• Please review our updated Terms and Rules here

Commodore PET 8032 Monitor problem(s)?

daver2 said:
Vision is a wonderful thing...

Looks like something now in the ODD video RAM circuit is causing a problem (to start with of course).

See the schematic here: http://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/8032/8032029-09.gif.

Check for activity on: UB8 pins 1 and 11 and see what the signals are like on all of the pins around UB8.

It could be one of UC6 or UC7 (the video RAMs themselves). This is the most likely cause.

Dave
Click to expand...
And so says the all mighty Google!

I found this site: https://www.apuliaretrocomputing.it/wordpress/?p=6250

The site shows a PET with a similar issue, only theirs showed exclamation points and had an issue with the keyboard. But the rest of the problems were very similar to mine. Unfortunately, all my 2114s are soldered in. So I wondered what the story was about piggybacking these to diagnose them. I know it works for 4116 DRAM. But what about SRAM?
 
It CAN work. I would guess 30%-40% of the time.

It will only work if the faulty bit(s) are HIGH (and the piggyback RAM can pull them LOW) or the faulty bit(s) are floating.

If you get my PETTESTER working, I can tell you which device is at fault :)...

Dave
 
Interesting, the numbers are OK... As is the text "bytes free"...

There appears to be 'blocks' of 15 or 16 bytes that are not correct - and most of the characters I see are composed of Romulan text that are unobtainable from the character generator. This implies that the faulty character cells should be changing the character in the cell on a fast basis - possibly indicating that the faulty bits are floating...

Dave
 
Last edited:
daver2 said:
Interesting, the numbers are OK... As is the text "bytes free"...

There appears to be 'blocks' of 15 or 16 bytes that are not correct - and most of the characters I see are composed of Romulan text that are unobtainable from the character generator. This implies that the faulty character cells should be changing the character in the cell on a fast basis - possibly indicating that the faulty bits are floating...

Dave
Click to expand...


Yeah. But it's not the characters themselves. I should mention that the garbage on screen is shimmering. Any text that intersects with this garbage comes out wrong. Text in the clear areas of the screen are fine. Delete, however, adds new characters about 20 spaces after the curser. The string of which seems to grow with every press.

The good news is that my keyboard is working perfectly.
 
More good news! There appear to be four 2114s on the board, and guess how many spars I have! Still need sockets though.

Will piggy back tomorrow. Since it's a bit past midnight.
 
Last edited:
Two are used by the EVEN video RAM (that appears to be OK) and two for the ODD video RAM (that appears duff in someway).

Dave
 
I have mentioned before, that it seems, that nearly every PET repair involves defective 2114 SRAMs. I'm not sure why the high mortality rate.

One very odd thing is most of the 2114 SRAMs I have bought as spare parts are ok.

I am a "spare parts lunatic" and probably should be medicated for it, I buy up a lot of IC's & transistors used in my valued equipment and test them before I put them into my spare parts inventory.

Generally too, I go for the ones that are genuine vintage 70's and 80's date codes. But then, I often go for mil spec style parts with ceramic IC packages and that way I don't get too many duff IC's.

But even the common garden resin package ITT 2114's I have bought as spare parts are ok.

But it certainly appears, that it is hard to find a vintage PET motherboard with original 2114 IC's on its pcb that still work.

It makes one wonder what is going wrong inside these IC's.
 
Our company is a member of JEDEC (or used to be I should say now) and a number of suppliers had problems before the JEDEC Committee produced their Standards.

As a result, it could be that one or two manufacturers had problems with their production lines, and Commodore bought into these as parts suppliers for 2114s.

The toss of a coin, and it came down heads rather than tails. Or, more correctly, they got a better price from that supplier than the others...

Dave
 
Almost had it!!! Then the power cut out. I'm assuming it's CR2 (since removd) as that was the one overheating. CR4, which is the same component and right next to it was fine. The 12v rail was high at 17v. The +5v was gone. And the -5v was low at less than -0.1v. Also, I checked the fuse first thing, and it's still good.
 
DistantStar001 said:
Almost had it!!! Then the power cut out. I'm assuming it's CR2 (since removd) as that was the one overheating. CR4, which is the same component and right next to it was fine. The 12v rail was high at 17v. The +5v was gone. And the -5v was low at less than -0.1v. Also, I checked the fuse first thing, and it's still good.
Click to expand...

The usual cause for the power rails crashing is the blue Tant capacitors shorting out. Fortunately, unlike a lot of modern switching regulators (and replacements for analog regulators claiming high efficiency) these early design analog regulators were amazing in that they have great over-current and over-temperature shut down functions that save them from damage if their output are overloaded. In this respect they are impossible to beat and were a masterpiece of analog design. For example look up the design and specs of the original National semiconductor LM309k 5V regulator from the 1970's era, they were a masterpiece of design.

It might not be that, but if your board has these blue Tants they should all be replaced. Luckily some of these boards used yellow jacket electrolytics and generally these are fine.

If you are powering your board from anything other than the original power transformer, trying to inject DC voltages into it from other power supplies, it creates many opportunities for accidents. This is why I never recommend it when people have suggested it on other threads.
 
As far as I can tell, there is only one such capacitor on the board. And it doesn't appear to have anything to do with power (C51). Everything else is electrolytic. Even so, I'll replace it to be safe.
 
DistantStar001 said:
As far as I can tell, there is only one such capacitor on the board. And it doesn't appear to have anything to do with power (C51). Everything else is electrolytic. Even so, I'll replace it to be safe.
Click to expand...
I have not found a faulty electrolytic cap on either of my two PET boards yet. Though the main 23,000 uF filter cap in my PET was defective. On one PET board, two of the Blue Tants were shorted out.

The failure mode of electrolytic caps is generally not shorting. They tend to dry out, lose capacity and their ESR goes high. You can check that of course with in circuit testing with an ESR meter.

I have rarely have seen electrolytics short, but they can, one notable case, I disassembled the capacitor to find out why. At the factory a tiny metal fragment got into the insulation between the foils and that is how the short occurred.
 
Hugo Holden said:
I have not found a faulty electrolytic cap on either of my two PET boards yet. Though the main 23,000 uF filter cap in my PET was defective. On one PET board, two of the Blue Tants were shorted out.

The failure mode of electrolytic caps is generally not shorting. They tend to dry out, lose capacity and their ESR goes high. You can check that of course with in circuit testing with an ESR meter.

I have rarely have seen electrolytics short, but they can, one notable case, I disassembled the capacitor to find out why. At the factory a tiny metal fragment got into the insulation between the foils and that is how the short occurred.
Click to expand...
A complete recap is likely going to be my next step. For now, I've replaced both bridge rectifiers and the tant capacitor. -5v and 12v are back! And spot on. +5v... Not so much. I'm only getting about 0.9v.
 
Finally... SUCCESS!!!!!!!!!!!!!!!

And the +5v issue was so stupid! The back screw between the two cassette ports had worn through the solder mask. A simple insulating washer and bingo! A working Commodore PET has re-entered the world!
 
Cool!

And aren't the 5V analog regulators great, you short them out and their internal protective mechanisms save them and no smoke is evolved.
 
Hugo Holden said:
I have mentioned before, that it seems, that nearly every PET repair involves defective 2114 SRAMs. I'm not sure why the high mortality rate.

.......

It makes one wonder what is going wrong inside these IC's.
Click to expand...
They were made by MOS
 
FWIW,
R752 and R753 are a very common failure in the 12" monitor. They are 56 Ohm metal film resistors that get too hot and open up.

p.s. I always say, check the voltages first.
 
Last edited:
You must log in or register to reply here.
Back
Top