neoncluster.com

I saw this PET 64 in non working condition on Yahoo Auction Japan. I had never heard of a PET 64 until then and I started to do some research. SImply put, a PET 64 is a Commodore 64 in a PET body, with a few minor differences. How this machine ever got to Japan is a mystery. It was offered as a cheap and sturdy computer to schools, and was a marketing scheme by Commodore to use up refurbished Commodore 64 motherboards and un-needed supply of surplus PET cases. I bid and was surprised that I won the machine at a remarkably low price. Maybe the Japanese collectors were aware how rare these machines are.

Once the machine arrive I spent some time trying to get it to work. It was totally dead. Not even a flicker on the screen, Although I could hear the whine of the monitor, and rubbing the back of my hands over the screen I could feel a static charge, that told me that the CRT high voltage circuit was up and running at least.

I couldn't get my head around the motherboard at all. It was certainly in trouble as there were large areas where there was no pulsing at all, although I had confirmed that the crystal was producing a strong pulse. I contacted Ray Carlsen who is the leading expert on repairing Commodore 64’s. After sharing a few emails back and forth trying to trouble shoot the system, we decided it was probably best if I send him the motherboard to fix. He had never seen a PET 64 before and was interested in finding out more.

It took Ray less than a week to fix the board, which is surprising considering the amount of damage it had sustained. Here is an email from Ray explaining what he did to get the motherboard up and running:

I got the opening screen back but (now the bad news) to get that much I had to replace about a dozen chips: all eight RAM, the PLA, one of the CIA's, the BASIC ROM and the VIC... so far. Two of the RAM chips were overheating (indicating internal shorts) and the other chips simply didn't work. That many IC's bad clearly points to a failed power supply. The black "brick" used with the C64 fails like that sometimes (by over-voltage on the +5V line) and the same thing happens... lots of chips bad. I didn't expect that with your computer. An alternative scenario is that someone got into that computer and started diddling, which damaged those chips. Not knowing the history of that PET64, we must assume the PS still has the problem and it needs to be tested and fixed before the board is reconnected. There must be a 5 volt regulator IC in there somewhere... that's the main suspect. If someone has been playing around in there, rewiring, whatever, all bets are off. You can't know what they did but it has to be corrected. Otherwise, all my work will be wasted.

If you've been on my CBM repair website, you may have seen something I call the "computer saver". It's a circuit I designed that is installed on a C64 motherboard, a self-resetting circuit breaker. It monitors that supply voltage and "trips" if the voltage goes too high for whatever reason.  I'm thinking of installing that mod on your board... just in case.

To run your board on a standard C64 supply for bench tests, I installed a power socket. Since it doesn't extend beyond the board end, it should fit back in the PET case just as it is, so if it's OK with you, I'll leave it that way. If not, I'll extract it. Your choice.

Lastly, I don't know if the SuperPLA will work in this board... it should, but I ran tests on a similar board with my version of a mod PLA (made with an EPROM) and I had trouble getting it to work because of timing issues. Early C64 boards were very fussy about timing. That said, another after-market PLA replacement worked fine in that board. I have two of those and may just install one of them so you'll never have to worry about replacing it again.

Whew! I'll be running more tests on this board to make sure all other components are OK. Back to you....

Ray

I’ve never heard of a machine needing that many repairs. It was either hit by lightning or as Ray suggests there may be a serious issue with the power supply. With Ray’s guidance I tested the PSU and could not see any issues, just incase Ray installed his ‘Computer saver’ hardware before shipping the board back to me.

Upon installing the repaired motherboard into the original case, it quickly became obvious that the monitor was also non-functional. I could confirm that the motherboard was working by hooking up an external monitor to the video out port.

Again, I could confirm that the CRT high voltages were present from feeling static on the front of the screen.  I took some readings from the back of the CRT. and I could see that the +400V supply source was missing. It comes off the flyback as a pulse at pin 4, is "fused" via resistor R752, rectified by diode D752 and goes to the circuits that use that voltage such as the CRT screen (which now reads zero). Next I had to see if the resistor was open, and also if the diode were not shorted.

Finally I replaced the resistor on the monitor circuit board at R752. I also decided to replace R753 as it was also reading open. As predicted by Ray Carlsen, the monitor fired up perfectly in a beautiful hue of green. The machine now seems to be 100% fixed and the results can be seen above.

PET 64 after repair

Conclusion

Even with Rays help this was a difficult fix. I think it took over 3 months to get the machine up and running. With so many problems I was surprised that we ever got it running at all. Unfortunately I don’t have this machine any more. I was just too tight on space and eventually something had to go. The buyer seemed very motivated to get it for his large Commodore collection in Germany. I spent about 2 hours boxing it up (triple boxed) to make sure it would survive the journey. About 2 weeks later I received an email from the buyer. The machine had survive the trip physically intact with no external damage, however the screen had died and was showing a dark black area in the centre. I can only imagine that it suffered some failure due toy the low pressures associated with air travel. I was very upset. The new owner however felt that he could get a new screen for it from someplace and didn’t seem overly concerned.

Position of R752 and R753 on the video board