Différence entre le SN7490 et le SN7490A

[Daudet78]

Bonjour
Je suis confronté, sur un ancien équipement, au remplacement d'un SN 7490 par un SN7490A (le SN7490 n'est plus trouvable !). Ce circuit est utilisé dans un diviseur de la fréquence secteur (50 hertz écrêtè par une résistance et une Zener 5V). Avec un SN7490A, le comptage par dix fonctionne mal et j'ai une réponse folklorique ( division par un nombre plus petit que 10 !)
Si quelqu'un a une idée du problème ....
Merci d'avance
Daudet

 

[davenn]

Hi Daudet

welcome to the forums

please keep to English, then we can understand your questions and comments

most likely just a variation in the type of case material

Dave

 

[(*steve*)]

With the 4000 series, there were A and B variants, the A variant being unbuffered and the B variant being buffered.

I can't find any reference which suggests a meaning for the A at the end of 74xx logic and I note that some datasheets refer to (say) DM5490/DM7490A which seems to indicate that the 54 series (mil?) and the 74xxA are identical.

I always thought that the A was a package designator (AN is typically seen on plastic DIP versions) but it's only the "N" that is significant there.

 

[davenn]

agreed, Steve

any link that stated 7490 ( no A) when clicked on always took me to a datasheet that had the A suffix

 

[Daudet78]

I'm back
I think I found the problem. In the schematic in question, I have the 7490 clock signal which is made with 6.3V AC clamped with a resistance of 2.2K and 5V Zener. And I think that is not compatible with a TTL signal. The 7490 had to be a more slower circuit (and therefore more tolerant!) divider by 10. With the 7490A (F = 32MHz), it need a clock with steep edges. My equipment working again with a 7490A if I put a schmitt trigger (74LS14) between the clipped signal and pin 14 (clock A)
It's not a nice hardware modification but effective ...
Thank you for yours replies
Daudet

 

[(*steve*)]

Interesting.

What was the behaviour you were seeing with the 7490A before you added the Schmitt trigger?

Most TTL documentation is scanned and hard to search on the internet. But my curiosity is piqued now!

 

[Daudet78]

With a simple replacement of 7490 by a 7490A, I count a random (and not by 10!) Of the circuit. I have not found the datasheet of 7490. But recollection, he was 10 MHz max.
My explanation is that during the slow transition of the input 7490A, there is an oscillation of the input stage and therefore a multiple counting
With a Schmitt trigger, the problem is fixed

 

[Daudet78]

I found an explanation http://serge.dusausay.free.fr/ss/complement/art29/art29.html
Excuse me, but it is in French. But the last image clearly explains the oscillation phenomenon with a slow Dv/Dt in input

 

[(*steve*)]

Many circuits require a fast edge to their clock lines to prevent slight differences in behaviour.

I went looking for mains derived clock circuits for TTL and found this. I note the capacitor to remove high frequency components from the signal - but which would also slow the edges. That aside, it seems fundamentally the same as the circuit you describe.

This circuit provides a faster edge for CMOS, and it should also work for TTL. I'm not sure it would be more convenient than the extra chip. You would probably want to reduce the resistor values by an order of magnitude for TTL though.

The Schmitt trigger is still the best solution though as it will give extremely fast edges and noise immunity.

 

[Daudet78]

This is exactly my case ...... so be a trigger schmitt between 50Hz and the input of 7490A.
But on a link, a CD4017 is used . This product as an internal schmitt trigger (unless it is sold by Texas Instruments ...)