Partially working 800 keyboard - non-mylar?

[Arcanis-Will]

I have an Atari 800 with a keyboard that has 8 or 9 non-working keys. I have read on several threads about having to repair the mylar layer so I took everything apart and attempted the separation of PCB from key frame, but the PCB did NOT want to come off. I applied enough force that I feared cracking the edges of the PCB so I gave up. It was like the PCB was glued or soldered in place. While it was apart I cleaned all the card edge contacts and reseated all the boards (I did not reseat the ICs).

 

So then I reassembled everything including the case, added a small shot of De-Oxit into a few non-working keys. That did not help so far.

 

Further reading shows that there were several different makers. I did not know that while I had it open, or I would have written down the name from the board. I don't remember it.

 

Are there other variants that are not mylar based? And if so, what would be best? Just the repeated key activations?

 

I'll describe what I remember:

- The attached ribbon cable was fairly thick and multicolored with red-brown female connector.

- 9 attachment screws, could not find one under the tape area. However, removal of the screws did NOT loosen the PCB in the slightest.

- Heavy grey plastic frame - keyboard unit as a whole was very heavy and solid.

 

My next steps would probably be to spray under each non-working key with deoxit (I only did 2 so far), exercise them for a long(er) time, let it dry and retest. Should I open it back up and get the name of the board?

 

Any other suggestions are welcome.

thank you!

[Rybags]

Most of the keys fall within an 8x8 matrix - having 8 that don't work sounds conveniently like a single severed connection, the keys in each row/column aren't necessarily adjacent.

 

Nice little representation of the matrix is on this page http://members.casema.nl/hhaydn/howel/logic/burched/b5_800xl_kbd.htm

[ClausB]

I've had some 800 keyboards with cracked solder joints. Look for fine circular cracks with a magnifier. Simply resolder them.

[Arcanis-Will]

Most of the keys fall within an 8x8 matrix - having 8 that don't work sounds conveniently like a single severed connection, the keys in each row/column aren't necessarily adjacent.

 

Nice little representation of the matrix is on this page http://members.casema.nl/hhaydn/howel/logic/burched/b5_800xl_kbd.htm

Nice chart. Can I assume that the scan matrix of the 800XL is the same for the 800?

 

Failing keys are (from left to right, row by row)

3 9 0 e o p d L ; + * c b

 

Some fall into rows or columns but not all - as in, some but not all of a column are failing. Could be a break partway through.

 

If the other data is correct then I could find the 4051 multiplexers and reseat them. I don't think I need to reseat the POKEY since at least one key works out of each row and column.

[Arcanis-Will]

I've had some 800 keyboards with cracked solder joints. Look for fine circular cracks with a magnifier. Simply resolder them.

 

Thanks, I will do that.

[Roydea6]

http://atariage.com/forums/topic/190251-keyboard-connector-documentation/?p=2408846

 

 

Check this page for info on which trace effects which Keys..

 

[ACML]

I'll describe what I remember:

 

- The attached ribbon cable was fairly thick and multicolored with red-brown female connector.

 

From this description It sounds to me like you have a Hi Tek keyboard. These are typically very reliable save the notorious spacebar failure. The pattern of keys that are not working does not sound like an easy fix that I can think of. Your best bet may be to buy a replacement keyboard. Stay away from Mitsumi. If its a Hi Tek (whiter plungers), verify that the spacebar is working specifically along with the rest of them. The Stackpole (yellow plungers) is also a good reliable mechanically switched keyboard, but can suffer from cracks in the aging plungers that can cause the keys to occasionally pop off or bind. There are some NOS Stackpoles out there. I'd image that any Hi Teks will have an 800 wrapped around them and just avoid the Mitsumi type altogether.

Edited February 15, 2014 by ACML

[Arcanis-Will]

I've had some 800 keyboards with cracked solder joints...

 

I re-opened my 800 tonight and yes indeed there are breaks in the continuity across the keyboard traces. There are at least 2 spots where there is a break in the traces, and several other places where the pad or solder is popped up. Using a multimeter I can trace the continuity back up through the matrix starting from where the ribbon cable connects, and also from key to key, and it is showing quite a few dead spots to fix. (It took me about 45 minutes to remove most of that double-sided foam tape that was under the ribbon cable.)

 

So good news - I tested a few of the non-working keys and they are actually making good contact when measured at the key itself. Just not getting across the board. Tomorrow's project will be the resoldering and jumpering.

 

Thanks for the suggestion!

[ClausB]

Glad to help. I forgot about the sticky foam. Yes, that's the worst part.

[Arcanis-Will]

 

Tomorrow's project will be the resoldering and jumpering.

 

Well I said "tomorrow's project" but I had other stuff come up so it took an extra day. Tonight I spent about 2 hours fixing a surprisingly large number of broken pads. Looks like there was some kind of impact to the keyboard that lifted a lot of the keys' pads.

 

Long and short of it -- all keys work 100% now!

 

I took pictures of the finished product. Should I post them? (need I ask?)

[ClausB]

Seems to be common on older 800s. I had three like that. Did the owners pound on the keys? Or maybe fatigue from thermal cycling? Or just crummy solder?