Repairing a beautiful 4032, spooky symptoms, any ideas?

[fantomas]

Hi guys,

I'm trying to repair a beautiful 32k 4032 and have a really really strange problem. I have a "spare 4032" which I used to check some parts (by exchanging them with the working 4032), so I can say that

* All ROMs are ok
* Both 6520 are ok
* 6522 is ok
* 6502 is ok
* CPU CLK is ok (checked with an scope)

Symptoms:

* After power up screen is garbled and turns black after a couple of seconds
* I changed 4x 4116 RAMs until now (sockets J2, I2, J3, I3)
* Before changing the 4x 4116 the screen sometimes showed "counting characters" or single, static characters instead of the black screen

But now comes the spooky part:

When I see the black screen and I switch the PET off, I sometimes see the start screen characters (### Commodore Basic ....###) for a glimpse of a second on the screen before it vanishes.

What could cause this kind of symptoms? Does switching off the machine "release" a kind of IRQ? Strange ...

Any ideas?
Kind regards,
Ingo

 

[dave_m]

* I changed 4x 4116 RAMs until now (sockets J2, I2, J3, I3)
* Before changing the 4x 4116 the screen sometimes showed "counting characters" or single, static characters instead of the black screen

So you have a 9" CRT 4000 PET that does not have a 6545 CRT Controller as the RAM are in the J and I positions?

You must start with the 8 lower 16K X 1 bit RAM chips as bad zero page will not allow the PET to boot. They are I2 thru I9. Forget about the upper RAM for now (J2 - J9). Piggy back a good 4116 on the I chips one at a time and power on computer. If something different happens, that may be the bad chip. To avoid so many power cycles, you may want to hook up a switch to the RESET line. It can be found on the expansion connector at J4-22. A ground pin will be on the adjacent row.

 

[KC9UDX]

Try powering it up without the 6520s. It does sound like a stuck IRQ or reset. Of course, you can check both of those signals with your oscilloscope first, too.

 

[Eudimorphodon]

I'd definitely check out the "stuck reset" idea. The basket case 4032 I was repairing some years ago had that issue; solved it by replacing the capacitors in the 555 timer section.

 

[fantomas]

Thanks for your answers so far, I was on a business trip last week so could not continue working on the PET.

I tested booting without the 6520s, also changed them with my spare 4032, it did not change the boot behaviour.

 

[fantomas]

Yes, the RAMs in both of my PETs are in the J and I positions. Thanks for the hint about the replacement order for the RAMs, I will continue changing the I-row.

Another possible hint:

Directly after booting into the black screen I have no signal on CAS0 which I traced back to the 74LS244 in B3. Does this make any sense? Should I try to replace it or is this the symptom and not the cause?

Thanks for all your help,
Ingo

 

[fantomas]

Good news!

I replaced the RAMs in the I-row as dave_m suggested. After replacing the 4116 chip in I4 the PET came to life

Thanks for all your help, it feels so great to see it up and running again.
Ingo