• Please review our updated Terms and Rules here

Commodore PET help!

Max1227

Experienced Member
Joined
Jul 4, 2015
Messages
78
Location
UK
Hello all,

I recently acquired a commodore pet 8096 upgraded from a 8032, 8050 duel disk drive and a 8023P printer.
None of any of the above components work, they all have problems and I was wondering if any of you could help me.
Firstly the PET, it chirps on startup but the monitor doesn't switch on and as a consequence nothing is displayed. From time to time I get a green light on the screen but this then drifts off the screen after 5 mins. The disk drive switches on and the lights come on but they don't change colour to green. Also, drive 1's green light flickers sometimes and I have only got the centre LED to go to green two times. This tells me it is probably a fault with the processor at UN1. The centre LED also went to orange once or twice?
The printer does absolutely nothing. I checked inside the printer and one red wire was cut so I fixed that and the transformer 'crackled' on startup, so not sure how to fix that.

Any help would be much appreciated.
Thanks in advance
 
As this is your first post - welcome to the forum.

Firstly, I would be tempted to put the printer and the disk drive away for the time being until you have the PET working.

Secondly - do you have any electronics experience or access to tools (such as a multimeter, logic probe or oscilloscope)?

From the chirp at start-up we know that the power supply is probably working enough for the logic circuits and that the processor etc. is at least running for long enough to give us the 'chirp'. We don't know whether the video and synchronisation signals are present from the Pet to the monitor - or whether the monitor is working.

Have you found the schematics for your particular version of the PET at ftp://www.zimmers.net/pub/cbm/schematics/computers/pet/index.html and can you read schematics?

The first thing (in my book to do) is to disconnect the power supplies to the PET logic board and the monitor and use a multimeter (set to read AC volts) to ensure that all the voltage rails are available from the transformer (and roughly reading right).

Beware, there will be mains voltages inside the case with it open!

Secondly, I would connect the logic board back up to the transformer (but NOT the monitor) and use the multimeter again (this time set to DC volts) to measure the voltage rails of the regulated supplies for being correct.

Next, I would use an oscilloscope to probe the video and synchronisation signals (horizontal and vertical) from the main logic board to the monitor. If they are all present and correct - the problem lies with the monitor. If they are not present and correct - the initial problem lies with the PET main board (there may also be a fault with the monitor as well - but we will leave that aspect for now).

If the problem lies with the monitor - there are extremely high voltages in here that are lethal and can kill. If you have no experience of high voltage electronics - and you suspect the problem is with the monitor - I would seek the help of a local TV repair shop (if they still exist within your area).

Dave
 
Welcome to the PET Forum.

Follow daver2's excellent advice. If the problem is no horizontal (20 KHz) and vertical (50 Hz) timing signals, then most likely the CRT Controller chip (6545) is not being in initialized by the boot process. Although recently there was the case of a bad CRTC. Since you are getting the power up chirp, a lot is working. We may have a bad ROM in the $D000 thru $FFFF space (3 ROMs), or bad zero page RAM (lower 16K bank).

Are any of the RAM/ROM memory chips on sockets? How about the two PIAs (6520) and VIA (6522)?

A question for daver2 and others is "should the 64K memory board be removed until the main board is fixed?" I am not familiar with the 8096.
 
Last edited:
Good morning Dave - it is evening time in the UK so I am about to go to bed...

I have never seen one of the 8032 64K expansion boards myself - so this is new territory. Just downloaded the user manual from Bo's website and 'had a quick nosy' at what it entails. It seems to just require power and swapping the 6502 CPU out onto the memory board - no PROM changes or anything to the main board.

Based on that I would think it is a good idea to remove the 64K memory expansion card by doing the reverse of the installation instructions (providing the OP has an antistatic bag and a suitable cardboard box to keep the board safe from harm).

My logic (and I am sure it is yours as well) is to reduce the amount of hardware you are debugging to an absolute minimum and build the system back up from there.

Good call Dave - missed that one.

Dave
 
Hello all,

May I take this opportunity to thank you for all you help!
Firstly in response to daver2 I do have a multimeter but no oscilloscope or logic probe. I don't think I could read schematics but I could probably learn to. My Pet does indeed have the 64k expansion board but it is easier said that done to remove it. The plugs that connect he power supply to the expansion board are different so you cannot simply 'swap' it out unfortunately. Power first goes to the expansion board as you would expect since it houses the 6502 CPU. I can do all the things you suggest with the multimeter. If we locate the problem to be in the CRT I'm afraid I will have to give up since I cannot deal with such circuitry.


In response to dave_m I too think this is the problem. However the CRT doesn't switch on at all. I don't hear any static or feel it on the screen when I switch the PeT on and off. I think power to the CRT is dodgy and there is a problem with the video ROM not sure about the RAM yet. I think all the ROM chips are socketed but none of the RAM is.

Thanks again.
 
However the CRT doesn't switch on at all. I don't hear any static or feel it on the screen when I switch the PeT on and off. I think power to the CRT is dodgy and there is a problem with the video ROM not sure about the RAM yet. I think all the ROM chips are socketed but none of the RAM is.

Don't worry about the CRT yet. Without the 20 KHz Horizontal Drive signal from the CRT controller (6545) to the video board, there is no switching signal to create the high voltage. As daver2 says, disconnect the J7 cable going to the video board until we know there are correct timing signals to the CRT.

Here is the schematic (sheet 10) which shows the Horizontal and Vertical timing signals coming from the 6545 CRTC to the J7 connector.

Are you able to check signals with the voltmeter on the main board or is the 64K memory board in the way? Is there no way to leave the main board with power and disconnect the memory board? You might have to transfer the 6502 CPU back to its main board socket at UB14. Does anyone have an assembly drawing of the 8096?
-Dave
 
Assuming the 8096 is an 8032 plus the 64K add-on board - the user manual for the 64K add on board is available at ftp://ftp.zimmers.net/pub/cbm/manuals/pet/CBM_8032_Expansison_Memory_Board_64K_Users_Guide.pdf. It looks a relatively simple matter to install the board - and removing it should be just as simple.

From a picture I have been able to find at http://www.primrosebank.net/computers/pet/photos/8096_open_4.jpg (and the manual) it looks as though the power is not even modified to the 8032 main board - so it should just be a simple matter of removal and re-inserting the 6502 CPU back into the main board. However, this is different to the statement made in post #5 regarding the distribution of the power supply cables. Can you double check (a) if this is correct and whether the PSU cable goes to the expansion board or the main board and (b) can you double check that the boards inside the case are actually an 8032 + 64K ram expansion and not a SuperPET(!?) In a SuperPET - the power does go to the 'expansion' board first and then to the main board... It wouldn't hurt to identify the part numbers on the boards and post them for us to check (or even to take a photograph of the insides - although you may not be able to post photographs on the forum until you have 10 or so posts).

+1 for trying to get the main board alone working and then worrying about the monitor later. It is interesting that you say in your first post that you sometimes get a green glow on the monitor - but then it 'drifts off the screen'. This is actually good and could indicate that the monitor may actually function if the main board was working (but no guarantees of course).

Do you need any help to identify where and how to measure the AC and DC voltages?

Dave
 
Hello,

I should be able to do all the things you suggest, however I cannot perform any tests for a while since I don't have the computer with me.

Power definitely goes into the expansion board first and I cannot switch it out (already tried). Any help with performing the tests would be much appreciated since I am not an expert in this field. If I can attach photos I will since a picture would be incredibly useful.

Thanks
 
..or even to take a photograph of the insides -

As daver2 says YES! Take a lot of photos as you disassemble (especially of the connectors). That way you can put everything back together correctly without plugging connectors in backwards. Marking pin 1 of connectors would be good too. Take your time, it is not a race. Get familiar with the assembly drawings and schematics. Understand things like the 6502 CPU is at UB14, which is B row, 14th chip over.
 
Hello,
Ok so I did some research and the expansion board is definitely a 64k expansion board, however power still goes to the expansion board before the main board and I cannot remove it as the power pin connectors are different, so it is essential to have the expansion board plugged in for power.

Thanks
 
Hello,

Thanks for the help. Would it be possible if we could discuss a solution to fixing the disk drive? (This is the part I really want to work)

Someone else on another forum suggested that the problem was bad caps?

Thanks,
 
Hi Max1227.

I am not disagreeing with you - but what you are saying doesn't agree with what I am reading about the 64K memory expansion unit.

It is possible that we are talking about a different 64K expansion board - in which case we need to identify what you really have so that we are all talking about the correct parts. Have a look at http://www.vintagecomputer.net/commodore/pet/64K_expansion/ and confirm that is what you have. If you scroll to the bottom of the page you will find links to more detailed images.

There were multiple memory expansion boards made for the PET from suppliers other than Commodore - so we may all be talking about a different item.

Please take a photograph of the PET insides and post it when you can (or have a nice word with the moderators). As a minimum, please identify the part numbers either printed or etched onto the boards (e.g. ASSY 8032110 PET 64K MEMORY). I see you don't have the PET with you - so that is probably a moot point at the moment until you do.

It is also always possible that one of the previous owners of the PET didn't follow the instructions and have 'modified' the installation of the board from what it should have been. If this is the case - then we need to work out what has been done from standard.

Dave
 
Hello,

I'm only going on what i saw when I opened the unit up. I immediately thought I'd remove the expansion board as it could be causing the problem, and it isn't nessesary for operation. It doesn't look like anything has been modified, otherwise it has been well executed.

Sorry for my delayed responses but everything I type has to be approved first. I will take pictures and see if I can upload them when I have the pet with me. In the mean time, if you type in 64k pet expansion board and go to images, you will find the one that is in my pet.

Thanks,
 
If we keep a little 'chat' going - eventually the mods will let you post unhindered... I see I can send you PMs now - so it may be worth looking at uploading any old image to see if you can do it.

I used Google to do exactly what you said - and most of the hits came up with what I had found already. The installation manual for the Commodore 64K memory expansion board has the installation as (paraphrased by me):

1. Remove the 6502 CPU from the PET main board and plug it into the expansion memory board.
2. Connect the ribbon cable from the multiway socket on the memory expansion board to the socket vacated by the 6502 CPU.
3. Connect the short power cable from the memory expansion board to J10 on the main PET logic board.
4. Connect the long power cable from the memory expansion board to J11 on the main PET logic board.

There is nothing about disconnecting the power connector from the main board and attaching it to the memory expansion board.

Manual: ftp://ftp.zimmers.net/pub/cbm/manuals/pet/CBM_8032_Expansison_Memory_Board_64K_Users_Guide.pdf pages 2-1, 2-2 and 2-3.
Images: http://www.vintagecomputer.net/commodore/pet/64K_expansion/

As you can see we have a discontinuity somewhere - so either your board is installed in a non-standard way or what you have is not what I have identified.

We will only know for sure when you identify what is on the actual board inside the PET and post a photograph.

Rather than give us some text to search for in Google can you actually provide a URL to what you think is inside your PET. This will make life a bit easier.

Dave
 
Hello,

Okay, I'll take a look at the PET this evening and get some photos taken. I'll edit this post when I've finished.

Regards,

Okay so it turns out that i was looking at it wrong and power does first go to the main motherboard. The jumper cable that takes DC power from the power supply is right at the front so not sure how I missed that, however the other jumper cable that take power to the expansion board is tucked away right at the back, easy to miss hence why I couldnt work out where power was coming from. In the meantime I recently bought another PET model 4032. This PET works perfectly but I have noticed some difference between both PET's, for example, the 4032 humming noises from the transformers is a lot louder, and humming comes from the back of the CRT, this isn't present in the 8096 which makes me think that the CRT isn't even switching on??

So just to clarify the expansion board is definitely a 64K expansion board and has been installed correctly. I was wrong.

Also, if you would still like me to take pictures I can now, but I don't see the point now that we have correctly identified the board.
 
Last edited:
Excellent,

We are both talking about the same 64K expansion card. No need for pictures. I just wanted to make sure we were not talking about different cards.

The CRT will not 'switch on' for a number of reasons. The first one is that the CRT (or the monitor circuit) is faulty. The second reason is if the PET main board is faulty. If there is no drive signal from the PET logic board to the monitor it will not 'switch on'. Hence it is very difficult to determine which of the two scenarios we are dealing with unless we measure some of the monitor's drive signals first. Before we do that, however, we first need to explore the power supply section of the PET and make sure that is operating correctly. It is good that you get an initial 'beep' at start-up - but that is not a foolproof way of identifying that the main logic board is fully operational.

My recommendations are as before:

Disconnect the power supply from the monitor.

Disconnect the power supply connector from the PET main logic board.

Check that the AC voltages are correct on the disconnected power connector to the PET main logic board (which I think is J8). See schematic http://zimmers.net/anonftp/pub/cbm/schematics/computers/pet/univ2/8032087-11.gif. Pin 1 to pin 4 should be about 8V AC. Pin 5 to pin 4 should also be about 8V AC. Pin 8 to pin 9 should be about 15V AC. For a 12" monitor, the voltage from the transformer to the monitor should be about 20V AC.

Turn the mains power off and reconnect the power connector to the PET main logic board.

You will now probably have to remove the 64K expansion board to get to the main logic board. Follow the reverse instructions for installation. Don't forget to transplant the 6502 CPU from the 64K expansion board back to the main logic board.

Power up the PET again (leave the monitor power supply disconnected at this time). You should still get the 'beep' as before.

Measure the power supply voltages on the PET main logic board with your multimeter.

VR1 pin 3 to VR1 pin 1 should be -5V DC.
VR2 pin 2 to VR2 pin 3 should be +12V DC.
VR4 pin 2 to VR4 pin 3 should be +5V.

Something doesn't look right about the pinout for VR1 - but that is what is on the schematic...

If the voltages look about right - measure the voltages on the video connector of the main logic board (with your multimeter set to 5V DC range - or the nearest available range to that).

Video J7 pin 1 (relative to J7 pin 7).
VERT DRIVE J7 pin 3 (relative to J7 pin 7).
HORIZ DRIVE J7 pin 5 (relative to J7 pin 7).

Can you post the voltages that you measure (for all the points) and we'll see where that takes us.

Before you make any measurements, can I ask someone to check over my post to make sure I have not screwed up something somewhere (it is getting late in the UK!)

Dave
 
Hello,

Thanks I will try and do all the things you suggest and post my findings but I will need help along the way.

I am about 1hour 15 minutes away from Worcester, so a meet up is definitely possible if that's what you are suggesting?

I don't have the specific tools for removing IC's and I do not want to damage any of the chips hence why my progress is slow.

If you have any further suggestions do let me know.

Thanks,
 
No problem - take it as slow as you like and we will help you as much as we can.

It is always better to think about each step and be perfectly happy with what you are doing before you do it and potentially make a mistake in haste. Most mistakes don't cause damage - but when I was away this week with work I did see a poor defenceless tantalum bead capacitor 'die a horrible death like a volcano' when the 5V power supply was connected to it the wrong way round...

I often travel around with my work (usually up the M5 and M6 from Worcester to Lancaster) so if you get into real difficulties (or need someone to help kickstart you on the right path) I am happy to do this if it is on my way.

Removal of ICs (in sockets) is usually accomplished by a small flat-bladed screwdriver. Gently 'walk' the IC out of the socket a bit at a time from both ends to avoid bending the pins. I have a wrist band (from Maplin) that I put on to ground myself before removing ICs and put them straight into antistatic foam and a protector bag when removed to prevent electrostatic and/or mechanical damage.

I would not suggest removing a soldered in IC until you are absolutely sure it is faulty - and then only when you are proficient at soldering and desoldering. If you aren't, I would suggest buying some veroboard and some cheap ICs from Maplin, soldering the ICs onto the veroboard and then trying to desolder them and resolder them back in without making a mess or damaging the copper tracks. You may have to try this a few times before you get 'the knack'.

[Incidentally, I don't own shares in Maplin...]

Dave
 
It is always better to think about each step and be perfectly happy with what you are doing before you do it and potentially make a mistake in haste.

To Max: Yes, and keep a logbook and write down each step you take, so you can backtrack if necessary. And take lots of photos. As daver2 says, plugging a connector in wrong, can ruin your day.
 
Back
Top