Theres always a meta possible over any

existing sWhen the Clock input goes from 0 to 1, the value of D which is the
same as

! is transferre" to #, thus also changing

an" D for the next transition
of the Clock input from 0 to 1$
%f the fre&uency of the oscillator is '0 () which means '0 cycles per secon"!, the
fre&uency of the # output is half that, or 10 ()$ *or this reason, such a circuit+in
which the

output is route" back to the Data input of a flip,flop+is also known as
a frequency divider$
-f course, the output from the fre&uency "ivi"er can be the Clock input of another
fre&uency "ivi"er to "ivi"e the fre&uency once again$ (ere.s an arrangement of three
of them/

0et.s look at the four signals %.ve labele" at the top of that "iagram/

%.ll a"mit that %.ve starte" an" en"e" this "iagram at an opportune spot, but there.s
nothing "ishonest about it/ The circuit will repeat this pattern over an" over again$ 1ut
"o you recogni)e anything familiar about it2
%.ll give you a hint$ 0et.s label these signals with 0s an" 1s/

Do you see it yet2 Try turning the "iagram 30 "egrees clockwise, an" rea" the 4,bit
numbers going across$ 5ach of them correspon"s to a "ecimal number from 0 through
16/
Binary Decimal
0000 0
0001 1
0010 '
0011 7
0100 4
0101 6
0110 8
0111 9
1000 :
1001 3
1010 10
1011 11
1100 1'
1101 17
1110 14
1111 16
Thus, this circuit is "oing nothing less thancounting in binary numbers, an" the
more flip,flops we a"" to the circuit, the higher it will count$ % pointe" out
in Chapter : that in a se&uence of increasing binary numbers, each column of "igits
alternates between 0 an" 1 at half the fre&uency of the column to the right$ The
counter mimics this$ ;t each positive transition of the Clock signal, the outputs of the
counter are sai" to increment, that is, to increase by 1$
0et.s string eight flip,flops together an" put them in a box/

This is calle" a ripple counter because the output of each flip,flop becomes the Clock
input of the next flip,flop$ Changes ripple through the stages se&uentially, an" the flip,
flops at the en" might be "elaye" a little in changing$ <ore sophisticate" counters
aresynchronous, in which all the outputs change at the same time$
%.ve labele" the outputs #
0
through #
9
$ These are arrange" so that the output from the
first flip,flop in the chain #
0
! is at the far right$ Thus, if you connecte" lightbulbs to
these outputs, you coul" rea" an :,bit number$ ; timing "iagram of such a counter
coul" show all eight outputs separately, or it coul" show them together, like this/

;t each positive transition of the Clock, some # outputs might change an" some
might not, but together they reflect increasing binary numbers$
% sai" earlier in this chapter that we." "iscover some way to "etermine the fre&uency
of an oscillator$ This is it$ %f you connect an oscillator to the Clock input of the :,1it
Counter, the counter will show you how many cycles the oscillator has gone through$
When the total reaches 11111111 '66 in "ecimal!, it goes back to 00000000$ =robably
the easiest way to use this counter to "etermine the fre&uency of an oscillator is to
connect eight lightbulbs to the outputs of the :,1it Counter$ >ow wait until all the
outputs are 0 that is, when none of the lightbulbs are lit! an" start a stopwatch$ ?top
the stopwatch when all the lights go out again$ That.s the time re&uire" for '68 cycles
of the oscillator$ ?ay it.s 10 secon"s$ The fre&uency of the oscillator is thus '68 @ 10,
or '6$8 ()$
;s flip,flops gain features, they also gain in complexity$ This one is calle" an edge-
triggered D-type flip-flop with preset and clear/

The =reset an" Clear inputs overri"e the Clock an" Data inputs$ >ormally these two
inputs are 0$ When the =reset input is 1, # becomes 1 an"

becomes 0$ When the
Clear input is 1, # becomes 0 an"

becomes 1$ 0ike the ?et an" Aeset inputs of
an A,? flip,flop, =reset an" Clear shoul"n.t be 1 at the same time$! -therwise, this
behaves like a normal e"ge,triggere" D,type flip,flop/
Inputs Outputs
Pre Clr D Clk Q

1 0 B B 1 0
0 1 B B 0 1
0 0 0 C 0 1
0 0 1 C 1 0
0 0 B 0 #

The "iagram for the e"ge,triggere" D,type flip,flop with preset an" clear looks like
this/

We have now persua"e" telegraph relays to a"", subtract, an" count in binary
numbers$ This is &uite an accomplishment, particularly consi"ering that all the
har"ware we.ve been using was available more than a hun"re" years ago$ We have
still more to "iscover$ 1ut let.s now take a short break from buil"ing things an" have
another look at number bases$
Chapter 15. Bytes and Hex
The two improve" a""ing machines of the last chapter illustrate clearly the concept
of data paths$ Throughout the circuitry, :,bit values move from one component to
another$ 5ight,bit values are inputs to the a""ers, latches, an" "ata selectors, an" also
outputs from these units$ 5ight,bit values are also "efine" by switches an" "isplaye"
by lightbulbs$ The "ata path in these circuits is thus sai" to be 8 bits wide$ 1ut
why 8 bits2 Why not 8 or 9 or 3 or 102
The simple answer is that these improve" a""ing machines were base" on the original
a""ing machine in Chapter 1', which worke" with :,bit values$ 1ut there.s really no
reason why it ha" to be built that way$ 5ight bits Dust seeme" at the time to be a
convenient amount+a nicebiteful of bits, if you will$ ;n" perhaps % was being Dust a
little bit sneaky, for % now confess that % knew all along an" perhaps you "i" as well!
that : bits of "ata are known as a byte$
The wor" byte originate" at %1<, probably aroun" 1368$ The wor" ha" its origins in
the wor" bite but was spelle" with a y so that nobo"y woul" mistake the wor" for bit$
*or a while, a byte meant simply the number of bits in a particular "ata path$ 1ut by
the mi",1380s, in connection with the "evelopment of %1<.s ?ystemE780 their large
complex of business computers!, the wor" came to mean a group of : bits$
;s an :,bit &uantity, a byte can take on values from 00000000 through 11111111$
These values can represent positive integers from 0 through '66, or if two.s
complements are use" to represent negative numbers, they can represent both positive
an" negative integers in the range F1': through 1'9$ -r a particular byte can simply
represent one of '
:
, or '68, "ifferent things$
%t turns out that : is, in"ee", a nice bite si)e of bits$ The byte is right, in more ways
than one$ -ne reason that %1< gravitate" towar" :,bit bytes was the ease in storing
numbers in a format known as 1CD which %.ll "escribe inChapter '7!$ 1ut as we.ll
see in the chapters ahea", &uite by coinci"ence a byte is i"eal for storing text because
most written languages aroun" the worl" with the exception of the i"eographs use" in
Chinese, Gapanese, an" Horean! can be represente" with fewer than '68 characters$ ;
byte is also i"eal for representing gray sha"es in black,an",white photographs because
the human eye can "ifferentiate approximately '68 sha"es of gray$ ;n" where 1 byte
is ina"e&uate for representing, for example, the aforementione" i"eographs of
Chinese, Gapanese, an" Horean!, ' bytes+which allow the representation of '
18
, or
86,678, things+usually works Dust fine$
(alf a byte+that is, 4 bits+is sometimes referre" to as a nibble an" is sometimes
spelle"nybble!, but this wor" "oesn.t come up in conversation nearly as often as byte$
1ecause bytes show up a lot in the internals of computers, it.s convenient to be able to
refer to their values in as succinct a manner as possible$ The eight binary "igits
10110110, for example, are certainly explicit but har"ly succinct$
We coul" always refer to bytes by their "ecimal e&uivalents, of course, but that
re&uires converting from binary to "ecimal+not a particularly nasty calculation, but
certainly a nuisance$ % showe" one approach in Chapter :that.s fairly straightforwar"$
1ecause each binary "igit correspon"s to a power of ', we can simply write "own the
"igits of the binary number an" the powers of ' un"erneath$ <ultiply each column
an" a"" up the pro"ucts$ (ere.s the conversion of 10110110/

Converting a "ecimal number to binary is a bit more awkwar"$ Iou start with the
"ecimal number an" "ivi"e by "ecreasing powers of '$ *or each "ivision, the &uotient
is a binary "igit an" the remain"er is "ivi"e" by the next smallest power of '$ (ere.s
the conversion of 1:' back to binary/

Chapter : has a more extensive "escription of this techni&ue$ Aegar"less, converting
between binary an" "ecimal is usually not something that can be "one without a paper
an" pencil or lots of practice$
%n Chapter :, we also learne" about the octal, or base,:, number system$ -ctal uses
only the "igits 0, 1, ', 7, 4, 6, 8, an" 9$ Converting between octal an" binary is a snap$
;ll you nee" remember is the 7,bit e&uivalent of each octal "igit, as shown in the
table on the next page$
Binary Octal
000 0
001 1
010 '
011 7
100 4
101 6
110 8
111 9
%f you have a binary number such as 10110110!, start at the rightmost "igits$ 5ach
group of 7 bits is an octal "igit/

?o the byte 10110110 can be expresse" as the octal "igits '88$ This is certainly more
succinct, an" octal is in"ee" one goo" metho" for representing bytes$ 1ut octal has a
little problem$
The binary representations of bytes range from 00000000 through 11111111$ The octal
representations of bytes range from 000 through 799$ ;s is clear in the prece"ing
example, 7 bits correspon" to the mi""le an" rightmost octal "igits, but only ' bits
correspon" to the leftmost octal "igit$ This means that an octal representation of a 18,
bit number

isn.t the same as the octal representations of the ' bytes that compose the 18,bit
number

%n or"er for the representations of multibyte values to be consistent with the
representations of the in"ivi"ual bytes, we nee" to use a system in which each byte is
"ivi"e" into e&ual numbers of bits$ That means that we nee" to "ivi"e each byte into
four values of ' bits each that woul" be base 4! or two values of 4 bits each base 18!$
Base 16$ >ow that.s something we haven.t looke" at yet, an" for goo" reason$ The
base,18 number system is calle" hexadecimal, an" even the wor" itself is a mess$
<ost wor"s that begin with the hexa, prefix such as hexagon or hexapo" or
hexameter! refer to six of something$ (exa"ecimal is suppose" to meansixteen$ ;n"
even though The icrosoft anual of !tyle for Technical "ublicationsclearly states,
JDo not abbreviate as hex, J everyone always "oes an" % will too$
That.s not the only peculiarity of hexa"ecimal$ %n "ecimal, we count like this/
0 1 ' 7 4 6 8 9 : 3 10 11 1'K
%n octal, you.ll recall, we no longer nee" "igits : an" 3/
0 1 ' 7 4 6 8 9 10 11 1'K
?imilarly, the base,4 number system also "oesn.t nee" 4, 6, 8, or 9/
0 1 ' 7 10 11 1'K
;n" binary, of course, nee"s only 0 an" 1/
0 1 10 11 100K
1ut hexa"ecimal is "ifferent because it re&uiresmore "igits than "ecimal$ Counting in
hexa"ecimal goes something like this/
0 1 ' 7 4 6 8 9 : 3 2 2 2 2 2 2 10 11 1'K
where 10 pronounce" one-#ero! is actually 18
T5>
$ The &uestion marks in"icate that
we nee" six more symbols to "isplay hexa"ecimal numbers$ What are these symbols2
Where "o they come from2 Well, they weren.t han"e" "own to us in tra"ition like the
rest of our number symbols, so the rational thing to "o is make up six new symbols,
for example/

Lnlike the symbols use" for most of our numbers, these have the benefit of being easy
to remember an" i"entify with the actual &uantities they represent$ There.s a 10,gallon
cowboy hat, a football 11 players on a team!, a "o)en "onuts, a black cat associate"
with unlucky 17!, a full moon that occurs about a fortnight 14 "ays! after the new
moon, an" a knife that remin"s us of the assassination of Gulius Caesar on the i"es the
16th "ay! of <arch$
5ach byte can be expresse" as two hexa"ecimal "igits$ %n other wor"s, a hexa"ecimal
"igit is e&uivalent to 4 bits, or 1 nibble$ The table on the next page shows how to
convert between binary, hexa"ecimal, an" "ecimal$
Binary Hexadecimal Decimal Binary Hexadecimal Decimal
0000 0 0 1000 : :
0001 1 1 1001 3 3
0010 ' ' 1010

10
0011 7 7 1011

11
0100 4 4 1100

1'
0101 6 6 1101

17
0110 8 8 1110

14
0111 9 9 1111

16
(ere.s how to represent the byte 10110110 in hexa"ecimal/

;n" it "oesn.t matter if we.re "ealing with multibyte numbers/

-ne byte is always represente" by a pair of hexa"ecimal "igits$
Lnfortunately or perhaps, much to your relief!, we really aren.t going to be using
footballs an" "onuts to write hexa"ecimal numbers$ %t coul" have been "one that way,
but it wasn.t$ %nstea", the hexa"ecimal system ensures that everybo"y gets really
confuse" an" stays that way$ Those six missing hexa"ecimal "igits are actually
represente" by the first six letters of the 0atin alphabet, like this/
0 1 ' 7 4 6 8 9 : 3 ; 1 C D 5 * 10 11 1'K
The following table shows the real conversion between binary, hexa"ecimal, an"
"ecimal/
Binary Hexadecimal Decimal
0000 0 0
0001 1 1
0010 ' '
0011 7 7
0100 4 4
0101 6 6
0110 8 8
0111 9 9
1000 : :
1001 3 3
1010 ; 10
1011 1 11
1100 C 1'
1101 D 17
1110 5 14
1111 * 16
The byte 10110110 can thus be represente" by the hexa"ecimal number 18 without
your "rawing a football$ ;s you.ll recall from previous chapters, %.ve been in"icating
number bases by subscripts, such as
10110110TW-
for binary, an"
'71'*-LA
for &uaternary, an"
'885%M(T
for octal, an"
1:'T5>
for "ecimal$ To continue the same system, we can use
18?%BT55>
for hexa"ecimal$ 1ut that.s clearly excessive$ *ortunately, several other, terser,
metho"s of "enoting hexa"ecimal numbers are common$ Iou can in"icate the
numbers this way/
18(5B
%n this book, %.ll be using mostly a very common approach, which is a
lowercase h following the number, like so/
18h
%n a hexa"ecimal number, the positions of each "igit correspon" to powers of 18/

The hexa"ecimal number 3;4:Ch is
3;4:Ch N 3 x 10000h O
; x 1000h O
4 x 100h O
: x 10h O
C x 1h
This can be written using powers of 18/
3;4:Ch N 3 x 18
4
O
; x 18
7
O
4 x 18
'
O
: x 18
1
O
C x 18
0
-r using the "ecimal e&uivalents of those powers/
3;4:Ch N 3 x 86,678 O
; x 4038 O
4 x '68 O
: x 18 O
C x 1
>otice that there.s no ambiguity in writing the single "igits of the number 3, ;, 4, :,
an" C! without a subscript to in"icate the number base$ ; 3 is a 3 whether it.s "ecimal
or hexa"ecimal$ ;n" an ; is obviously hexa"ecimal+e&uivalent to 10 in "ecimal$
Converting all the "igits to "ecimal lets us actually "o the calculation/
3;4:Ch N 3 x 86,678 O
10 x 4038 O
4 x '68 O
: x 18 O
1' x 1
;n" the answer is 871,34:$ This is how hexa"ecimal numbers are converte" to
"ecimal$
(ere.s a template for converting any 4,"igit hexa"ecimal number to "ecimal/

*or example, here.s the conversion of 93;Ch$ Heep in min" that the hexa"ecimal
"igits ; an" C are "ecimal 10 an" 1', respectively/

Converting "ecimal numbers to hexa"ecimal generally re&uires "ivisions$ %f the
number is '66 or smaller, you know that it can be represente" by 1 byte, which is two
hexa"ecimal "igits$ To calculate those two "igits, "ivi"e the number by 18 to get the
&uotient an" the remain"er$ 0et.s use an earlier example+the "ecimal number 1:'$
Divi"e 1:' by 18 to get 11 which is a 1 in hexa"ecimal! with a remain"er of 8$ The
hexa"ecimal e&uivalent is 18h$
%f the "ecimal number you want to convert is smaller than 86,678, the hexa"ecimal
e&uivalent will have four "igits or fewer$ (ere.s a template for converting such a
number to hexa"ecimal/

Iou start by putting the entire "ecimal number in the box in the upper left corner$
That.s your first "ivi"en"$ Divi"e by 4038, the first "ivisor$ The &uotient goes in the
box below the "ivi"en", an" the remain"er goes in the box to the right of the "ivi"en"$
That remain"er is the new "ivi"en" that you "ivi"e by '68$ (ere.s the conversion of
71,14: back to hexa"ecimal/

-f course, "ecimal numbers 10 an" 1' correspon" to hexa"ecimal ; an" C$ The result
is 93;Ch$
-ne problem with this techni&ue is that you probably want to use a calculator for the
"ivisions, but calculators "on.t show remain"ers$ %f you "ivi"e 71,14: by 4038 on a
calculator, you.ll get 9$80443'1:96$ To calculate the remain"er, you nee" to multiply
4038 by 9 to get ':,89'! an" subtract that from 71,14:$ -r multiply 4038 by
0$80443'1:96, the fractional part of the &uotient$ -n the other han", some calculators
can convert between "ecimal an" hexa"ecimal$!
;nother approach to converting "ecimal numbers through 86,676 to hex involves first
separating the number into ' bytes by "ivi"ing by '68$ Then for each byte, "ivi"e by
18$ (ere.s a template for "oing it/

?tart at the top$ With each "ivision, the &uotient goes in the box to the left below the
"ivisor, an" the remain"er goes in the box to the right$ *or example, here.s the
conversion of 61,388/

The hexa"ecimal "igits are 1', 10, 16, an" 14, or C;*5, which looks more like a
wor" than a numberP ;n" if you go there, you may want to or"er your coffee
68,436$!
;s for every other number base, there.s an a""ition table associate" with hexa"ecimal/
+ 0 1 ! " 5 # $ % & ' B C D ( )
0 0 1 ' 7 4 6 8 9 : 3 ; 1 C D 5 *
1 1 ' 7 4 6 8 9 : 3 ; 1 C D 5 * 10
' 7 4 6 8 9 : 3 ; 1 C D 5 * 10 11
! 7 4 6 8 9 : 3 ; 1 C D 5 * 10 11 1'
" 4 6 8 9 : 3 ; 1 C D 5 * 10 11 1' 17
5 6 8 9 : 3 ; 1 C D 5 * 10 11 1' 17 14
# 8 9 : 3 ; 1 C D 5 * 10 11 1' 17 14 16
$ 9 : 3 ; 1 C D 5 * 10 11 1' 17 14 16 18
% : 3 ; 1 C D 5 * 10 11 1' 17 14 16 18 19
& 3 ; 1 C D 5 * 10 11 1' 17 14 16 18 19 1:
' ; 1 C D 5 * 10 11 1' 17 14 16 18 19 1: 13
B 1 C D 5 * 10 11 1' 17 14 16 18 19 1: 13 1;
C C D 5 * 10 11 1' 17 14 16 18 19 1: 13 1; 11
D D 5 * 10 11 1' 17 14 16 18 19 1: 13 1; 11 1C
( 5 * 10 11 1' 17 14 16 18 19 1: 13 1; 11 1C 1D
) * 10 11 1' 17 14 16 18 19 1: 13 1; 11 1C 1D 15
Iou can use the table an" normal carry rules to a"" hexa"ecimal numbers/

Iou.ll recall from Chapter 17 that you can use two.s complements to represent
negative numbers$ %f you.re "ealing with :,bit signe" values in binary, the negative
numbers all begin with 1$ %n hexa"ecimal, ',"igit signe" numbers are negative if they
begin with :, 3, ;, 1, C, D, 5, or * because the binary representations of these
hexa"ecimal "igits all begin with 1$ *or example, 33h coul" represent either "ecimal
167 if you know you.re "ealing with 1,byte unsigne" numbers! or "ecimal F107 if
you.re "ealing with signe" numbers!$
-r the byte 33h coul" actually be the number 33 in "ecimalP This has a certain appeal
to it, of course, but it seems to violate everything we.ve learne" so far$ %.ll explain how
it works inChapter '7$ 1ut next % must talk about memory$
Chapter 1#. 'n 'ssem*la+e ,- .em,ry
;s we rouse ourselves from sleep every morning, memory fills in the blanks$ We
remember where we are, what we "i" the "ay before, an" what we plan to "o to"ay$
These memories might come in a rush or a "ribble, an" maybe after some minutes a
few lapses might persist J*unny, % "on.t remember wearing my socks to be"J!, but all
in all we can usually reassemble our lives an" achieve enough continuity to
commence living another "ay$
-f course, human memory isn.t very or"erly$ Try to remember something about high
school geometry, an" you.re likely to start thinking about the ki" who sat in front of
you or the "ay there was a fire "rill Dust as the teacher was about to explain what #5D
meant$
>or is human memory foolproof$ %n"ee", writing was probably invente" specifically
to compensate for the failings of human memory$ =erhaps last night you su""enly
woke up at 7/00 ;$<$ with a great i"ea for a screenplay$ Iou grabbe" the pen an"
paper you keep by your be" specifically for that purpose, an" you wrote it "own so
you woul"n.t forget$ The next morning you can rea" the brilliant i"ea an" start work
on the screenplay$ J1oy meets girl w$ car chase Q explosionsJ2 That.s it2! -r maybe
not$
We write an" we later read$ We save an" we later retrieve$ We store an" we
later access$ The function of memory is to keep the information intact between those
two events$ ;nytime we store information, we.re making use of "ifferent types of
memory$ =aper is a goo" me"ium for storing textual information, an" magnetic tape
works well for music an" movies$
Telegraph relays too+when assemble" into logic gates an" then flip,flops+can store
information$ ;s we.ve seen, a flip,flop is capable of storing 1 bit$ This isn.t a whole lot
of information, but it.s a start$ *or once we know how to store 1 bit, we can easily
store ', or 7, or more$
%n Chapter 14, we encountere" the level,triggere" D,type flip,flop, which is ma"e out
of an inverter, two ;>D gates, an" two >-A gates/

When the Clock input is 1, the # output is the same as the Data input$ 1ut when the
Clock input goes to 0, the # output hol"s the last value of the Data input$ *urther
changes to the Data input "on.t affect the outputs until the Clock input goes to 1 again$
The logic table of the flip,flop is the following/
Inputs Outputs
D Clk Q Q/*ar
0 1 0 1
1 1 1 0
B 0 # #,bar
%n Chapter 14, this flip,flop was feature" in a couple of "ifferent circuits, but in this
chapter it will be use" in only one way+to store 1 bit of information$ *or that reason,
%.m going to rename the inputs an" outputs so that they.ll be more in accor"ance with
that purpose/

This is the same flip,flop, but now the # output is name" Data -ut, an" the Clock
input which starte" out in Chapter 14 as $old That Bit! is name" Write$ Gust as we
might write "own some information on paper, the Write signal causes the Data %n
signal to be written into or stored in the circuit$ >ormally, the Write input is 0 an" the
Data %n signal has no effect on the output$ 1ut whenever we want to store the Data %n
signal in the flip,flop, we make the Write input 1 an" then 0 again$ ;s % mentione"
inChapter 14, this type of circuit is also calle" alatch because it latches onto "ata$
(ere.s how we might represent a 1,bit latch without "rawing all of the in"ivi"ual
components/

%t.s fairly easy to assemble multiple 1,bit latches into a multibit latch$ ;ll you have to
"o is connect the Write signals/

This :,bit latch has eight inputs an" eight outputs$ %n a""ition, the latch has a single
input name" Write that.s normally 0$ To save an :,bit value in this latch, make the
Write input 1 an" then 0 again$ This latch can also be "rawn as a single box, like so/

-r to be more consistent with the 1,bit latch, it can be "rawn this way/

;nother way of assembling eight 1,bit latches isn.t &uite as straightforwar" as this$
?uppose we want only one Data %n signal an" one Data -ut signal$ 1ut we want the
ability to save the value of the Data %n signal at eight "ifferent times "uring the "ay, or
maybe eight "ifferent times "uring the next minute$ ;n" we also want the ability to
later check those eight values by looking at Dust one Data -ut signal$
%n other wor"s, rather than saving one :,bit value as in the :,bit latch, we want to save
eight separate 1,bit values$
Why "o we want to "o it this way2 Well, maybe because we have only one lightbulb$
We know we nee" eight 1,bit latches$ 0et.s not worry right now about how "ata
actually gets store" in these latches$ 0et.s focus first on checking the Data -ut signals
of these eight latches using only one lightbulb$ -f course, we coul" always test the
output of each latch bymanually moving the lightbulb from latch to latch, but we."
prefer something a bit more automate" than that$ %n fact, we." like to use switches to
select which of the eight 1,bit latches we want to look at$
(ow many switches "o we nee"2 %f we want to select something from eight items, we
nee" three switches$ Three switches can represent eight "ifferent values/ 000, 001,
010, 011, 100, 101, 110, an" 111$
?o here are our eight 1,bit latches, three switches, a lightbulb, an" something else that
we nee" in between the switches an" the lightbulb/

The Jsomething elseJ is that mysterious box with eight inputs on top an" three inputs
on the left$ 1y closing an" opening the three switches, we can select which of the
eight inputs is route" to the output at the bottom of the box$ This output lights up the
lightbulb$
?o what exactly is JWhat %s This2J2 We.ve encountere" something like it before,
although not with so many inputs$ %t.s similar to a circuit we use" in Chapter 14 in the
first revise" a""ing machine$ ;t that time, we nee"e" something that let us select
whether a row of switches or the output from a latch was use" as an input to the a""er$
%n that chapter, it was calle" a ',0ine,to,1,0ine ?elector$ (ere we nee" an 8-%ine-to-
1-%ine Data !elector/

meta