Digital net and De Morgan curiosity

[flanguasco]

While implementing a SN74196 (Presettable Decade Counter) as a hierarchical block I have found a strange behavior that I can't understand.

Looking at the attached circuit you see that the IC inputs, LOAD/COUNT and CLEAR, are connected to an OR gate with two inverted inputs.

I have firstly used an OR gate and two NOT at the inputs but it didn't work: a negative pulse on the CLEAR input correctly set all the Q outputs to zero on the falling edge but the following rising edge set all the Q outputs to the values currently applied to the Data inputs (wrong).

After a few trials I have replaced the OR gate and the two NOTs with the De Morgan equivalent, just a NAND gate, and everything was then working fine.

Now my question is: why this? Where is the difference in the two versions of this circuit?

Is it related to the sequence used by Multisim 13 in evaluating the net events?

Ciao and thanks for the attention.

Franco

[FatBoyMonchi]

Hi, I think the OR and double NOT is not equivalent to a NAND, so the behaviour wasn't correct before.

[flanguasco]

Hi FatBoyMonchi:

sorry for the delayed reply but only now I saw your post.

If you look in the left-bottom corner of the attached circuit you see the LOAD/COUNT (I1) and CLEAR (I2) connected to an OR gate trough two inverted inputs so, I think, the OR output (U) is:

U = Not(I1) OR Not(I2)

But, for the De Morgan equivalence,

U = Not(I1) OR Not(I2) = Not(I1 AND I2) = I1 NAND I2.

If my previous expressions are correct, I still don't understand the different Multisim behavior.

Ciao, thanks for the reply and to the next.

Franco