Nooo.. formerly healthy PET 2001 is sick

[bitfixer]

A formerly healthy PET 2001-N suddenly became very sick..
Just turned it on today after a few weeks of inactivity. It came up to a mostly blank screen with a few characters (mostly the greek letter pi) along the top few rows. I removed all the ROMs and reseated them, and then got the line "*** COMMODORE BASIC ***" on the top but no cursor. After a few seconds a machine language monitor popped up with some addresses. It was in this state for a while, and thinking it might have something to do with bad RAM, I piggybacked some 4116s on one of the rows of memory. Next time I turned it on, I got a garbage screen, and now I am consistently getting that garbage screen. No real pattern I can see except the top left character seems to always be '@'.
Seems like it's possible there is a bad socket. Is there a particular bad socket that would have such symptoms?
Frustrating.. just when you think the PET is happy..

 

[Eudimorphodon]

For the "stuck on the garbage screen" symptom, have you checked out the RESET circuitry? That just randomly went bad on my 4032, and turned out to be a rotten capacitor.

 

[bitfixer]

Thanks! I just checked out RESET on the 6502, and indeed it appears to be stuck low.. (restrained excitement) So now I have something to dig further into.

 

[bitfixer]

Eud, dead on with the RESET circuitry. This was pretty much the same problem I was having here. One or more of the capacitors is bad, it was keeping the 6502 in reset. In addition to that, it looks like one of the ROMs has died. With a spare set, the PET is ok again. I'll have to sort out exactly which ROM (or ROMs) failed. Thanks again for the insight.

 

[Eudimorphodon]

Yay!

I was going to ask which capacitor of the three associated with the 555 went, if you knew. Money was on it being the tantalum one but it turned out to be one of the ceramic disks. (I don't know which because I replaced them in the same go.)

When it was crashing to the debugger do you happen to remember what the PC was? (Or if it was consistent?) I had a broken data line on the ROM bank between the D000 and C000 ROMs and the PC reading I'd get at the crash (CD76) actually helped clue me in to where the issue was. (Coincidentally the ML debugger lives in the D000 ROM. If the data line had been bad one socket over I wouldn't of gotten that far.)

http://www.vintage-computer.com/vcf...s.-(2001-8-4032-and-8032)&p=183909#post183909

 

[dave_m]

gub,
Good work. As Eud said, if the machine is working well enough to get to the ML Monitor with a blinking cursor, one should read samples of the various ROMs, read/write RAM with some test patterns, etc to get as much info as possible before power cycling the machine.