Keys c,v,n,m don't work when pressed once

I've recently been working on modding my A1200, and only just realised that 4 of the keys aren't working as expected.

The first mod I've made has been to replace the capacitors (with branded Panasonic caps), and as far as I can make out, everything is working fine.

However, only recently, as I've been using the Amiga more, I've noticed that certain keys only work under certain conditions:

- keys affected c,v,n,m
- they work if held down and another key pressed (i.e. if I hold c and press v, then "c" appears on the screen")
- they work if held down (i.e. when the key repeat starts, they work)

This makes me feel that it isn't a connector issue. However, I have cleaned the ribbon cable and connector, and also gave the whole motherboard another clean with alcohol, and the issue still remains.

My thoughts now are somehow the membrane has something affecting it... although I'm not confident to open the keyboard, or if it's the issue. The named keys do work consistently, even if they don't work when pressed at first.

I do have a feeling that the keys may not have worked previously (before I replaced the caps), although I can't be certain. Prior to that, I hadn't really used the Amiga for close to 20 years (I guess).

So I have a couple of questions if I may:

1. Is it worth taking the keyboard apart and cleaning it?
2. If I do so, am I likely to damage it?
3. Based on the symptoms, is it even likely to be the culprit?

Many thanks! I'm really hoping I can fix this...

No response, oh well.

Well, just to update, I tried multiple things (including buying a replacement keyboard membrane), and also found out that my power supply was not delivering correct voltages. Didn't look like the voltages were high, but a little concerned that it might have surged in the past. Replaced it for now with my old Amiga 500 power supply, which despite looking bad, actually still works fine.

Someone has mentioned that the problem may be the CIA chips... does this sound likely?

[Leffmann]

The A1200 keyboard is like that, it's probably the cheapest part of everything ever put into these computers. I think it has to do with the connection matrix inside the keyboard, where most of the keys on each row are connected on the same line and in effect become mutually exclusive.

http://eab.abime.net/showthread.php?t=71099 etc.

Thanks for your reply. However, it's not the keyboard. As mentioned (albeit not explicitly), I replaced the membrane, and exactly the same issue occurs with both membranes. The keyboard connector is also okay - I've checked that several times.

[ajk]

- they work if held down (i.e. when the key repeat starts, they work)

This is what got me stumped. Obviously the key press is being detected so there was probably nothing wrong with the membrane in the first place. As far as I am aware the key repetition is purely a software feature. Yet having all the affected keys on the same row sounds like it would be some hardware related fault...

Have you been doing all your testing in Workbench? Is there any difference if you boot a game where you can type in a name or something, do those keys behave differently?

[nobody]

The membrane of the keyboard is damaged in some places. You will need a lot of work if you want to repair that and a set of tools like small screwdrivers, multimeter and conductive glue. Get ready to take out 40 or so really small screws to release the membrane. Probably better to buy a replacement keyboard.

[ajk]

@nobody

How have you come to that conclusion if he has already swapped the membrane once, and even with the original one the keys worked if you pressed them for long enough?

[Toni Wilen]

Originally Posted by ClockworkCoder

Someone has mentioned that the problem may be the CIA chips... does this sound likely?

It is almost impossible. CIA only works as a serial port for keyboard 8-bit serial scan code data, it has no knowledge of key codes or columns or rows. Software does the rest, including key repeat.

[Jope]

Perhaps the keyboard MCU's IO lines are damaged?

Originally Posted by ajk

This is what got me stumped. Obviously the key press is being detected so there was probably nothing wrong with the membrane in the first place. As far as I am aware the key repetition is purely a software feature. Yet having all the affected keys on the same row sounds like it would be some hardware related fault...

Have you been doing all your testing in Workbench? Is there any difference if you boot a game where you can type in a name or something, do those keys behave differently?

It is odd. I've done most testing in Workbench, although someone else had suggested that I press both mouse buttons on boot to test from the command prompt. There was the same issue there too.

Another issue - not sure if this is new, or if it was happening intermittently before - the left mouse button isn't always working. I can press the left Alt to work around it though. This issue only seems to happen in Workbench, and the odd thing is that I can still hold the left mouse button down to drag.

The left mouse button also always to work without fail for any loading screens that require pressing it too.

The keyboard (c,v,n,m) issue does occur in games though - still working if I hold until it starts repeating.

Originally Posted by Toni Wilen

It is almost impossible. CIA only works as a serial port for keyboard 8-bit serial scan code data, it has no knowledge of key codes or columns or rows. Software does the rest, including key repeat.

I'm really encouraged if so. I was steeling myself in case I needed to, and was starting to research putting in sockets (PLCC44) to make further tests easier. However, it would be a really challenging fix for me to attempt.

Originally Posted by Jope

Perhaps the keyboard MCU's IO lines are damaged?

Static or over voltage damage? Are you able to point me towards some way of checking that?

[Jope]

You could take an oscilloscope to each line in turn and check to see what levels they output.

Or perhaps just blindly replace the keyboard MCU. :-)

Originally Posted by Jope

You could take an oscilloscope to each line in turn and check to see what levels they output.

Or perhaps just blindly replace the keyboard MCU. :-)

Thanks for the suggestions.

The good news is I have an oscilloscope. The bad new is I don't really know how to use it, and to be honest, I'm a little cautious about using it with something that's mains earthed. I do really want to be able to use it for such occasions though...

I guess a replacement MCU may not be easy to find either...

[hooverphonique]

Originally Posted by ClockworkCoder

Thanks for the suggestions.

The good news is I have an oscilloscope. The bad new is I don't really know how to use it, and to be honest, I'm a little cautious about using it with something that's mains earthed. I do really want to be able to use it for such occasions though...

Only connect oscilloscope probe ground to amiga ground, and you should be fine. Otherwise, you'll need to unearth one of the units, or get an isolation transformer

Originally Posted by hooverphonique

Only connect oscilloscope probe ground to amiga ground, and you should be fine. Otherwise, you'll need to unearth one of the units, or get an isolation transformer

Thanks. After I read your comment, I re-watched EEVBlog's "How not to blow up your oscilloscope" and feel a little more confident.

I'll give it a go... although now I'll attempt it after I've completed the Gotek mods - I now have the OLED and rotary encoder At least want to see them working before I break anything else!

Small update. I haven't really changed anything, but it seems that "c" and "v" at least are now working. "n" and "m" are the same as before.

Not sure if I mentioned, but there's also an issue with the mouse, at least in Workbench, which I feel may also be related to the keyboard, basically:

- When I press the left mouse button it won't register a click on icons or anything, however,
- It moves the Workbench backdrop if I hold the LMB and move it, and
- If I press the "Left Alt" on its own, a LMB click is registered.

This makes me wonder if somehow one of the Amiga keys is permanently registering somehow (shorted?), as usually I think you should need to press Amiga + Left Alt for a Left Mouse Click - is this correct?

Incidentally, I got a new power supply from https://www.c64psu.com/, and, yes, I did tentatively check the output voltages before plugging it in. It seems to work really well

[hooverphonique]

Originally Posted by ClockworkCoder

Thanks. After I read your comment, I re-watched EEVBlog's "How not to blow up your oscilloscope" and feel a little more confident.

Yes, that's a very good watch when in doubt about these things