Commodore PET 4032/8050/3022

[SiriusHardware]

Regarding the results from the reset circuit, sorry but due to the difference in the diagrams I need you to do further checks - instead of

UE5 pin1
UE5 pin2

UD15 pin5
UD15 pin6

Could you please instead check and report the voltages on

UE15 pins 1 and 2
UD15 pins 5 and 6 (I know you already reported this one, I have a reason for asking again)

Do you have a pair of fine nosed tweezers? If you do, put your scope on the output of UD16 (pin 3) and momentarily short out C50 by joining its legs together with the tweezers. What happens to the voltage on UD16 pin 3 when you do that?

Also: With power off, measure the resistance from UD16 pin 7 down to 0V.

[ScottishColin]

16MHz confirmed on UE4 pin 6 and UE5 pin 6.

Colin.

[SiriusHardware]

OK, and the revised results from the reset circuit?

[ScottishColin]

Got to nip out - it'll be 2-3 hours before I get to these.

Colin.

Originally Posted by SiriusHardware

OK, and the revised results from the reset circuit?

[SiriusHardware]

I may also be busy by the time you get back to it.

When you short out C50 the output of UD16 (Pin 3) should go high for a certain (fairly short) amount of time, then go low again.

Moving to UE15 pin 1, you should see the same thing there when you momentarily short C50. Initially low, high for a bit, then back low.

On UE15 pin 2, you should see it go the opposite way: normally high, low for a short time after you momentarily short C50, then high again.

Same on UD15 pin 5 (high, low for a short time, high)

Then on UD15 pin 6 the same (high, low for a short time, high).

At the moment your CPU reset pin is held low all the time, so seems as though something in the reset chain is not working, hence these measurements.

You can also (with power off) locate R13 and measure its in-circuit resistance. It is a 1K resistor but will probably measure somewhat less in-circuit.

Is there anything connected to connector J4, and if so, what?

[ajgriff]

My apologies for muddying the waters with the wrong schematic. Ironically I also have a copy of the Technical Manual you're now using and it would have been much more helpful if I'd posted that in the first place.

Moving on, it does seem that this machine is an 80 column PET with a 12" monitor. The attached image is a crop of one of the photos originally posted by Colin. This being so the PCB waveforms shown later in the Technical Manual may prove useful as fault finding progresses.

Alan

[SiriusHardware]

I think the main mistake was not making sure we had visually identified the right PCB in the first place. It was only when Colin said this one didn't have a tin can next to the beeper that I thought... 'Uh Oh'. That was unusually imprecise of us.

[ajgriff]

My memory banks (almost as slow as a PET) have finally stirred and I think my apology was too hasty as the 4032 schematic I posted originally was extracted from the technical manual. Do you remember 'pudwink' and his 4032 from a year ago Sirius? Pretty sure it was the same PET version as Colin's.

Alan

[ScottishColin]

I'm pretty sure it's a 40 column display as it's a 4032. Also the jumpers look like they're set to 40 column to me; I believe they're called "Fat 40" machines.

It'll be a surprise for us all as and when it finally bursts into life.

I've not got to the measurements tonight and won't - they'll be done in the morning.

Colin.

[SiriusHardware]

I think my apology was too hasty as the 4032 schematic I posted originally was extracted from the technical manual.

Yes, the first diagram set you linked to is I think identical to the first one of two in that technical manual. I only belatedly realised myself that there were two distinct board versions covered in it. I kept not getting past the first one. I only discovered that manual because by chance the google summary of the search result contents included the phrase 'cut pins 10 and 11 of UDZ".

they'll be done in the morning.

No problem Colin, I don't imagine there is the same sense of urgency about this one as you have your original one to go and play with at any time.

I'm hoping this one will only have one or two faults as most do, and you are in a better place in terms of test equipment and knowledge than you were when you started out on the original one so it ought to go a little bit more quickly and easily.

It's funny though, I distinctly remember you passing up the chance to own a certain EPROM programmer on the grounds - among others - that it would take up a bit too much space, so I wonder how you managed to talk yourself into acquiring a whole second PET along with its peripherals?

Does this machine have an interesting life story or is it just a case of 'bought it off a bloke on the Internet', past history unknown?

I still can't get over that dangerous bare wire which was wrapped around the mains on-off switch. The most benign explanation I can think of is that whoever did that decided the machine was most likely 'dead' because the on-off switch wasn't working and decided to try bypassing it. It was certainly incredibly irresponsible of whoever originally did that to leave it that way.

[ortek_service]

I seem to recall once reading that Commodore had been known to have fitted one of upto four different mainboards - depending what they had to hand - in a particular model of the PET. So the model on the front isn't necessarily going to map to a unique board / circuit.

And it may not have been that a previous owner was a 'board-swapper'.

It may also be possible to have an '80 Column' marked mainboard, that has been reconfigured / had less memory fitted to be 40 column ?
(Could be an early case of a manufacturer providing enhanced hardware, but disabling it in lower-end models - like us now commonplace wuth test equipment etc, so they can charge you extra for a 'software-unlock' code

[ScottishColin]

I believe that this PET was one of 20 used at a coal mining office. I have some disks too in the purchase which are dated 1985 so I'll see what I find when I get to them too. I have asked the chap I purchased from for any history and I'll post it here if I get anything.

I actually bought the set as I really wanted the printer - Commodore dot-matrix printers come up very rarely that I have seen and it's a great match for my 3016 - the printer is a 3022. The whole purchase was £300 so I'm very happy with that for these three devices.

Colin.

Originally Posted by SiriusHardware

Yes, the first diagram set you linked to is I think identical to the first one of two in that technical manual. I only belatedly realised myself that there were two distinct board versions covered in it. I kept not getting past the first one. I only discovered that manual because by chance the google summary of the search result contents included the phrase 'cut pins 10 and 11 of UDZ".

No problem Colin, I don't imagine there is the same sense of urgency about this one as you have your original one to go and play with at any time.

I'm hoping this one will only have one or two faults as most do, and you are in a better place in terms of test equipment and knowledge than you were when you started out on the original one so it ought to go a little bit more quickly and easily.

It's funny though, I distinctly remember you passing up the chance to own a certain EPROM programmer on the grounds - among others - that it would take up a bit too much space, so I wonder how you managed to talk yourself into acquiring a whole second PET along with its peripherals?

Does this machine have an interesting life story or is it just a case of 'bought it off a bloke on the Internet', past history unknown?

I still can't get over that dangerous bare wire which was wrapped around the mains on-off switch. The most benign explanation I can think of is that whoever did that decided the machine was most likely 'dead' because the on-off switch wasn't working and decided to try bypassing it. It was certainly incredibly irresponsible of whoever originally did that to leave it that way.

[SiriusHardware]

I believe that this PET was one of 20 used at a coal mining office.

If so, the monitor may have spent hours on end displaying exactly the same thing so there is a possibility of 'screen burn', where an image of the software's front end / control panel / menu ends up burned into the phosphor on the screen.

If you look at the screen carefully you may be able to see a 'fossil' image of whatever it used to display all day long, and if so that may provide some insight into how it was used.

At one point in my career I was a field engineer working on CCTV systems, among other things. We had brought a CRT CCTV monitor back into the workshop to repair but the part took a while to arrive and by the time it was fixed nobody could remember which particular branch of company 'x' it had come from. Rather than ring around them all asking if they were missing a monitor (...bit embarrassing), we put it in a dark room and turned the brightness up a little bit and there was a faint scene etched on the screen - somebody said 'Oh! I know where that is, it's....' and that did prove to be the case.

So screen burns, while generally unwanted and undesirable, can be useful.

[ScottishColin]

OK - measurements:

UE15 pin1 - 0V (well, wobbling between 125mV and 157mV - I'll read that as zero)
UE15 pin2 - 4.55V

UD15 pin5 - 4.55V
UD15 pin6 - 1.38V

UD16 pin3 before/after shorting C50 - 0V to 4.38V then low

UE15 pin1 before/after shorting C50 - 0V to 4.45V then low

UE15 pin2 before/after shorting C50 - 4.42V to 0V then back up

UD15 pin5 before/after shorting C50 - 4.45V to 0V then back up (this pin had a square wave on it)

UD15 pin6 before/after shorting C50 - 1.73V to 0V then back

R13 resistance - 465Ω

Those low voltages on UD15 pin 6 don't look right to me - have I got that right?

Colin.

[SiriusHardware]

AJ mentioned Pudwink's 4032 from a year ago - to be honest I had completely forgotten it but that was because we didn't really 'repair' it as such, we just diagnosed it as a monitor, rather than mainboard, fault and Pudwink went off and found the solution for the monitor fault himself.

For reference, the thread is here:-

https://www.vintage-radio.net/forum/...d.php?t=191848

AJ did helpfully separate the circuit diagrams for this particular variant (no tin can / no UF1 next to the beeper) into a single PDF, which I hope he won't mind me attaching here again as it will save us from having to scroll up and down through the whole manual, much of which is not relevant to this version.

One of the other differences between the two variants is that the one we initially thought we were working on only has a single 5V regulator whereas the one Colin is actually working on has twin 5V regulators. Therefore, in addition to the supply voltage measurements we requested earlier, we also need to check the voltages on:-

CR6 ringed end.
CR6 non-ringed end.

[SiriusHardware]

Those low voltages on UD15 pin 6 don't look right to me - have I got that right?

Yes, UD15 is a buffer, not an inverter, so the output should follow the input (high when input high, low when input low). Although the output is changing state, it is not changing by enough.. The simplest explanation is that the UD15 buffer is faulty.

However...

You mentioned having nothing (which I assumed to mean 0V, or near enough), on the reset pin of the CPU. This doesn't really square with there being 1.73V on the output of the UD15 buffer, so could you check (with power off) for continuity between UD15 pin 6 and the CPU RESET pin and if there is continuity, check the standing voltage on the CPU RESET pin again.

Whether 0V or 1.73V, both of those voltages are low enough to keep the CPU held in reset so that's starting to look like a proper fault.

-Is there anything connected to the J4 connector, if so, what?

You may note that reset output from pin 6 of UD15 doesn't only go to the CPU reset pin, it goes to a lot of other places besides and so any one of the many chip inputs connected to this circuit node may have failed low resistance and be causing that fault.

Subject to verification by checking the continuity, pin 6 of UD15 is one output driving many inputs which makes it the hardest-working IC pin on this circuit node, therefore I would be looking at ruling out UD15 first.

How do you feel about taking UD15 out in one piece? If you can do that and put it in a socket with pin 6 bent out sideways, that may rule UD15 in or out.

[SiriusHardware]

I should say - at some point in the past you mentioned that you had been using a Dremel to cut the pins off ICs - for the reasons I gave then - the risk of fine conductive metal powder ending up everywhere and the risk of generating static which could damage other, healthy ICs which are connected to the one you are trying to remove, please don't use that method if you are going to cut pins and remove ICs. If you can get them out in one piece without causing any damage to the PCB then that's great.

[ScottishColin]

In reverse order, nothing on J4.

And no continuity between UD15 pin6 and UB14 (6502) pin3.

I do however have continuity between UD 15 pin 6 and the following:
J4 22
UB14 (6502) pin 40
UB13 pin 2

Colin.


Colin.

[ScottishColin]

Could you educate me - how does this schematic show connectivity between UD 14 pin 6 and pin 3 of the 6502?

Colin.

[ajgriff]

For ease of reference I'm attaching the file I intended to post earlier. It's an extract from the Technical Manual with the PCB layout and schematic in landscape format. It's much easier to refer to than constantly delving into the full manual.

I have on file a number of extracts and 4032 related documents accumulated during the 'pudwink' thread. This is a link to that thread should it prove useful moving forward:

https://www.vintage-radio.net/forum/...d.php?t=191848

Alan