'.
This form of representation depends on numbers consisting of a
fixed number of digits. As we are concerned with the Z80
microprocessor we will consider 2's complement numbers consisting
of eight binary digits.
98 Programming in Z80 Assembly Language

In the 2's complement system, a negative number is represented by

taking the 2's complement of its equivalent positive value; this
is done by converting all O's to l's and l's to O's, and then
adding 1. For example,

+5 is 00000101B

so -5 is 11111010B
+ 1
--------
11111011B
--------
Hence, -5 is held as 11111011B in a register or memory byte. The
mechanism for producing a negative number in 2's complement form
is equivalent to subtracting the equivalent positive value from
2.
Exercise 10
Calculate the binary equivalent of -1, -2 and -126 and the
decimal equivalent of 10000000B and 10000001B, assuming an 8-bit
2's complement system.

When performing arithmetic with numbers within a 2's complement

system, numbers are added bit by bit as normal but any carry out
of the most significant bit is ignored. For example, adding +5 and
-5 looks like

00000101B +5
+ 11111011B + -5

[I] OOOOOOOOB 0

The one carry out of the addition of the eight two bits is
ignored and the result is contained in the 8 bits - that is,
zero, which you would expect to obtain when adding -5 to +5.

Exercise 11
Calculate -60 + 70, - 23 + -46, 85 - 96 and 5 - -121, in binary
using an 8-bit 2's complement system.

The range of numbers which can be held in a byte, using 2's

complement, goes from -128 to +127 as follows

-128 10000000
-127 10000001

-2 11111110
-1 11111111
0 00000000
+1 00000001
+2 00000010

+127 01111111
 Append ix A: Binary and Hexadecimal Number Systems 99

There are other ways of representing negative numbers in

computers, but the 2's complement method is the most common and
the one used by the Z80 microprocessor. However, rather than
using the inelegant phrase '2's complement' we shall refer to
'signed numbers' rather than 2's complement numbers from now on.

It may help you in your understanding of signed numbers and

unsigned numbers to look at the weighting of the bits in a byte
for each of the representations. They are

unsigned numbers 128 64 32 16 8 4 2

signed numbers -128 64 32 16 8 4 2

so that, for example, the unsigned number 10010001B is equ ivalent

to

1 x 128 + 1 x 16 + 1 x 1 which equals 145

whereas, the signed number 10010001B is equivalent to

1 x -128 + 1 x 16 + 1 x 1 which equals -111.

Exercise answers

974H is equivalent to 2420.

101B is equivalent to 5.

2 0 and 1.

3 E8A5H E x 4096 + 8 x 256 + A x 16 + 5 x

14 x 4096 + 8 x 256 + 10 x 16 + 5 x
57344 + 2048 + 160 + 5
59557

4 C7BAH OllOllOlB
- 9FF8H + 01011110B
27C2H 1l001011B

5 62EH
100 Programming in Z80 Assembly Language

6 FBH is equivalent to 251

A3B2H is equivalent to 41906
142 is equivalent to 8EH
9467 is equivalent to 24FBH

7 9AB3H is equivalent to 1001101010110011B

110011101111B is equivalent to CEFH

8 1290 is equivalent to 50AH and 10100001010B

101110111101B is equivalent to BBDH and 3005

9 Unsigned numbers in the range 0 to 1111111111111111B (FFFFH

and 64535) can be represented in two bytes.

10 -1 is equivalent to 11111111B
-2 is equivalent to 11111110B
-126 is equivalent to 10000010B
10000000B is equivalent to -128
10000001B is equivalent to -127

11 11000100 -60
+ 01000110 + +70

[1) 00001010 +10

11101001 -23
+ 11010010 + -46
[1) 10111011 -69

01010101 +85
- 01100000 - +96
--------
[ 1) 11110101 -11
--------
00000101 +5
- --------
10000111 --121

[ 1) 01111110 +126
--------
Appendix B: Hexadecimal-Decimal
Conversion Tables

The table below provides for direct convers ion between

hexadecimal numbers in the range 0 to FF and decimal numbers in
the range 0 to 255.

o 234 5 6 7 8 9 ABC D E F

00 000 001 002 003 004 005 006 007 008 009 010 011 012 013 014 015
10 016 017 018 019 020 021 022 023 024 025 026 027 028 029 030 031
20 032 033 034 035 036 037 038 039 040 041 042 043 044 045 046 047
30 048 049 050 051 052 053 054 055 056 057 058 059 060 061 062 063
40 064 065 066 067 068 069 070 071 072 073 074 075 076 077 078 079
50 080 081 082 083 084 085 086 087 088 089 090 091 092 093 094 095
60 096 097 098 099 100 101 102 103 104 105 106 107 108 109 110 III
70 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
80 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143
90 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159
AO 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175
BO 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191
CO 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
DO 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223
EO 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239
FO 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255

For conversion of larger numbers use the following in conjunction

with the table above.

101
102 Programming in Z80 Assembly Language

Hexadecimal Decimal

100 256
200 512
300 768
400 1024
500 1280
600 1536
700 1792
800 2048
900 2304
AOO 2560
BOO 2816
COO 3072
DOO 3328
EOO 3584
FOO 3840
1000 4096
2000 8192
3000 12288
4000 16384
5000 20480
6000 24576
7000 28672
8000 32768
9000 36864
AOOO 40960
BOOO 45056
COOO 49152
DOOO 53248
EOOO 57344
FOOO 61440
Appendix C: Summary of zao
Instructions

This appendix contains a summary of the complete Z8D instruction

set.

The first table, C.l, gives a summary of the flag operations.

In tables C.2 to C.12, the instructions are logically arranged

into functional groups. Each table shows the assembly language
mnemonic OP code, the numeric OP code, the symbolic operation,
the content of the flag register following the execution of each
instruction, the number of bytes required for each instruction as
well as the number of memory cycles and the total number of T
states (external clock periods) required for the fetching and
execution of each instruction. Care has been taken to make each
table self-explanatory without requiring any cross reference with
the text or other tables.

The following pages have been reproduced by permission of Zilog,

Inc. 1977. This material shall not be reproduced without the
written consent of Zilog, Inc.

103
104 Programm ing in Z80 Assembly Language

Instruction C Z \,s N H Comments

ADD A, s; ADC A,s t t v t ot g·bit add or add with carry
SUB s: SBC A, s, CP s, NEG t t v t 1 t g·bit subt ract, subtract with carry , com pare and
negate accumul ator
AND s 0 P t o J Logical operat ions
DRs; XORs 0 t P t o0 And set's differen t flags
INC s
• t
v t ot g·bit increment
•t t v t I t g·bit decrement
DEC m
ADD DD, ss ••• 0 X
o
16·bit add
ADC Il l , ss t t v t X 16·bit add with carry
SBC Ill, ss t t V t I X 16-bit subtract with carry
RLA ; RLCA, RRA , RRCA t • • • o 0 Rot ate accumulator
Rl m; RlC m; RR m; RRC m t t P t o 0 Rotate and shift location m
SlA rn; SRA m; SRl m
RLD,RRD
• tt
P t 0 0 Rotate digit left and right
DAA P t • t Decimal adjust accumulator
CPL • • • • 0I J Complement accumulat or
SCF ••• 0 Set carry
CCF ••• 0 x Comp lement carry
IN r,(C) P t
• t xx
o 0 Input register indirect
INI ; IND; OUTI ; OUTD I X Block inpu t and output
INIR ; INDR ; OTJR ; OTDR
• t X X I X Z = 0 if B "" 0 othe rwise Z = I
• I
LDI,LDD
• X
t X o 0 Block transfer instructions
LDIR ,LDDR
CPI, CPIR , CPD, CPDR
• x ot xt o IX
0 P/V = I if BC "" 0, otherwise P/V = 0
Block search instructio ns
• t = =
Z I if A (HL) , otherwise Z 0 =
=
P/V I if BC "" 0 , otherwise P/V 0 =
LD A, I ; LD A, R ·tFFtOO The content of the interrupt enable flip-flop (IFF)
is co pied into the P/V flag
BIT b , s • t xix 0 I The state of bit b ofloc ation s is copied into the Z flag
NEG ttVtl t Negate accu mulato r

The followin g no tati on is used in th is table :

Symbol Operation
c Carry/link flag. C= 1 if th e operatio n produced a carry from th e MSB of the operand o r result .
z Zer o flag. 2=1 if the result of the ope ratio n is zero.
S Sign flag. S= 1 if th e MSB o f the result is one .
PfV Parity or overflow flag. Parity (P) and overfl ow (V) share the same flag. Logical ope ration s affec t th is flag
with the pari ty of the result while arithme tic ope rations affect th is flag with the overflow of the result. If P/V
holds parity , PfV= l i f the result of the operation is even, PfV=O if result is odd . If P/V holds overflow, P/V=1
if the result of the operation produced an overflow.
H Half-carry flag . H=1 if the add or subtract operation produ ced a carr y into or borrow from into bit 4 of the accu mulator.
N Add/Subtr act flag. N=1 if the previous operation was a subtr act
Hand N flags are used in conjunction with the decimal adjust instructio n (DAA) to prop erly correct the re-

.
sult into packed nco format following addition or subt raction using operands with packed BCD form at.
I The flag is affected according to the result of the operation.
The flag is unch anged by the operation.
o The flag is reset by the operatio n.
1 Th e flag is set by the operation.
X The flag is a " do n' t care."
V PfV flag affected accordi ng to the overflow result of the operation.
P P/V flag affected according to the parity result of the operation.
r Anyone of the CPU registers A, B, C, D. E, H. L.
S Any a·b it location for all the addressing mode s allowed for the particular instru ction. ,
"IRI Any 16-bit locati on for all the addressing modes allowed for that instruction.
Any one of the two index registers IX or IV.
Refresh count er,
n 8-bitvalue in range <0, 255>
no 16-bitvalue in range<D. 65535>
m Any 8- bit location for all the addressing modes allowed for the partic ular instructio n.

Summary of flag operation

Table C. l Courtesy Zi log, Inc.
 Appendix C: Summary of Z80 Instructions 10 5

Flap OP-Code No . No . No .
Sy mbolic or or M err
Mn emonic Ope ratio n C Z P/V S N H 76 54 3 210 By tes Cycl es Cycles Com ments
LO r. r'
LD r.n
r _ r'
r -n .0 0

0
0

0
0

0
0

0
0 01 r
00 r
·-
r'
110
I
2
I
2
4
7
r,
000
~ R"l!.
B
-
.·
n 001 C
LD r, (HL) r-( HL) 0 0 0 0 0 0 01 r 110 I 2 7 010 D
LD r, (IX+d) r- (IX+d) 0 0 0 0 II 0 11 10 1 3 5 19 011 E

-
01 r 110 100 H

LD r, ( IY+d ) r - (IY+d ) 0 0 0 0
d
0 0
-
II III 101 3 5 19
10 1
III
L
A

-
0 1 r 1 10

LD (HL), r (HL)- r 0 0 0 0 0 0
- d
0 1 110 r 1 2 7
LD (IX+d), r (IX+d )- r 0 0 0 0 0 0 II 0 11 101 3 5 19

-
01 110 r

LD (IY+d), r (IY+d) -r 0 0 0 0 0 0
d
I I II I 10 1
- 3 5 19

-
0 1 110 r
- d

-
LD (HL), n (HL)- n 0 0 0 0 0 0 00 110 110 2 3 10

LD (IX+d), n (IX+d ) - n 0 0 0 0 0 0
- n
I I 0 11 101 4 5 19
00 110 110

-- -
d

LD (IY+d). n (IY+d) - n 0 0 0 0 0 0
n
II I I I
-
10 1 4 5 19

--
00 110
d
n
110
-
-
LD A, (DC) A-( BC) 0 0 0 0 0 0 00 00 1 010 I 2 7
LD A, (DE) A - (DE) 0 0 0 0 0 0 00 0 11 0 10 I 2 7

-- --
LD A. (nn ) A -(nn ) 0 0 0 0 0 0 00 III 0 10 3 4 13
n
n
LD (BC), A (BCl - A 0 0 0 0 0 0 00 000 0 10 I 2 7
LD (DE), A (DEj-A 0 0 0 00 0 10 010
0 0 0 I 2 7
LD Inn), A l nn) -A 0 0 0 0 0 0 001 100 10 3 4 13
n - -
LD A, I A- I 0
n
I IFf' I 0 0 II 101 101
- - 2 2 9

LD A,R A -R 0
I 01 010 III
I IFF I 0 0 II 10 1 10 1 2 2 9

·
01 0 11 II I

.
LDI ,A I -A 0 0 0 0 0 II 101 101 2 2 9
01 000 III
LD R, A R- A 0 0 0 0 0 II 101 101 2 2 9
01 00 1 III

No les : r, r' means any o f the registe rs A. B, C, D, E, H, L

IFF the co nte nt o f t he interrupt enab le flip-fl op (IFF) is co pied into the P/V !lag

fla B Notation : • = nag no t affec ted , 0 = flag reset, I = n ag set, X = nag is unknown,

; = nag is affected according to the result o f the oper at ion .

8-bit l oad group

Table C.2 Courtes y Zi l og , Inc .
106 Programming in Z80 Assembly Language

F n..r... No. No. No.

Symbolk 01 oIM oIT

· ·· ··· --
C .Pt" S N H 76 54) 110
...
MnftMH&ic Opcraltc. 1)'leli ",<I.. Sbl. Commeatl

--
t Odd.,nn dd ~nn 00 ddO 00 1 3 3 10 PIliI
n 00 Be

· ·· ··· -
n 01 DE
to IX,nn IX - nn II 011 10 1 • • " '0 HL

--
00 100 001 11 SP

······ --
n
n
LO IY,n n I Y-nn 11 III 101 • • "
--
00 100 001

·· ··· · --
n
n

--
LOU L ,tnn) H - (n n+ l ) 00 101 010 3 S I.

· · ·· ·· --
l-(nn) n
n
t o dd, tnn ) dd H"" (nn +l )
dd ....(nn )
I I 101 101
0 1 ddl 01 1
• • '0

-
L

···· ·· -- -
n
n
to IX. ( fi n ) IX - (nn+l)
H
IX - ( n n )
II 01 1 101
00 101 010
• • '0

-
L

-
· ·· ··· -
n
n
to IV, (n n ) IV - (nn+ 1)
H II I I I tO I • • '0

-- --
IYl - (n n ) 00 101 0 10
n

··· ··· -- -
n
lO (nn ) , HL (nn+ l) .... H 00 100 010 3 S I.
-
{nnl - l n

·· ····
n
t o (nn),d d IM+I)-dd
{nn ) .... dd
H
11 101 10 1
01 ddO 0 11
• • 20

- -
L

- -
n

·· ··· ·
n
LO Inn), IX lnn+ l ) - I X
H
11 011 101 • • '0

- -
(nn) -IX 00 t OO 010
L

· · ···· -
n
-
·
n
lD(n n}, IY (M +D -IV
H
II I II 101 6 '0

- --
I MI -IY 00 100 0 10
-
L
n

······
n

, ,
······
t o SP. HL SP - HL 11 III 001 I I s
i o SP, IX SP-IX 11 011 ) 01 10

, ,
··· · ··
11 111 00 1
t o SP, IV SP- IV II III 101 10

······
11 111 00 1 qq Poi,
PUSH qq (SP- 2) - qql It qqO 101 I 3 11 00 Be

,
······
(SP-I ) - qqH 0' DE
PUSH IX (5P- 2) - IX II 01 1 101 4 IS 10 HL

,
L

······
(5P-1) - I X II 100 10 1 II AF
H
PUSH IY (SP- 21 -I Y II III 10 1 4 IS
L

······
151-I )-I Y II 100 10 1
H
POP qq q qH - (SP+ I) II qqO 001 I J 10

,
······
qq l - (SP)
pop IX IX -~SP + 11 II 011 ' 0\ 4 14
H

,
·· · · ··
IX - (SP)
l
II 100 OOL
POPIY IY -(S'+! )
H
IY -(SP)
II 111 101
11 100001
• "
L

No tes : dd is any of t he repU e' pain BC, DE, HL , SP

qq is an y of the reJi ster pai n AF, BC, DE , HL
(PAI R) H' (PAI R) L refer to hilh order and lo w or de r e ilh t bits of the relute, p air respec t ively.
E... Be L c C, AFH " A

Ftaa No ta nce: • c flaa nat aff ect ed , 0 c fll a reset , I "fla, set . X • flaa is unk nown ,
t flaa is af fec ted acco rd ina t o the rescn o r the o pe ratio n.

16-bit load group

Table C.3 Courtesy Zilog, Inc.
 Appendix C: Summary of Z80 Instructions 107

Fla gs Op.code
No. No. No .
Symb oli c ~ 01 ol M 01T

·· ·· ·· ·· ··· ···
····
Mnem o n ic Ope ration C Z V S N II 76 543 21 0 B y t~$ Cycles Sta tes Commen ts
EX DE . ilL DE · · II L II 10 1 0 1 1 I I 4
EX AF . A F' AF · · AF' 00 00 1 0 00 I I 4
EXX
(Rv{}iD I I 0 11 0 0 1 I I 4 Regis te r bank an d
au x iliary regist er

·· ····
ban k ex ch ange
EX ISP), II L 11- (SP + I) II 100 0 11 I 5 19

· · ·· · ·
1 - (SPI
EX (S P), IX IX - (SP+ II II 0 1 1 101 2 6 23
II

· · · ···
IX ... (Sll) II 100 Oil
L
I· X ,S PI, I" IY Il ~ · ( SP+ II II I I I 10 1 2 6 23
IY - (S I') II 1110 01 1
L

·· ·
Q)
W I ( I.)F)- I I Il.1 I II 0 I I 10 1 10 1 2 4 16 Load ( H L) in to
(DE), incremen t the
UE - DE. I III 100 000
pointer s.and
Ill ..· II L+ I decrement the by te
Be ..· IH.·· 1 counter UK )

LDIH WFI -t IlLJ

IlF ..-· t>HI
II I. ... 111..1
·· · II II 0 I I 111 1 101
10 1111 000
2
2
5
4
21
16
If 8 (' ¢ 0
Ir o c =0

He - HC· l
Rq ' l'i1t unl it
Il l" '" U

·· ·
Q)
I DO IUI-I ..- tl lL l I II 0 I I ttl I 10 1 2 4 16
Ill · · 1"' · 1 III 101 000
II l · · II I.·1
Bl ' " HC- )

L OD R (I>U "-l ll l.l

Il E - 1ll-· 1 ·· · II II 0 I I lUI 10 1
10 III 000
2
2
5
4
21
16
If OC ¢ 0
IfH C =0
111.· 111 · 1
11(' - 8 (' · 1
Re p ea l un ul
H( =0
C) Q)
( PI A - (Ii Ll
1Il - II L+ 1 · I 1 I 1 I 11 101 101
10 10 0 00 1
2 4 16

HC - 8 (' - 1

·
C) I
CPH< ,, - UIl ) I I I I I I I 10 1 10 1 2 5 21 If BC -j:. 0 and A ,. (H L)
In - B lot' 10 110 00 1 2 4 16 If 8 (' = 0 or A =(il L)
Il l " 0 - U<,'·I
RI,.'p ,·.)1 un nl
,\ := (l ll . l m
He = II

·
C)IQ;
(' I' D A - (111., I I 1 I I I I 10 1 10 1 ! 4 16
II I -· 11\ · 1 10 10 1 0 01
B(' - 0 (' · 1
I~
·
I
CPOH A - (111) I I I I I II 101 10 1 2 5 21 IfBC " 0 and A .. (II L )
II l .. II L· 1 10 III 00 1 2 4 16 If 8 C = 0 or A =(ilL)
B(' - B(' · I
Rep eat 11111,1
A =lULl o r
0(' =0

No tes: CD p." fla g i ~ U if the resu lt of H<.'-1 =O. o the rwise P/V = I
Q) I fl a ~ I .. 1 if A =lllll . o the rwise Z =(I .

H iI~ No ta tion : _ = l1 a ~ ne t af fec ted . (l = 11:lg re set . I : flag \CI, X '= tlag 1\ u nkn ow n,
J ;:. n at! i.. a fft= ~'h: d acco rdin g to th e rc..u lt o f the ope ra tio n.

Excha nge grou p and block transfer and s earch group

Tabl e C.4 Court esy Zilog, Inc .
108 Programm ing in Z80 Assembly Language

r lags O p-Cod e
No . No . No.
Sy mbo lic ~ of of M ofT
Mnemo nic Op era lio n C Z V S N H 76 543 2 10 Byt es Cy cles Sta les Com me nts

ADDA , r A -A + t I I V I 0 I IO IQQQ] r I I 4 r Reg.

000 B

- IQQQJ -
ADD A, n A - A +n I I V I 0 I II IQQQ] 110 1 1 7
00 1 C
n 0 10 D
ADD A,(HL) A · A+ (HL) I I V I 0 I 10 010 1 2 7 0 11 E
ADD A,(IX+d) A-A + (lX+d) II 0 11 101 3 5 19 100 1/
I I V I 0 I
101 L
IQQQ]
I
10 110

-
II I A
d -
ADD A,(IY+d) A -A +(JY+ d ) I I V I 0 I 11 I I I 10 1 3 5 19
10IQQQ] 1 10

ADC A,s A ..... A + s + CY I I V I 0 I

- @] - d
~ 1\ any o f r, n .
SUB , A -A - 5 I I V I 1 I IQIQJ IIIL,. /I X- d).
I IY +d J a-, ..ho wn I 0 '
SBC A, s A "'" A - s - CY I I V I 1 I lIT!II A D I) i n xt ruc uo n
AND s A-A A s n I P I 0 I o::lli!l
OR s A-A v s 0 I P I 0 0 [illjJ Th e mdi ca tcd bu v
[illjJ repl ace the OGO in
XOR s A- A . 5 (l I P I 0 0
11ll' ADD \0: 1 abo ...
WIJ
··
CP s A -s I I V I I I
I NC r r - r+ 1 I V I 0 I on r l:IQ!il 1 1 4
IN C (HL) I on I 10 [[]ill]

·
CIlL) - (HL)+ I I V I 0 1 3 II
I NC (I X+d J n X +d ) ... I V I 0 I J 1 0 1 1 101 J b 2.1
(I X + d )+l
UU 11n[@ill

·
d
I NC HY +d) HY+d ) - I V I 0 I II III WI ] b 23
lIY "d ) + I
on ' I I U lTIE!l
- '.
·
d
DEC m m.. m-l ; V I I I lTIill m is any of r, (HL ),
f lX+d) .lIY +dIJ \
sno wn for INC
Same f orm al .1l1lJ
\ I;l t~·\ a\ I N C
K,' p l.I" " 100 wi th
111 1 III UP dKh:

NOles: Th e V sy mbol in urc P/ V O••g c o lumn indrca t c-, th at th e It V na~ co n t arn \ th e ove r flow ot thc rv...ul t 01 th.;
opera tion Simil ar ly t he P sym b o l ind u-atc x pan rv. V =: J mcunv ove rflow . V =0 mean ...llll l l l\' l ' r ll i llol. P =:I
mean s pa rit y of th e re..u lt is eve n, p ..: 0 rncauv p ant y o t th c rcvul t ,...odd ,

Fl ag Notat ion : _ = flag no t affec ted . 0 :: n ag rc-et , I -= na t! <rt . X ;: n at! IS un kn ow n .

t ;: flag hi affe c ted acco rdi ng 10 the tv.. iil l o f the o pc rauon

8-bit a r i t hme t i c and logical group

Tabl e C.S Courtesy Zilog, Inc .
 Appendi x C: Summary of Z80 Instructions 108

FlaS' Op-{:o<Ie
No. No. No.
Symbolic ~ of ofM ofT
Comments
Mn emonic Operation C Z V S N H 76 543 210 Byte s Cycl es Sta tes

DAA Converts ace.

content into
p ack ed BCD
I I P I
· I 00 100 111 I I 4 Decimal adjust
acc u m u lator

foll owing add

or subtract
with packed

· · ··
BCD operands
CPL A-A I I 00 101 III I 1 4 Complement
accumulator
(one 's complemen t)
NEG A-O -A I I V I I I I I WI 10 1 2 2 8 Negate ace. <two 's
comple me nt)

···
0 1 000 100
CCF Cy - a I 0 X 00 III I II I I 4 Complement carry

·· ····
flag
4 Set carry flag

· ··
CY - I I 0 0 00 110 III I I

··
SCF
4

· · ····
NOP No ope ration 00 000 00 0 I 1

··· 01 110 110 I I 4

······
HALT CPU ha lted
4

· ······
DI IFF-O II 110 OIl I I

EI IFF - I II III 011 I I 4

IM O Set inte rrupt 11 WI 101 2 2 8

·· · ···
mode 0 01 000 110
IMI Set interrupt II 101 10 1 2 2 8

· ·· · ··
mode 1 01 010 110
1M2 Set interrupt II 101 101 2 2 8
mode2 0 1 OI l 110

No tes : IFF indicates the interrupt enable flip-flop

CY indicate s the carry flip-flop .

Flll Notation: • = flag no t affected. 0 = flag reset . 1 = flag set, X = flag is unkno wn.
*= flag is affec ted according to the result of the operation .

General pur pose a r i t hme t i c and CPU control gro ups

Tabl e C. 6 Cour tesy Zi log , I nc.
1 10 Program ming in Z8 0 Assembly Language

Op-Code No . No. No.

Symbolic
Flag.
or orM err
Mnemonic Operation C Y N H 76 543 210 Bytes Cycles States Comments

···
Z I" S
ADDH l , " HL - HL+ss I 0 X 00 ..1 001 I 3 II II Rei ·
00 DC
01 DE
ADCH l , .. HL-HL+ $I +CY t I V I 0 X II 101 101 2 4 15
10 Hl
01 ..1 010 II SP
s ac Hl , ,, Hl-H l· .. -cv I I V I I X 11 101 101 2 4 15

· ··
01 ..0 010
AD D IX, PP IX -IX + PP I 0 X 11 0 11 101 2 4 15 PP Reg.
00 pp l 001 00 DC
01 DE
10 IX

·· ·
II SP
AD D I Y.n I Y-IY+rr I 0 X I I III 101 2 4 15 rr Reg.
00 rr1 001 00 IlC
01 IJE
10 IY
II SP

INC .. SS ... IS + I
·· · · · · 00 ssO 0 1 I I I b

INC IX IX - IX + I
······ II 0 11 10 1 2 2 10

·· · ···
0010001 1
INC IY IV -IY + 1 II III 101 2 2 10

······
0010001 1
DEC ss 55'" 5S- ) 00 ..1 0 11 , I I 6

IJEC IX IX - IX · I
··· ··· I I 011 10 1 2 2 10

······
00 101 011
DECIY IY- IY ·I II III 101 2 2 10
00 101 011

Note s: 55 is any o f the regis ter pairs BC, DE, HL. SP

pp IS any of the register pairs Be. DE. IX. SP
11 i\ an y of the register pain Be, DE. IV. SP.

folll No tatio n: . :: l1a~ net ..tf ccted , 0:: I1lg reset , 1 :: nag set , X = flag 1'\ unknown .
t = fhg i.. aff ect ed acco rding to the resul t of the opera tion.

16-bit arithmetic group

Table C.7 Courtesy Zilog, Inc.
 Ap pendix C: Sum mary of Z80 Instructions 111

Flags Op-Code
No. No. No.
Symbolic ~ 01 aIM olT

···
Mnemonic Operation C Z V S N H 76 543 210 Bytes Cycles States Comments
RLCA I 0 0 00 000 III 1 I 4 Rotate left circular
~ accumulator

RLA
~ I
·· · 0 0 00 010 III I I 4 Rotate left
accumulator

RRCA
~ I
··· 0 0 00 00 1 I II I I 4 Rotate right circular
accumulator

RRA
~ I
·· · 0 0 0001 1 III 1 1 4 Ro tate right
accum ulator

RLCr I I P I 0 0 11 001 011 2 2 8 Rotat e left circular

OO[QQQ]r register r
RLC (HL) I I P I 0 0 I I 001 011 2 4 15 r ~I·
00 [QQQ] 110 000 B
RLC (IX'd) ~ I I P I 0 0 11 0 11 101 4 6 23 001
010
C
D

- -
•. \ HU . lI l1;....' . fIY••l!
11 001 0 11 011 E
d 100 H
00 [QQQ] '10 101 L
111 A
RLC (IY+d) I I P I 0 0 11 111 101 4 6 23

- -
11 001 011
d
00 ~ 1l 0
RL m
~ m '! • • IHLl . IIX"" I. lI YOdl
I I P I 0 0 [ill]] Instruction format an
states are as shown
for RLC,m. To form
§ OP-code replace
RRC m ~m < r,(H l I.lI X. c1I, Il Yodl
I I P I 0 0 lQill of RLC,m with
shown co de

RR m ~ m ~ r. f H LI . (I X . 4 1 . I I Y " 1
I I P I 0 0 (QITJ

SLA m ~o I I P I 0 0 [iN]
. " I. l HLI.I IX"<l'. llY · c1,

SRA m ~", · • . IHlI. lIX. ol" f1Y· c11

I I P I 0 0 llQI)

SRLm ~
" ",. I HLI, f IX· c1 1. / IY. "
I I P I 0 0 lIITl

RLD '~'"U . I P I 0 0 11 101 101 2 5 18 Rotate disit left and

.
01 101 III right between the
accumulator
and location (HL).
RRD
'~'"U I p I 0 0 11 101 101
0 1 100 III
2 5 18 The conten t of the
upper half of the
accumulator is
unaffected

All Notation : -,. tlag not affected, 0 = Oag reset, 1 = Oagset, X = flag is unknown.
; = nag is affected according to the result of the operation.

Rotate and shift group

Table C.8 Courtesy Zilog, Inc.
11 2 Programming in Z80 Assemb ly Language

Flap Op{"ode
1"0. No. No.
Symbolic: ~ oC
Bytes
orM
Cycles
oCT

·
Mnemonic Opention C Z V S N H 76 543 210 States Comments

BITb. , Z-rb I X X 0 I 11 001 011 2 2 8 , JleI.

·
01 b r 000 B
BITb,lHLl Z-IHL\. I X X 0 I 11 001 011 2 3 12 001 C
010 D

·
01 b 110 011 E
BITb, (IX+d) Z-(iX+djb I X X 0 I 11 011 101 • 5 20 100 H

- -
11 001 011 101 L
111 A
d

·
01 b 110 b Bi.Teatecl
BIT b, (IY+d1 Z -IIY+dlb I X X 0 I 11 111 101 4 5 20 000 0

- -
11 001 011 001 I
010 2
d 011 J
01 b 110 100 4
IOJ S
110 6

·· · ·· ·
111 7
SETb" 11 001 011 2 2 8
" -I

···· ··
IITIb r
SETb, (HLI IHLlb -I 11 001 011 2 4 15

· ·. ·· ·
(j]b 110
SETb,IIX+dl (IX+d)b - I II 011 101 4 6 23

- -
II 001 011
d

·· ··· ·
(j]b 110
SETb. (IY+dl IIY+dl b - I 11 III 101 4 6 23

- -
11 001 011
d
IIiJb 110

RES b , m 'b -0 [ill To fonn new OP·

m=',(HL l , code repla ce IIIl
(IX+dl. or SET b,m with
(IY+dl [QJ. Flags and time
sta tes for SET
instruction

Not..: The notatlon 'b indicate. bil b (0 10 71 or Iocatlon L

FIaa Notation : • = n., not affected. 0 • flag reset . I ... flag ~ 1. X • flag is unknown.
* = flag i' affected according to the result or Ihe operation .

Bit set, reset and test group

Table C.9 Courtesy Zilog, Inc.
 App endix C: Summa ry of Z80 Instructions 113

Flags Op-Code
No. No. No.
Symbo lic: ~ of ofM ofT
Mnemonic Operation C Z VS N H 76 543 210 Bytes Cycles Stales Comments

If nn PC-nn
······ - - II 000 OIl
n
3 3 10

- -
· ··· · · .- -
n cc Condi tion
If cc, nn If conditio n cc I I cc 0 10 3 3 10 000 N Z non zero
is true PC -enn , 001 Z zero

-
n
o therwise
continue - n
010
Oil
100
Nen oo carry
C carry
PO par it y odd
101 PE parity even

····· · -
110 P sign pc suwe
JR e PC -PC+e 00 01 1 000 2 3 12 III M sign negative
-
······ -
<-2
JR C, ' If C " O. 00 III 000 2 2 7 If co ndition n ot met
continue
IfC: I ,
.-2 - 2 3 12 If condit ion is met

···· · · -
PC ..... PC+e
JR NC•• IfC: I. 00 110 000 2 2 7 If c o nd itio n not mel
cont inue
If C: 0,
.-2 - 2 3 12 If cc nduion j.. met

· · ·· ·· -
PC -PC+ e
JR Z.< If Z: 0 00 101 000 2 2 7 I f condit io n not met
con tinue
IfZ = I.
<-2 - 2 3 12 If co ndition h me t

· · ·· ·· -
PC ....PC+e
JR NZ, < If Z = I. 00 100 000 2 2 7 If co nditi o n n OI m,
cont inue
IfZ: O.
<-2 - 2 3 12 If con di tio n me l
PC ....P('+e
l P (HL) ('( '-HL
······ II 101 00 1 I I 4

JPOX) ('('-IX
······ II 0 11 101 2 2 8

·· · · · ·
II 10 1 00 1
JPO Yl PC-IY II III 101 2 2 8
II 101 001

DlNZ.e 0-0-1
If 0 =0,
continue
······ - 00 010 000
<-2 -
2 2 8 Ir B =0

If 0 "0, 2 3 13 IF B _ 0
PC-PC+e

Notes: e represen ts the exten sion in the relative add ressing mode .
e is a signed two's complement number in the range <-126. 129>
e-2 in the op-code provides an effective address of pc +e as PC is
increment ed by 2 prior to the additio n of e.

Flag Notation : - = flag not affe cted, 0 = flag reset , t :: flag set, X = flag is unknown .
S = flag is affec ted acco rding to the result of the operation.

Jump group
Table C.l O Co urtesy Zilog . Inc.
114 Programming in zao Assembly Language

Flap OP<:OO.
No. No. No.
Symbolic: ~ ol olM ofT
Mnemonic Opencion C Z V S N H 76 543 210

···· · · - -
Byt.. Cycles States Comments
CAll nn (SP.IH'C 11 001 101 3 5 17
H

-" -
(SP.2)-PC n
l
PC-nn

CALL cc, nn If condition

·· · ·· · - " - 11 ee 100 3 3 10 If cc is false

-" -
cc is false
continue .
otherwise 3 S 17 If cc is true
sameas

· ·· ·· ·
CAll nn
RET PC -(SP) 11 001 001 I 3 10
l
PCH-{SP+I)

RETee If condition
cc is false
continue ,
otherwise
·· ·· ·· 11 ee 000 1

1 3
1 5

11
If cc is false

If cc is true
sameas ee Condition
RET 000 NZ non zero

····· ·
001 Z zero
010 NC non carry
RETI Return from 11 101 101 2 4 14 011 C carry
interrupt

······
01 001 101 100 PO parity odd
RETN Return from 101 PE parity even
11 101 101 2 4 14 110 P
non maskable sign positive

· ··· · ·
interrupt 01 000 101 111 M sign negative
RSTp (SP.I)-PC 11 t 111 1 3 11
H
(SP-2)-PC l
PCH-O
PCl -P
t P
000 OOH
001 08H
010 IOH
011 18H
100 20H
101 28H
110 30H
111 38H

fla. Notation: • = flag not affected, 0 =flag reset. I = nag set, X = nag is unknown
S = flag is affected according to the result of the operation.

Call and return group

Table C.II Courtesy Zilog, Inc.
 Appe ndix C: Summary of Z80 Instructions 11 5

FollgS Op-Code
No. No. No.
Sy mbo lic: ~ 01 ol M ofT

·· ·· · · - -
Mnemonic Opera tio n C Z V S N H 76 543 210 Byt es Cycles Stales Co m me nts

IN A.(n) A - In) II Oil 0 11 2 3 II ntoAo-A,

·
n Ace to AS - A 1S
IN , , (e) , - ( C) I P I 0 I I I 101 101 2 3 12 Cto A - A
ifr = 110 only , 01 000
O 7
BtoA - A I S
S
the flags will
be affected
'\'
·
'-"
INI (HL) -(C) I X X 1 X I I 101 101 2 4 16 C to A - A7
O
8 - B· 1 10 100 0 10 BtoAS- A
IS

·
HL-HL+ 1
INIR (HL)- (C) 1 X X 1 X 11 101 101 2 5 21 Ct OAO - ~
B-B ·1 10 110 010 (118"'0)
8 '0 AS - Al 5
HL - HL + 1 2 4 16
Repeat until (118 = 0)
8 =0

G::
IND (HL) -(C)
B- B -1 · I X X I X I I 101 10 1
10 101 010
2 4 16 C to A - A
O 7
B to AS - A I S

·
HL-HL - 1
INDR (HL) -(C) I X X I X 11 101 101 2 5 21 Ctu A - A
O 7
8 - 8-1 10 II I 010 (lfB "'0) 8 to AS - A I S
HL -HL -I
2 4 16
Repeal until (lfB = 0)
8=0
Otf T (n), A (n) _ A
· · ···· 11 0100 11 2 3 II n to AU - A
7

· · · ···
. - n -+ Ace to AS - A l 5
OUT( C),r (C)-, 11 101 10 1 2 3 12 C ta A .... A,
O
0 1 r 001 B '0 AS - A 1S
CD
OUTI (C)-( HL)
8 -B-1 · I X X I X 11 101 101
10 100 OI l
2 4 16 (' to AU "" A,
B to As .... A
1S

·
HL - HL + I
OTIR (C) -( HL) I X X 1 X 11 101 101 2 5 21 ( '1 0 AU .... A
7
8 - B- 1 10 110 0 11 (If 8'" 0) 8 t o As "" A
1S
HL - HL + I
2 4 16
Repeat until (If 8 = 0)
8- 0

CD
OUTD (C) - IHL)
8 - 8 -1 · I X X I X I I 101 101
10 10 1 Oil
2 4 16 Chi AO- A,
8 (0 AS - A
1S

·
HL - HL - I
OTDR (C) - IHL) I X X I X 11 101 10 1 2 5 21 c ro AU - A7
8-8 -1 10 111 0 11 (If 8 '" 0) 8 t o A g .... A
15
HL-H L -I
2 4 16
Repeat until (l f B = 0)
8=0

No tes: CD If the result o f B-1 is zero the Z flag is set . o therwise it is reset .

Flag Not ation: • = nag not affe cted , 0 = flag reset, 1 "" flag set , X = nag is unknown .
t "" nag is affe cted according to the result of the operation.

Input and output group

Table C.12 Courtes y Zilog, I nc.
Appendix D: Display and Keyboard
Character Codes

The ASCII code character set is shown below

b7 --+ 0 0 0 0 I I I I
b6 --+ 0 0 I I 0 0 I I
b5 --+ 0 I 0 I 0 I 0 I
b b b
4 3 2 I b~+ 0 I 2 3 4 5 6 7
,
0 0 00 0 NUL OLE SP 0 @ P p

0 0 01 I SOH DCI I I A Q a q
II
0 0 10 2 STX DC2 2 B R b r
0 0 1I 3 ETX DC3 ~ 3 C S C 5

0 10 0 4 EOT DC4 $ 4 0 T d t

0 101 5 ENQ NAK 0/0 5 E U e u

0 I I 0 6 ACK SYN a 6 F V f v
I
0 I 1 I 7 BEL ETB 7 G W 9 w
10 0 0 8 BS CAN ( 8 H X h x
10 01 9 HT EM ) 9 I Y i y

··
10 10 10
10 1 I II
LF SUB
VT ESC *+ ·,
J
K
Z
[
j
k
z
{
I 10 0 12 FF FS , < L <, I I
I

I 101 13 CR GS - - M ] m }
I I I 0 14 so RS . > N " n rv

I I I I 15 Sl us / ? 0 - 0 DEL

For example, the ASCII code for the character K is binary

1001011, hexadecimal 4B and decimal 75 .

116
Appendix E: Expression Operators

The following table lists these operators which may be used in

an operand expression. The list is in order of precedence of
evaluation .

OPERATOR FUNCTION

+ Unary plus
Unary minus
.NOT. or \ Logical NOT
.RES. Result
** Exponentiation
*/ Multiplication
Division
.MOD. Modulo
.SHR. Logical shift right
.SHL. Logical shift left
+ Addition
Subtraction
.AND. or & Logical AND
.OR. or A Logical OR
.XOR. Logical XOR
.EQ. or Equals
.GT. or > Greater than (signed)
.LT. or < Less than (signed)
.UGT. Unsigned greater than
.ULT. Unsigned less than

The Result operator (.RES.) causes overflow to be suppressed

during evaluation, so that an assembly error does not result
from an overflow condition.

The Modulo operator (.MOD.) is defined as

.MOD.B = A-B*(A/B)
where the A/B is an integer division.

The Shift operators ( .SHR. and .SHL.) are followed by two

arguments. The first argument is shifted by the number of bits
specified by the second argument.

The five comparison operators (.EQ., . GT. , .LT., .UGT. and . ULT. )
evaluate to logical TRUE (all ones) if the comparison is true,
and a logical FALSE (zero) otherwise.

117
Index

ADC 13.2, 13.4 END 3.4

ADD 8-bit 2.2, 2.3 ENDC 11.5
16-bit 13.2, 13.3 EQU 4.4
AND 11.1, 11.2 Exchange instructions
Auxiliary registers 16.2 registers 16.2
stack 6.4
BCD arithmetic 15.2 Expressions 9.5, E
instructions 15.3 Extended addressing 2.7
numbers 15.1
Binary numbers A.l, A.2 Flag register 5.1
BIT 9.1
Block search instructions 14.2 HALT 2.8
Block transfer instructions 14.1 Hexadecimal numbers A.l, A.2
Bytes A.7
Immediate addressing 2.7
CALL 3.2, 7.5, 8.5, 12.5 extended 7.2
Carry 8.1 Implied addressing 7.2
flag 8.2 INC 8-bit 2.5
Character codes 3.3, 4.3, D 16-bit 7.4, 13.3
Comments 2.8 Index registers 9.4, 13.3
COND 11.5 Input from keyboard 4 . 2
Conditional jump 5.2 instructions 16.3
Conditional pseudo operations 11.5 numbers 6.2
Conversion - numbers A.3 to A.6, B Interrupt instructions 16.4
CP 5.3
CPD 14.2 JP conditional 5.2, 8.2, 8.4,
CPDIR 14.2 12.5
CPDR 14.2 unconditional 4.1
CPI 14.2 Jump tables 9.6
CPL 11.2
Labels 2.8
DAA 15.3 LD 2.1, 2.4
DEC 8-bit 2.5, 6.2 LDD 14.1
16-bit 7.4, 13.3 LDDR 14.1
DEFB 2.8 LDI 14.1
DEFL 11.5 LDIR 14.1
DEFM 7.3 Logical operations 11.1
DEFS 9.3 Loops 4. 1, 5.4, 6. 1, 7. 1, 13. 3
DEFW 13.1
Division 10.4 Macros 11.4
DJNZ 6.1 Masking 11.3

118
 Index 119

Memory 1.3 Relative addressing 7.2

Modular programming 4.2 RES 9.2
Multiple byte arithmetic 13.4 RET 3.2, 7.5, 8.5, 12.5
Multiplication 10.4 RLD 15.4
Rotate i ns t r uc t i ons 12.1, 12.2
NEG 2.6 RRD 15.4
Nibble 15.1
NOP 16.1 SBC 13.2, 13.4
NOT operation 11.1 SET 9.2
Sign flag 5.1
OR 11. I, 11. 2 Signed numbers A.8
ORG 3.4 SLA 10.3
~utput to display 3.3 SRA 10.2
of text 7.4 SRL 10.1
instructions 16.3 Stack organisation 6.3
Overflow 8.3 instructions 6.4
flag 8.4 SUB 8-bit 2.2, 2.3
16-bit 13.2
Packing 12.3 Subroutine concepts 3.1
Parity 12.4 mechanism 7.5
flag 12.5
POP 6.4 Unconditional jumps 4.1
Pseudo operations 3.4,11.5 Unpacking 12.3
PUSH 6.4
XOR 11. 1, 11. 2
Register addressing 7.2
indirect 7.2 Zero flag 5. 1
Regis ters 1. 2