1
\$\begingroup\$

The problem is a homework problem from the EDX Digital Logic class. Please avoid looking at the rest of the post if you don't want to know the solution. I've already attempted the problem and know the solution because I got it wrong. I'm trying to understand the solution.

Consider the following circuit: enter image description here

Based on the problem, the propagation delay is 42. Can we simplify the circuit to have a lower propagation delay?

I recognize that the "bubbles" can be canceled between the inverters and the NAND2 gates such that Y = ABCD (an AND4 gate)

The solution to the problem says you should be able to rewrite circuit with a propagation delay of 18. Can someone please explain how?

I've tried to convert the AND gates into OR gates with inverted inputs, but that doesn't work. Assuming the answer is correct, the only combination(s) that can get a propagation delay of 18 are:

1. 3 NOT gates which can't convert 4 inputs into 1 output
2. 1 NAND3 and 1 NAND2 (which leads to !(!(ABC)D) which isn't ABCD
3. 1 NOR2 and 1 NAND2 which can't convert 4 inputs into 1 output 

and so forth

\$\endgroup\$
3
  • 1
    \$\begingroup\$ What is "Contamination delay"? \$\endgroup\$ Commented May 29 at 19:21
  • 1
    \$\begingroup\$ I've seen other people call it the minimum propagation delay if that helps? \$\endgroup\$ Commented May 29 at 19:35
  • \$\begingroup\$ @IbrahimRabbani 18 pretty much tells you exactly what to use. You are going to use two or more gates with a max of 10 among them, followed by exactly one 8. There's only one 8 to pick. So that's a given. And the NOR2 doesn't look like a good candidate for 10, so perhaps start with the NAND3 using 3 inputs and providing one input to a NAND2 and one other gate with 10 or less that can incorporate the remaining input and provide the other input to the NAND2. This can't be that hard. \$\endgroup\$ Commented May 29 at 21:52

1 Answer 1

0
\$\begingroup\$

You have four inputs, and only 2- and 3-input gates. So at least two gates must be used.

Consider the input signals: what other permutations of gates and inputs can be made?

1+1+1+1 (i.e., all NOTs; as you say, not useful)
2+1+1 (as shown)
3+1
2+2
etc.

Do the output terms of any of these combinations have useful values?

Another hint: symmetry is valuable here.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.