blog:commodore_pet_restoration

This is an old revision of the document!


I picked up this old PET-2001 from a clear-out of old electronic waste a colleague was doing at work. It had been stored in the back of a workshop along with old toner cartridges, ancient CRT monitors, 10base network switches, 286 laptops and more, that were all destined for waste disposal.

Although it looked pretty beat up, it seems to be mainly dirt/sticker residue - the case is in pretty good condition apart from a few minor scuffs on the corners.

Internally, it is surprisingly clean, with just a little dust on the board. The transformer is also very clean with no corrosion that seems typical for these computers. Also no damaged caps or traces on the board, which is promising.

It also has an add-on board fitted, piggy-backed on a ROM socket. This turns out to be a SuperSoft HR40 High Resolution board, which provides a 320×200 pixel drawing interface on 40-column PET displays; normally the PET cannot draw pixels at all and is restricted to character display… so this is quite an interesting find.

There's a project to reverse engineer this board on Github.

Towards the rear IO edge connectors, the board is a little more dirty, where it has been open to the elements/dust. In the far corner there is a very small amount of surface rust that will need taking care of.

After re-wiring the UK mains plug (which had become frayed and loose), I checked the fuse in the PSU and powered the system on. Initially I got no response at all and ended up pulling out the multimeter to check mainboard DC voltages… at least a nominal +5v DC and more were present, so my heart was starting to sink at the prospect of the monitor being faulty… but no!…. it was simply that the brightness had somehow been turned all the way down! m(

Turning it about half way up then showed the little PET sitting at a perfectly normal Commodore BASIC prompt:

The first thing to do at this point was to make sure that I was able to type and that the computer was actually responding to user input… well, this didn't quite go as planned:

That screen is the result of pressing every key on the keyboard. It's plainly obvious that the keyboard has an issue at this point. The first course of action with a malfunctioning keyboard is always to try cleaning the contacts, and it was quite simple to do this with the old PET:

Keyboard traces and plunger contacts cleaned with alcohol - they were dusty, but in really good condition.

Back together and every key tested:

Result! 100% working keyboard 8-)

Cleaned the inside of the case and sprayed the spots that were starting to rust with a clear lacquer to stop it spreading:

Motherboard before and after:

I took images of how the SuperSoft HiRes graphics board was connected, as it's not just inserted into a ROM socket, but has 3 chip test-leads clipped to the legs of 3 seperate chips.

Also something that I did not notice earlier is a stacked (DRAM?) module, soldered on top of another module. This is also definitely non-standard:

On the underside of the board are a small number of bodge wires. They're not held on particularly well, so I don't suppose they are factory fit:

The PET BASIC V2 easter egg indicates that it has a V2 BASIC ROM (which it should, for a 2001 with full-size keyboard, unlike the early chiclet key model):

# The command is:
WAIT6502,X 
# Where X is a positive integer and the number of times the easter-egg string is printed

  • blog/commodore_pet_restoration.1637875282.txt.gz
  • Last modified: 2021/11/25 21:21
  • by john