Remaining Chapter 2 topics:
So far we've looked at 3 functions in two variables, AND, OR and NOT.
| x | y | xy | x+y | x' | y' |
| 0 | 0 | 0 | 0 | 1 | 1 |
| 0 | 1 | 0 | 1 | 1 | 0 |
| 1 | 0 | 0 | 1 | 0 | 1 |
| 1 | 1 | 1 | 1 | 0 | 0 |
relate 2 variables. In fact, there are 16 possible combinations
(for each of the 4 lines, the result can be 0 or 1, meaning 2 to the
4th combinations). All 16 combinations are listed in your textbook.
Each one has a name and a particular functions, but we only will
focus on 4 additional ones: NAND, NOR, XOR, and XNOR. All 4
be derived from AND, OR, and NOT
| NAND | NOR | EXOR | XNOR | ||
| x | y | (xy)' | (x+y)' | x'y+xy' | (x'y+xy')'=xy+x'y' |
| 0 | 0 | 1 | 1 | 0 | 1 |
| 0 | 1 | 1 | 0 | 1 | 0 |
| 1 | 0 | 1 | 0 | 1 | 0 |
| 1 | 1 | 0 | 0 | 0 | 1 |

Transistors are three terminal devices which function essentially as a switch. Most of the details of how they actually work are abstracted away in this course. However, knowing a few things about transistors is useful in choosing what kind of logic devices to use in designing. The type of transistor and the way it is used to construct gates dramtically influences the performance of your circuits.
This is meant as an EXTREMELY high-level overview. Much more detailed discussions of transistors take place in ECE 320, 321, 404, 431, and 459. A slightly more detailed discussion is in chapter 10 of your text.
Transistors are created by making junctions of silicon materials that have been ``doped'' with different kinds of conducting particles, n-type (electrons), or p-type (``holes''). One junction of p- and n-type silicon is called a diode, two such PN junctions are a transistor. A transistor has 3 terminals, one on each type of silicon (either PNP or NPN). There are two fundamental types of transistors:
BJT (Bipolar Junction Transistors)
FET(Field Effect Transistors, the most common of which are MOSFETs, Metal Oxide Semiconductor FETs)
The symbol for each type is shown below with the 3 terminals labeled:

to conduct between the other two terminals, resulting essentially in this behaviour:


This circuit is an example of Resistor-Transistor Logic (RTL), the simplest way to build gates. While this approach works fine on breadboards, it is unfortunately very inefficient to build resistors in integrated circuits. To combat this, we use semiconductor material to function as resistor, replacing the resistor either with a diode (DTL) or another transistor (TTL, transistor-transistor logic). The gates in your lab kit are mostly TTL. All transistor designs are now used in almost all commercial systems. RTL and DTL are included for historical completeness, and because these technologies tend to be cyclical. New materials often require going back to the beginning. For instance, in the design of the Cray-3 supercomputer, silicon was abandoned in favor of Gallium Arsenide. Because of complications in using this new material, all the logic circuits were built using DTL. Future computing devices based on Indium Phosphide, optical switching circuits, biological materials or other new types of switches may require employing one of these currently obsolete techniques.
In addition to TTL, there are other kinds of all transistor logic. Each has its strengths and weaknesses. We compare these different kinds of logic based on several factors:
-Propagation delay: The time required for a signal to get from the input of a gate to the output. This is frequently the limiting factor in determining clock speed.
-Power Dissipation: The amount of power consumed by each gate. This power is dissipated in the form of heat, so cooling systems must be designed to compensate.
-Fan-out: This is the measure of how many gate inputs a single gate output can drive without buffering. For your TTL parts, the fan-out is around 10. This means that you can hook the output of one of your gates to 10 other gates. If you try to hook it to more than 10, the circuit will not function, or worse, function unpredictably.
-Noise-Margin measures how much external electrical noise a gate can withstand before producing an incorrect output. TTL will take anything below about 1 volt as a 0, and anything above about 2 volts as a high.
With this in mind, let's look at a few logic families:
TTL -Transistor-Transistor Logic, uses BJTs. easy to construct, high fan-out. Most boards try to be electrically compatible with TTL.
CMOS - (Complimentary Metal Oxide Semiconductor) - Much lower power, majority of modern integrated circuit chips, constructed from FETs.
MOS (NMOS, PMOS) - Higher density than CMOS, uses all one type of FET. I believe Intel is currently using an NMOS process.
ECL - Emitter Coupled Logic- Uses BJT with emitters connected. Very fast, but very hot!
BiCMOS - Modern hybrid, uses bipolar devices in critical sections, FETs elsewhere.
Even within a family, there is much subclassing. That's what a lot
of the letters on the chips in your lab kit are for. The ``74'' indicates
commercial grade TTL (``54'' is military grade - can withstand higher
and lower temperatures). ``40'' indicates CMOS. The last 2 numbers indicate
the logic function ``7404'' means inverter. If there are any letters
in between, it means its some kind of variation:
74H04 - High-speed
74L045- Low-Power
74S04 - Uses a Schottky Diode for improved performance
74ALS04 - Advanced low power Schottky,
etc., etc.
Among the many ways chips are classified is by their level of integration. ``Integrated Circuits'' (ICs) are merely chips which contain more than 1 transistor integrated on a single piece of silicon. They are classified into:
SSI (Small Scale Integration) <10 Gates per chip, most of your lab parts
MSI (Medium) 10-100 gates per chip, adders, subtractors, decoders, multiplexers
LSI (Large) 100-5,000 gates; Microcontrollers, small memories, the kind of thing that runs your microwave or calculator.
VLSI (Very Large - Engineers use really creative names) 5,000 - millions of gates; modern processors and memories.
ULSI (Ultra large)- a made up term which sounds good in promotional literature for the people who build them.
Everything we've done so far utilizes so-called ``positive logic'', which simply means 0V means 0, 5V means 1 (or whatever the ``high'' voltage is - nowadays it's frequently 3.3V or 2.1V instead of 5). In ``negative logic'', the high and the low value are interchanged. The low value means 1, and the high value means zero. Negative logic is typically indicated by the presence of a triangle on all the inputs and outputs to gates in a schematic. While everything functions logically the same, the tricky part is when you are asked to interface between positive and negative logic systems. For instance, take a gate which functions as follows:
| x | y | F |
| L | L | L |
| L | H | H |
| H | L | H |
| H | H | H |
| x | y | F |
| 0 | 0 | 0 |
| 0 | 1 | 1 |
| 1 | 0 | 1 |
| 1 | 1 | 1 |
| x | y | F |
| 1 | 1 | 1 |
| 1 | 0 | 0 |
| 0 | 1 | 0 |
| 0 | 0 | 0 |