Just got myself a commodore pet setup but not working

[SiriusHardware]

Pud has said that he intends to try to follow through on what has been posted but he won't be able to do it until the back end of the week, and maybe at that point he may choose to set up a meeting with Courtney. I'm afraid we will have to contain our excitement until then.

[ajgriff]

Originally Posted by SiriusHardware

You should maybe consider offering your cleaned-up version of the track side view to the zimmerman site as that's the first place anyone goes for PET technical info and it seems to be a missing piece of information there, unless we just didn't find it.

Bo Zimmerman’s site is a wonderful resource for PET owners but I still find it difficult to navigate even after a good deal of practice. There might be some track layouts hidden away in dark corners but I don’t remember seeing any other than for a couple for C2N PCBs included in a non-Commodore document. Presumably the best way to proceed would be via the contact form on zimmers.net. Copyright might be an issue I suppose but a discussion about that is way too far off-topic for this thread.

Alan

[SiriusHardware]

I did once find -a- PET monitor component layout which included what was supposed to be a transparent 'mask' showing the PCB tracks as though looking at them through the PCB but someone had managed to get the track overlay reversed east-west, so the LOPT was in (say) one of the upper corners and its PCB pads were in the opposite upper corner. Completely useless.

That's why that find by Colin is so vital, it is exactly what we needed in order to be able to indicate where to take the readings from.

I appreciated your work in summarising the places to measure from on the topside but I think access to those points with a scope probe with the PCB in situ would be very difficult with the tube and scan coils overhanging, you really would have to undo the board and slide it out and maybe fix it to the rear posts through the front holes to stop it from wandering off. That's not something I would have been comfortable asking Pud to try at this stage.

[ajgriff]

Originally Posted by SiriusHardware

I appreciated your work in summarising the places to measure from on the topside but I think access to those points with a scope probe with the PCB in situ would be very difficult with the tube and scan coils overhanging, you really would have to undo the board and slide it out and maybe fix it to the rear posts through the front holes to stop it from wandering off. That's not something I would have been comfortable asking Pud to try at this stage.

I certainly wouldn’t advocate anyone probing around on the top side of the board because of the risks you’ve highlighted. My suggestion was to use a test clip/hook attached to the test point with the machine powered off in order to avoid going near the monitor when live. It’s usually possible to attach the kind of clip I’m thinking of to a convenient component leg without creating a short.

My problem with taking measurements from the track side is still the size of the case apertures. With my 9” monitor the hole in the bottom of monitor’s case only allows access to about half the PCB. I’ve struggled to find images of a dismantled 12” monitor but did stumble across this video:

https://www.youtube.com/watch?v=X0xnPydHOuk

It’s not too long and it’s only the last half that is directly relevant. If pudwink’s monitor is the same there’s a cross shaped section dividing the bottom aperture in to four. From what I can tell this section makes a significant part of the PCB inaccessible. The only way I can see around this is to lay the monitor on its side on top of the main case having removed the four fixing nuts. Even then everything would depend on the wiring still reaching the PCB without the need for extensions.

Alan

[ScottishColin]

I think that video is an 8032. The photo on post 23 of this thread is a 4032.

Having said that, Commodore were pretty inconsistent in motherboards and cases across the range, even for the same model number sometimes, so we won't be sure until we see a photo of the PET in question.

Regardless, I don't think we will be able to do anything with the PCB in situ, certainly not if we need power for measurements.

Colin.

[ScottishColin]

This one's a bit clearer and is of a 4032:

https://www.youtube.com/watch?v=5asbfuc2OQE

Colin.

[SiriusHardware]

I followed that one as far as this frame. (Attached, with apologies to Rudy whose finger I have borrowed).

Although it doesn't extend as far out to the sides as I would have liked it does extend to all the way underneath the tube and that happens to be the area we need to get at, so I have hopes.

[ajgriff]

Originally Posted by ScottishColin

I think that video is an 8032. The photo on post 23 of this thread is a 4032.

I did realise that but it was the closest I managed to find and I'm pretty sure that the monitors will be the same. However having looked at the video again I didn't describe the issue very clearly. Firstly it's not a cross shaped section, it's T-shaped. Secondly the T-shaped section is part of the main case's pressing which is why I suggested that removing the monitor and laying it on its side might work.

Originally Posted by ScottishColin

This one's a bit clearer and is of a 4032:

https://www.youtube.com/watch?v=5asbfuc2OQE

Looking at your 4032 video the monitors do indeed look to be the same. However, there is no T-shaped section to get in the way so it may be possible to probe through the main case aperture although things could still a bit tight around the PCB edges. All useful stuff Colin thank you.

As you say only pudwink (or Courtney Louise) will be able to clarify the arrangement as far as his PET is concerned.

Alan

[SiriusHardware]

While looking through the book which Colin linked to in his other thread

http://www.primrosebank.net/computer...T_revealed.pdf

I noticed a very interesting little single-IC circuit to produce composite (monochrome) video from the PET's video and sync outputs. It's on PDF page 91 / Book page 86.

Could be very useful in a case where you suspect the monitor is faulty, but you don't have another one available to try. As far as I know PETs have the same low frame refresh rate as standard definition TV so you could theoretically run the output into any composite video input on a monitor, TV, or what have you.

[ajgriff]

Originally Posted by SiriusHardware

I noticed a very interesting little single-IC circuit to produce composite (monochrome) video from the PET's video and sync outputs.

Another similar design is shown in the attached images. The copyright owner has given blanket permission for individuals to copy the design for personal (not commercial) use only.

As you say a simple device like this could be a handy addition to the PET owner's tool kit. The monitor's bound to go wrong one day after all.

Alan

[pudwink]

Hi all
I am off now for a week so will get chance to have another look at the pet just a couple of things I want to put by you all.
Firstly I did try typing that line into the pet earlier on but nothing but I dont know if the keyboard is fully working yet but if I held the space bar down I did get a sound every few seconds.
I am not the bravest person when it comes to discharging the crt so as it has been switched off for a number of days so I was thinking of taking the crt board out and checking solder points etc but do I leave the motherboard in some say be careful as discharging the crt could spike the board any thoughts.
Finally I have found some posts of people using a petest rom which will also test the crt if there is some error on the memory I do have an epprom burner if anyone thinks this is a good idea and knows what I need to do to burn one.
Many Thanks

[SiriusHardware]

It sounds as though the system is kind of working when you say holding down a key makes the system make a noise.

What if you just hold down and then release CTRL and G together, does that make a noise? (you may need to press/release CAPS LOCK once first, if the machine has a CAPS LOCK key). Not knowing the machine well I can't guarantee that will work on a PET, but if it does work that will be useful to know.

Making a test EPROM, yes, good idea but no matter which one you make (I would suggest Daver2's test code as that can be programmed into a 2716 which will drop straight in) you will need a working display in order to see the output it generates, so, while a useful tool to have, you will not be able to use it until you have a working monitor.

Even if the PET has been switched off for days the tube will still be holding a full charge, in fact it will hold the charge indefinitely until it is intentionally discharged. After looking at various videos etc we think your best way to get at the measurement points on the monitor would be to unmount the whole monitor housing from the main plinth and lie it on its side so you can get to most of the underside of the monitor PCB through the 'Hatch' underneath. If you can take the monitor housing and its works off the plinth in one piece then you should not need to dismount the monitor PCB from inside the monitor housing at this stage.

Alternatively turn the whole machine onto one side and open the two halves of the case (upper and lower) up to 90 degrees apart, if that allows you access to the monitor PCB test points AJ and I indicated in earlier posts.

Have you given any consideration to Courtney's offer to look at it? We will assist you as much as we can for however long it takes but there is no substitute for having someone who knows their way around the machine sitting in front of it, and Courtney could easily try plugging the monitor of one of her working machines into your mainboard to see if your mainboard is producing video output, which we suspect it is. I don't know Courtney personally but I have the impression from reading her past posts that she has a lot of experience with all sorts of retro tech including CRT displays.

[ajgriff]

Originally Posted by pudwink

I am not the bravest person when it comes to discharging the crt so as it has been switched off for a number of days so I was thinking of taking the crt board out and checking solder points etc but do I leave the motherboard in some say be careful as discharging the crt could spike the board any thoughts.

There’s no need to fear the CRT. It only operates at a few kV which is low as these things go. Can still bite of course so some care is still needed. I’ve discharged my PET’s CRT a few times using a screwdriver with an insulated handle and a length of wire with croc clips on each end. One clip goes to a solid earthing point and the other to the blade of the screwdriver. Just slide the blade under the anode cap. I’ve never even heard the slightest click. The chances of spiking the board are practically zero in my opinion. None of this matters if it turns out that you can test the board without dismantling the monitor.

Originally Posted by pudwink

Finally I have found some posts of people using a petest rom which will also test the crt if there is some error on the memory I do have an epprom burner if anyone thinks this is a good idea and knows what I need to do to burn one.

Like Sirius I can’t see how any tester can diagnose a video or VDU fault unless the monitor is generating a display which this one clearly isn’t. I agree that PETtester V4 is the one to use but we haven’t got far enough yet to determine whether or not burning an EPROM is necessary. For all we know the PET might work flawlessly once we have a display. On which note can you please post oscilloscope photos of the three relevant signals as indicated in post #63 so that we can be absolutely sure that the monitor’s board is faulty before you go any further.

Alan

[pudwink]

Hi
Yes I will post all photos when I get them and I have made myself a purpose screwdriver with wire and croc clips I used on a sharp mz 80k my virgin moment on discharging a crt (oh baby).
My only what if is the video I posted at the begining https://youtu.be/sR6CnaRWPoI
where the same pet as mine no display and it was a ram chip at fault.
I am only a very poor novice and rely on the expert advice given by the members on here
which is very much appreciated but just a thought I had is if the crt is ok and it was a memory issue then the test rom would prove so.
shoot me down and tell me I am stupid and I will bow down to the forums superior knowledge I am just looking for the simple fix if possible.
many thanks

[SiriusHardware]

To be more precise: First of all see if you can get the whole machine into a position where you can see the waveform test points for video (1) and sync (3, 13), as indicated in this subsection from AJ's marked-up photo of the underside of the PCB (attached image #1). if you can't, see if you can unscrew the whole monitor housing from the plinth and separate it from the main body of the computer and lie it on its side so you can look at the monitor PCB through the square hole in the bottom.

If you can see them / reach them, scope the waveforms on those points.

Attached images 2,3 and 4 show what volts/div to set the scope to (ringed in green), what time/div to set it to (ringed in red) and where to position the line on the screen before making the measurement (ringed in pink) - in this case it says to place the trace so that it is sitting on the very bottom grid line on the scope before making the measurement.

When it comes to the video signal on waveform test point (1) don't worry if it doesn't look the same shape or length as the one on the diagram - that diagram may be intended to show what you should see when the display is showing a diagnostic 'test card' or some such. The main thing is that you see activity there and there are pulses about the right size, ie, number of squares high.

In addition to the settings mentioned above, scope channel 2 needs to be turned off and channel 1's AC/GND/DC switch needs to be set to 'DC'. It may be necessary to adjust the rotary TRIG control to get a steady 'picture' in each case.

[SiriusHardware]

shoot me down and tell me I am stupid

No-one on this forum will ever do that to you. The measurements we are asking you to make are to verify that the video and sync signals are as they should be. If they are, then you have a monitor fault rather than a computer fault. If that's so, we will try to walk you through the process of faultfinding the monitor.

Other than possibly separating the monitor case from the plinth, we are not asking you to dismantle the monitor into smaller parts, nor are we asking you to discharge the tube for the time being.

[ajgriff]

Originally Posted by pudwink

My only what if is the video I posted at the begining https://youtu.be/sR6CnaRWPoI
where the same pet as mine no display and it was a ram chip at fault.

This is why the video, horizontal drive and video drive signals are important. If the PET's main board isn't generating them correctly you'd conclude that there's a fault on the main board rather than a problem with the monitor. Does that make sense?

Alan

* Posts crossed and I echo the sentiments expressed by Sirius.

[ajgriff]

Originally Posted by pudwink

I am only a very poor novice .....

I am just looking for the simple fix if possible.

I competely understand the novice feeling as I only lifted the lid on a PET for the first time some 18 months ago and was very much out of my depth (still am in many ways). However, a couple of things I did learn early on is that it pays to be methodical and never be afraid to ask apparently dumb questions for fear of looking foolish. It's the best way to learn.

My PET's working well still (touch wood) but I found 8 faullty chips before it was finally functional. You might be more fortunate but be prepared. You're dealing with something that was cutting edge in its day but has suffered the ravages of time. Stick at it and you'll get there.

Alan

[SiriusHardware]

Just to recap, this (attached) is the front panel of the Hitachi V525 scope which I believe Pud is using. It may help us to be able to be more specific about scope settings when we are asking for any future measurements.

For example, to help the scope to trigger (to get a steadier 'picture' of each waveform of interest) the following settings will probably help.

MODE (beam mode) switch in lower centre of panel (CH1/CH2/ALT/CHOP/ADD) to CH1

INT TRIG switch at bottom centre of panel (CH1/CH2/Vert Mode) to CH1

MODE (trigger mode) switch in upper right of panel (AUTO/NORM/TV-V/TV-H) to AUTO or NORM normally, but for the VERT_DRIVE signal (waveform test point 13) you may get a better lock if you move it to TV-V

SOURCE switch in upper right of panel (INT/LINE/EXT) to INT

With the above settings in place it may still be necessary to both rotate the LEVEL (trigger level) rotary control and push it in or pull it out in order to get a steady picture of any waveform.

[ajgriff]

I've attached a slightly sharper image of the V-525 control panel. It’s a bit easier to read the legends.

Do please also ensure that the Volts/Div (Ch1) and Time/Div controls are adjusted in accordance with the settings circled in the attachments to post #95.

If it’s easier you can take the readings off the PET’s main board as you did previously. Once we’ve confirmed that the PET itself is generating the correct signals you can think about how you want to take readings from the monitor’s board.

Alan