|
September 1966 Radio-Electronics
 [Table of Contents]
Wax nostalgic about and learn from the history of early electronics.
See articles from Radio-Electronics,
published 1930-1988. All copyrights hereby acknowledged.
|
Digital logic has been
a fundamental part of electronics circuit classes for many decades - even in the
late 1970s when I began training in the USAF. It really wasn't until the 1960s -
about the time this "Computer Talk" articles appeared in Radio-Electronics magazine,
that terms such as "and," "or," "not," "nand," and "nor" - the language of Boolean
logic - became common parlance to the general electronics world. Many circuits implementing
the Boolean combinatorial functions were in existence, but people did not generally
refer to them by those terms. "Switching circuits," "relay circuits," and "ladder
circuits" were the lingo of the pre-digital era. Inputs typically drove diode, triode,
or pentode tubes (and later transistor) to implement the required functions. In
the early 1960s, Texas Instruments and Fairchild Semiconductor introduced some of
the first integrated circuit logic gates. DDL (diode-diode logic), DTL (diode-transistor
logic), and RTL (resistor-transistor logic) preceded TTL (transistor-transistor
logic) dominant in today's devices.
"And/Or ... Nand/Nor" ... Computer Talk
Fig. 1 - The AND gate can be set up to operate from as many inputs
as you would like; but every input must be active (on) before this gate will generate
an output of any kind.
Fig. 2 - In the OR gate, you can have any number of inputs; in
this logic circuit, however, an output will be produced if anyone of the inputs
becomes active (or turns on).
Fig. 3 - NAND gate circuit depends on inputs becoming inactive
(turning off or being removed). However, every input must be inactivated before
an output will be produced.
Fig. 4 - The NOR gate depends on removal of a single input. Can
have as many as is necessary, but losing just one of them is enough to cause the
gate to produce output "pulse."
Fig. 5 - This type of punch-card reader depends on lights shining
through the holes.
Fig. 6 - Headlight alarm shows how logic circuits can be turned
to everyday use. This setup takes one AND gate and one OR gate to turn on.
Fig. 7 - Solid-state version puts the same AND-OR logic principles
to work as above.
By Irwin Math, WA2NDM
Logic circuits aren't confined to computers. Here they are in everyday guise.
Automatic industrial control systems, complex computers, and a host of other
"thinking" machines use pre-planned programs to dictate their operations. These
programs tell the machine exactly what to do at each stage of an operation. Such
a program, besides commanding each step of a process, entails logical step-by-step
evaluation of exactly what effect each event has on following events.
Special circuits, called logic circuits, perform comparisons and make decisions.
Four of the elementary logic circuits are the and circuit, the or circuit, the nand
circuit, and the nor circuit. These circuits are the building blocks of complex
computers. If you know how they work, you are on your way to understanding computers.
In discussing applications, however, you'll see that these logic devices are used
in more everyday ways than you might think.
Fig. 1-a is a simple and circuit or and gate as it is usually called. Switches
and a relay are shown, to make the principle easy to understand. Switch A energizes
the lamp circuit, and switch B the relay coil. You can see that only when switches
A and B are closed can the relay close and light the lamp. Neither switch can light
the lamp alone. If you consider the act of closing a switch to be an energy input,
and the lighted lamp to be a result of output, you can visualize the underlying
operating principle of the and circuit: both input A and input B must be present
for an output.
Fig. 1-b is a semiconductor version of the and gate. Again, switch A must be
closed to supply collector voltage to the transistor, and switch B must be closed
to bias the transistor into conduction, before current can flow through the lamp.
This idea can be expanded to produce an and gate with any number of inputs. Fig.
1-c shows a four-input and gate. For the lamp to light (an output) all four of the
series-connected transistors must conduct. This requires that all switches be closed
(all inputs be present) - 1 and 2 and 3 and 4. If anyone (input) is missing, there
is no output. You can make the output (lamp lighting) depend on as many inputs as
you like.
Fig. 2-a is a simple or gate. In this circuit input A or B must be present before
you get an output. Since the two switches are in parallel, either one will energize
the relay and light the lamp. When you examine the solid-state version in Fig. 2-b,
you'll see that B+ is always supplied to the transistor so it is ready to conduct
all the time. All that is required for an output (make the lamp light) is that any
switch be closed, biasing the transistor into conduction. Expanding the number of
or-gate inputs is simply a matter of adding more switches in parallel with the ones
already present.
Up to this point, closing a switch has represented an input. In the next example,
you'll see situations where it is desirable to produce an output by opening a switch
(removing an input). These characteristics, basically the opposite of the circuits
just described, are called not-and (or nand) and not-or (nor) logic.
The not-and or nand circuit will produce an output only when inputs A and B are
not present. For instance, consider Fig. 3-a. Assume that opening a switch constitutes
an input. It is obvious that only when both switches are open will the relay contacts
close and the lamp light. The same function with semiconductors (Fig. 3-b) takes
a pnp transistor. When both switches are closed, the transistor base is essentially
tied to the emitter. With no base-to-emitter bias the transistor is cut off. Opening
both switches (not input A and not input B) removes this short and the transistor
conducts, lighting the lamp.
The nor gate (Fig. 4-a) produces an output when either switch A or switch B is
opened. When either is opened, the corresponding relay drops out, its contacts close,
and current flows to the lamp. The semiconductor version does the same thing. With
neither switch open, the bases of the pnp transistors are at emitter potential and
the transistors are cut off. When a switch is opened, however, its transistor is
biased on, current flows and the lamp lights.
While the operation of these circuits has been explained using the action of
closing and opening switches as inputs and the action of lighting lamps as outputs,
most computer circuits are more sophisticated. The usual logic circuit responds
to either voltage-level changes or to pulses of some sort.
A typical example of such a logic circuit is given in Fig. 5. This and gate is
part of one type of punch-card reader. The transistors are biased into cutoff by
the base resistors: the resistance of the photocells is very high. When a card moves
into place and the two holes on the card reach the right position, light shines
through them, illuminating the photoconductive cells. Their resistance drops to
a point where the two transistors are switched on. Current flows through the transistors
and the 1 K load resistor, producing a voltage across it. This voltage pulse is
fed to the next logic circuit where it produces another action, and so on. Whenever
the specific two holes on a card are in that particular position, there is an output
from the and gate. By selecting positions for lamps and photocells, you make the
many holes on a card represent specific data. The computer "reads" this data as
each card moves into position.
Now think about a simple but useful logic-type warning device. Consider this
situation: If you leave your automobile with the motor off and the headlights on,
the battery will discharge. How about an alarm to warn you before you leave the
car? Plan it so the passengers can leave the car without sounding the alarm. The
sources of information (inputs) can be: the ignition switch, which we will call
A; the headlight switch, B, and the door switches, C. The alarm should sound only
if A is open and B and C are closed. Any other condition should not trigger the
alarm.
Fig. 6 shows the relay approach. RY1 forms a nor gate with switch A as the input,
while RY2 forms an and gate with B and C as the input. The buzzer sounds only when
the headlights are on and one of the doors is open. B and C are closed and A is
open. The solid-state version (Fig. 7) operates in exactly the same manner. B applies
base bias for Q2, but Q2 can conduct only if Q1 does. C applies emitter voltage
to Q1, but it can conduct only if a door is open. So, with the headlights on and
a door open, Q1 and Q2 are both ready to conduct - except that a closed A holds
the base of Q1 at the same potential as the emitter and keeps it cut off. Turn off
the ignition switch and you add the final input that sets off alarm. The diodes
across the buzzer coil prevent voltage spikes generated by the buzzer coils from
damaging Q1 or Q2.
By thinking along these lines, you can construct similar and even more complex
control circuits for whatever special application you may have.
|
