Sharp PC-1211 (TRS-80 PC-1) Repair-a-thon #2. Turning 5 broken pokecoms into 4 working ones!

[@wallfris5285]

Thanks for another inspiring video. I'm always amazed by your tenacity in tackling these broken devices.

[@wimwiddershins]

Hmm, that looks like ancient Yutja. It appears to be a warning. "Collecting retro computer tech' is fraught with disaster!"

[@dugmeister65]

Great to see that you managed to get 4 out of 5 working…

[@HeyBirt]

That 5th one still bugs me though. I would love to find out what is wrong with it but now it is missing too many parts. Oh well, it gave its life for a good cause.

[@ohioterran7374]

Great job getting 4 out of 5 working! Great video and I love that you explain your process! Love it!

[@Fezzler61]

Who's on first? CPU #2! More amazing work!

[@benryves]

Good job on the repairs and it's interesting and useful to see the variety of board designs. I have a PC-1211 that suddenly stopped working properly; it functions as a calculator, but if you try to store something in "reserve" mode it immediately forgets it and if you enter a BASIC program line in "pro" mode it either hangs or translates it into nonsense (e.g. "10 A=1" becomes "600000:"). You can load a program from tape and it lists program lines at the end of the program correctly but early program lines are corrupt. My initial hunch is bad RAM so I guess I should start building a RAM tester like you demonstrated here!

[@HeyBirt]

The fixed variables, i.e. A-Z make use of some RAM in the display driver chips. Do A=5 in immediate mode and then PRINT A. Guessing though that it will work since it works like a calculator. Testing RAM is fairly easy I just had to add a few NOPs to account for how slow it is.

[@benryves]

@@HeyBirt I noticed that when you showed the service manual - fixed memories A-K in display chip 2, L-V in display chip 3 and W-Z towards the high end of main RAM. I can set those all in immediate mode and recall them fine, and also tried a bunch of flexible variables (e.g. A(100)) and those seem to work too. but it wasn't an exhaustive test as I can't write a program to test them on the computer itself! It looks like the reserve program is at the low end of main RAM followed by the user program, and as I can't store reserve programs and early program lines get scrambled I suspect it's whichever RAM chip that's responsible for the low end of memory that's to blame (RAM1, if they're numbered sensibly!)

[@MichaelEhling]

Would now be the time to drop the old cure-all: have you recapped it?

[@tenminutetokyo2643]

An Ard!