What is the difference between CMOS and Pass-Transistor Logic?

Question

My friend and I are taking our first digital systems designs course this semester. Our professor has introduced to us two different type of circuit designs; CMOS and Pass-Transistor Logic. The professor isn't very accessible, and we've searched the internet trying to understand the difference between CMOS and Pass-transistor logic, but a lot of the explanations were beyond us. We would really appreciate if someone would explain to us in layman's terms what the difference between the two is, the advantages of each, etc.

Thank you!

Answer 1

I really hate it when these "professors" confuse their students by mentioning terms like "Pass transistor logic" without first properly explaining what CMOS logic is.

I have been working with CMOS circuits and CMOS logic circuits since the 1980s when I was a teenager. Yet I had never heard of "Pass transistor logic" before. That gives you an indication how relevant the term "Pass transistor logic" is.

OK, so to educate myself I read the Wikipedia article on Pass transistor logic. Hmm, so it uses the same type of logic gates (same circuits) as standard CMOS logic but the connections between parts of the circuit is made/broken depending on what the circuit should do. This is of limited use but can be convenient for cells that need to memorize a certain value (zero or one).

But before you can understand Pass transistor logic you must first understand "normal" CMOS logic, so go read the Wikipedia article. Only when that 100% is clear to you should you read about Pass transistor logic.

CMOS logic is what is used in nearly all electronics these days. It will be a challenge to find a modern device with a battery, power lead or solar cell that does not contain CMOS logic circuits.

CMOS logic can be used to create a circuit that can memorize a value. In logic a value is either one or zero (a bit). A CMOS flip-flop can memorize a bit.

In order to save a couple of transistors we can arrange them in such a way that we would need less transistors (to make that flip-flop) yet still be able to memorize that bit. That is what Pass transistor logic is about.

But since modern CMOS transistors can be made so extremely small and cheap it isn't really needed to save those few transistors. Only there where a large number of bits needs to be stored would it make sense to use Pass transistor logic. I would not expect the find Pass transistor logic anywhere else but in memory cells.

So it might have helped if your professor had mentioned that using CMOS transistors we can make "Pass transistor logic cells" and that those are mainly used for memory applications.