• Please review our updated Terms and Rules here

Kaypro 16/2 "check keyboard" error

Thunderbird 98

New Member
Joined
Apr 20, 2021
Messages
3
So I won a Kaypro 16/2 at a local auction. It seems to boot to DOS, but always gives a "check keyboard" error. Someone told me it was bad capacitors in the keyboard logic. The capacitors used in the keybaord logic are marked "641" and I assume they are 640pf, which is an odd size that I cannot find anywhere. I'm hoping that's not the problem. Any other ideas? I really want to get this working!
Thanks
 
Welcome to these forums.

I see you have a related thread at [here], which I will also quote from.

Thunderbird 98 said:
At first I thought it was the keyboard itself, but now I know that there is something wrong with the keyboard logic on the processor board.
How do you know that the problem is the 'processor' board, and not the keyboard? To determine that (using no test equipment), you would have had to have used a 'known working' keyboard.

Thunderbird 98 said:
I found a schematic of the 16/2 which I've attached below.
Based on that schematic, the keyboard interface logic on the 'processor' board is part of the 8255 chip, plus what you see at [here], plus a decoupling cap for each chip. Any of those components could be faulty (assuming the 'processor' board contains the fault). A bad solder joint on the connector is also a possibility.

Thunderbird 98 said:
Someone told me it was bad capacitors in the keyboard logic.
Presumably they included the word "might" or "may". If not, how do they know the problem cause?

Thunderbird 98 said:
The capacitors used in the keyboard logic ...
* According to the partial circuit diagram that I pointed to earlier, C6, C7, and C8, are 47 pF. Likely to be of ceramic type.
* According to the partial circuit diagram that I pointed to earlier, C9 is 10 uF. Likely to be of tantalum or aluminium electrolytic type.
* A decoupling cap for each chip.

Thunderbird 98 said:
... are marked "641" and I assume they are 640pf, ...
See how in the photo, each is located at the end of a chip. C31, for example, is the decoupling cap for the pictured 74LS175.

They appear to be type ceramic of value 640 pf.

Thunderbird 98 said:
... which is an odd size that I cannot find anywhere.
I cannot see how an open-circuit failure of a decoupling cap could cause an always-appearing "check keyboard" error. In my opinion, at worst, maybe the occasional "check keyboard" error.

Thunderbird 98 said:
Any other ideas? I really want to get this working!
What test equipment do you have?
 
With testing equipment, all I have is a multimeter.
I assumed that the keyboard was fine, but it could be broken...
 
With testing equipment, all I have is a multimeter.
I assumed that the keyboard was fine, but it could be broken...
I am unfamiliar with the Kaypro 16/2. I assume that "check keyboard" is a result of the computer's power-on self test (POST) sending a signal to the keyboard, and not seeing the expected signal back from the keyboard. If that is the case, then I see the possibilities as:
* Faulty keyboard
* Faulty processor board
* Faulty cable that runs between J1 on the processor board and the DIN connector at the rear of the computer.
* Aforementioned cable not connected properly to the J1 connector on the processor board. Re-seat the plug on J1, and ensure it is plugged on correctly.

Do a thorough visual inspection.

It would be good if you can find someone local that has a working Kaypro 16/2, and trying their (known working) keyboard. That would quickly rule out/in half of the keyboard related circuitry. Maybe someone here knows of a different keyboard that is known to work with the Kaypro 16/2.

With a multimeter, you have the ability to see logic high and logic low levels within the processor board's keyboard interface circuitry. But a question becomes, "What are the normal levels expected to be seen on the various pins?" That is not always straightforward. For example, I found that in the IBM XT, certain failure modes of the XT's keyboard interface circuitry can result in the POST putting the 8255 chip's pins into unexpected states (including a case where the 8255's PB6 pin was constantly alternating).

You could gain some confidence in the processor board's keyboard interface circuitry, by finding someone else online who has a working Kaypro 16/2, and doing logic high/low comparisons between the two computers. But that could be wasted effort because the keyboard is what is at fault.
 
Apparently the 16/2 isn't very common, but I know someone who is knowledgeable on IBM PC stuff.
Thanks for the help!
 
Back
Top