A1200 Keyboard Diagnosis

[Elmo Putney]

My A1200 has in the past lost a few keys when I haven't used it for a while (like days), they eventually come back after a few hours. Just wondering if I can diagnose where the issue is (membrane, connector, controller etc...)
I've not used it for a Month or so, and currently it is refusing to co-operate, it really is a twat sometimes.

I have cleaned and re-seated the ribbon cable, and it looks good - so I don't think that is the issue.

I have an A500 in the cave - can I test that keyboard on an A1200 to start to isolate the issue? i.e. is it "plug compatible"?

Diagrom: no issues (other than the keys not registering)

It is normally the same keys that drop out, I looked at the schematics to see if it gave any clues as to maybe a specific trace is flaky but they are not very helpful.

Keys that are out:

F10
Del
Keypad: 0 (zero)
Enter (not the keypad one)
Backspace
and |\ (next to backspace)

Does that give any useful clues? (probably not)

Thanks all

Also have an A600 I could swap the keyboard from.

[Elmo Putney]

Obviously having just written this the little twat has decided to re-enable the lost keys, it's been switched on for about 3 to 4 hours running diagrom and with me poking/cleaning etc......none of which made any difference, until leaving it alone to sulk for a bit.

Question still stands though.

And yes it's recapped.

[Elmo Putney]

The one thing I think I can conclude is that it's not the membrane if all the missing keys came back at the same time? true?

[DanBy]

As I've built a few keyboards in my time and just finished a new A1200 mechanical keyboard... I can tell you that the keyboard circuit/membrane for the A1200 is incredibly simple and not likely to play a role in an issue that is time/waiting based.

Judging from the set of keys that don't work for you I think one or two of the circuits on the keyboard membrane does not have full continuity back to the keyboard controller chip (IC). Somewhat generally any issue that resolves after time might be the result of the operating temperature for some part of the circuit raising and resulting changes causing some broken trace/circuit to reconnect and regain continuity. One thing that might confirm this is to flex the motherboard gently in a direction while you're having the issue and see if the minor flexing restoes functionality. Or maybe pressing down on the keyboard controller chip might fix the issue. Either of those would lend some credence to this being a flaky trace or dodgy solder joint (or two). And worth trying as they are easy/crude tests that might give an indication of what is up.

That said the very, very first thing worth doing is "cleaning" the bottom of the key plungers that make contact with the membrane just to ensure they are making good contacts. These are graphite pads so you can refresh them by rubbing with a pencil. I don't think this will solve your issue but worth doing just to rule out something easy and basic and if it works you avoid a lot of time consuming diagnostics.

Next, I would look for dry solder joints anywhere on the path from the ribbon connector to the keyboard controller. The keyboard controller is on the motherboard at position U13, iirc. You might be able to spot a solder blob at one of the ribbon connector pins (underside of the motherboard) or at keyboard controller IC that looks dull/old/sad and reflowing it will fix this issue. You might even see a trace that looks kind flaky/poor. Though you might need magnification to catch that kind of thing.

If you really can't see anything then you'll want to continuity test each pin on the ribbon connector back to the relevant pin on keyboard controller IC and see if one of the lines isn't fully connected. You'll have to do this from a cold start when you know the keyboard is playing up. And if possible you'll want to try and test things from the contact inside the ribbon connector, not just the pin on the underside of the motherboard (that might be hella fiddly though).

To check where traces go try https://www.amigapcb.org/

If you're confident that all the traces, and sodler joints are fine then replacing the keyboard controller might be the next thing to do. There are 2nd hand ones or you can build a new one

https://github.com/OlegMishin/A1200_keyb_MPU

[aeberbach]

Unfortunately A600 and A1200 are not compatible, but you can grab my projects
A1200 Keyboard
A600 Keyboard
to see the keyboard matrix schematic of either. Based on the keys you said there isn't a likely broken trace or traces especially not ones that would all go in or out all together. You should probably look around your keyboard decoder chip for signs of anything, Or yep, try replacing the MPU with Oleg's as DanBy linked.

[Elmo Putney]

Thanks both, you have already helped me understand the problem.



Your very useful picture confirms that line "28" is the issue, is it safe to assume your numbering "28" corresponds to the pcb explorer numbering - i.e. pin 28 of the keyboard connector as per screen grab?

[DanBy]

Looking on aeberbach's blog

https://aeberbach.github.io/posts/20...3-a1200-kbd-2/

he states

I decided the ribbon connector that goes to the A1200 mainboard would be numbered arbitrarily from left to right

That said it is true the pins are labelled 1 to 31 left to right according to the schematics. You have 2 pins connected for every key, 28 is the common one here but there is also 1,2,3,4,5,7. Assuming the numbering is correctly mapped to the amigapcb then those pins are all routed to the same top right regiom of pins on the keyboard controller IC. Maybe the keyboard controller isn't fully soldered down around there? Or some subcircuit inside is blown.

TBH if I was continuity checking the pins I'd just do them all, its fiddly but once you're started you might as well do the lot.

[Elmo Putney]

Motherboard checks out.

Testing the end of the ribbon connector always see open circuit with any key combo that uses 28. Working keys I see about 200 Ohms, so looks like a broken trace on the membrane, I could probably bodge it but I'll order a new membrane.

Thanks for the help, really useful.

[Elmo Putney]

found it

[DanBy]

Nice!

Yeah they are needlessly fragile. I have a keyboard mod that passes a couple of thin wires over/alongside the membrane and I briefly crushed that wire in to membrane while replacing the top of the case back on and it manage to brake a trace