AY-3-8912A keyboard problem. Maybe pull-ups needed?

[Volo]

Hello!

I only recently bought my CPC 464 (large motherboard from 1984, two keyboard ribbons) and it seems to have weird issues not encountered with other computers in my collection. I kindly ask for help.

My CPC was bought as keyboard-only. I had to conjure up connectors myself. Power adapter is 5.5V - a bit too high, but no ripple. Worked for a couple of hours, then keyboard died. It traced the schematics – seems that Pin7 of AY-3-8912 got fried.



I couldn't find proper General Instruments AY-3-8912, and replaced it with later revision Microchip AY-3-8912A/P. Datasheets are a bit murky, but the chips should be quite similar, although the newer "A" revision is rumored to have tighter IO timing for better compatibility.

Soldered in Microchip AY-3-8912A/P. It worked fine for like 30 minutes. Then weird thing started happening: Whenever I press a button t first registers a button that I pressed, and then immediately the button to the right in the matrix (O results in OU, U results in UR, R results in RE). When the button is held, the "offset" symbol is on auto-repeat (REEEEEEEEEEEEEEEE).



I believe it is some sort of timing issue. Help me please.

1. Is such an issue common for CPC 464?
2. Is it the issue with AY-3-8912A?
3. It seems that unlike the older AY chip, "A" revision is rumored to require pull-ups on its IO lines. There seems to be a pad for pull-up resistor assembly (NR001 pictured). What is the recommended value of the pull-ups?



Thanks for your input!

[eto]

The A chips require the pull-up resistor array. 10K will do fine. 

[TotO]

@Volo Yes, put a 2K2 resistor network here to fix the issue.

[Bryce]

It's highly likely that the 5.5V PSU is the reason the AY got fried. Get a proper PSU before doing anything further.

Bryce.

[TotO]

Yes, I have not properly read the post, but the request.
You have another problem to damage the AY two times.

[Volo]

Thank you! 10k resistor assemby works just fine. Bought also 4.4k, but decided not to use it as it is yellow and ugly.

P.S. Thank you for your consern regarding voltages. Sadly, all Chinese 5V PSUs that I bought actually give at least 5.4V, only more noisy - there is static on the screen. My buying options are limited - there is a war in my country. The thing works from 5.4 (fingers crossed).

[GUNHED]

Fun thing, all the chinese psu's I bought give about 4,7 V (in use).
Can you take a Y cable and measure the Volts while the system is running? The result could be of interest. 

[Bryce]

Fun thing, all the chinese psu's I bought give about 4,7 V (in use).
Can you take a Y cable and measure the Volts while the system is running? The result could be of interest. 

You don't need a Y cable to do that, just measure across the power in socket.

And yes, fully agree. Most chinesium PSU's overstate their current capacity, so the voltage drops as soon as you get close to their fake claim.

Bryce.

[GUNHED]

Slip away... Shortcut! 

[Bryce]

Slip away... Shortcut! 

You would really have to try hard to short the two pins that are on different sides of the connector and 8mm apart with a DMM probe!

Bryce. 

[GUNHED]

Haha, you don't know my old ditter hands. 

In the wild west the know me as 'Old Ditterhand' 

Happy weekend! 

[Volo]

I measured between Pin 27 and Pin 49 of the expansion slot. It is 5.01 V. Though this PSU makes video noisy.