Commodore PET 8032 Keyboard problem

[magicappstore]

Hi All,

I have a strange problem with my 8032 and hope someone can help. When I boot it up, the cursor is an inverted @ symbol, I can type but when I type something invalid and press Return it takes much longer than it should to process and show the "?syntax error" message and if I do this 4 or 5 times all the blank spaces on the screen turn to @ symbols, I can continue to type but there is something very wrong.

I've done quite a few repairs to the PET. I purchased it in a none working state and with the help of some of you on another thread I managed to get it to boot to BASIC showing 32K of good RAM.

The following have been checked and verified as working okay in my 4032:
RAM 4116 & 2114
ROM
6502
6522
6520
CRT Chip

I've uploaded a video here to show the problem.

I'd be grateful for any help with this.


Thanks,

Dan

 

[dave_m]

I have a strange problem with my 8032 and hope someone can help. When I boot it up, the cursor is an inverted @ symbol, I can type but when I type something invalid and press Return it takes much longer than it should to process and show the "?syntax error" message and if I do this 4 or 5 times all the blank spaces on the screen turn to @ symbols,

A PETSCII 'space' is 20H (0010) while a PETSCII '@' is 40H (0100) so I would have guessed a problem with RAM F7, but if you have tested the chip, then the problem may be in the video data path which includes (if I am looking at the correct schematic) SD05 at F9 pin 14 Latch, LD05 at F10 pin 23 6316 Character ROM, and E11 pin 4 LS165 shift register.
-Dave

 

[magicappstore]

Hi Dave,

Thanks for the info, I was looking for F7 and E11 but couldn't find anything with those labels, I guess those are for the older 2001 model (Mines an 8032)?

Something else I have noticed is that if you do not type anything and just press Return instead of doing a newline it shows the "syntax error" message and has actually got an @ on the screen above the cursor, almost like it has typed the @ on its own.

 

[KevinO]

Could it be this board layout?

http://www.zimmers.net/anonftp/pub/cbm/schematics/computers/pet/8032/8032030-6.gif

If so, character ROM is at UA3, but I don't know where those other chips are without digging into it a bit. You might start with that one and see what happens. Dunno if that would cause the whole screen to go crazy though.

 

[dave_m]

Hi Dave,

Thanks for the info, I was looking for F7 and E11 but couldn't find anything with those labels, I guess those are for the older 2001 model (Mines an 8032)?

Something else I have noticed is that if you do not type anything and just press Return instead of doing a newline it shows the "syntax error" message and has actually got an @ on the screen above the cursor, almost like it has typed the @ on its own.

OK, I had the wrong schematic. If the four Screen RAMs are OK, look at the latch UB3, character generator UA3 and shift register UA2 if the problem is reading the data from the video RAM and turning it into a video stream for the screen. Of course maybe the problem is when the data in written into the screen memory when the keyboard is scanned. I'll check into the schematic for things to look at.

 

[magicappstore]

Hi Dave,

Thanks for that I've had a closer look at UB3, UA3 and UA2, removing them and swapping for some know working ones from my 4032 and they all check out fine. The video RAMs are all good so maybe you're right, it is something to do with the chip(s) that handle the keyboard. I've already replaced both 6520's (with 6821's) as they were both faulty from a previous thread.

I'm curious as to what the ROM code is doing when it scrolls the screen, as when this happens the screen fills with @ symbols which were previously just spaces and all the text characters I may have typed change to different characters?

 

[magicappstore]

Finally I have this fixed now. In the end it turned out UB5 was faulty so after replacing it I have a working 8032 with a working screen.

One last question, when its turned on, is there a way to stretch the rows vertically as they all seem a bit compressed (vertical only, horizontal seems fine), at the moment theres a large amount of black space above and below the screen content?

 

[dave_m]

Finally I have this fixed now. In the end it turned out UB5 was faulty so after replacing it I have a working 8032 with a working screen.

So the problem was not in the screen refresh path, but with the data path to/from the screen RAM and the CPU. Good job.

One last question, when its turned on, is there a way to stretch the rows vertically as they all seem a bit compressed (vertical only, horizontal seems fine), at the moment theres a large amount of black space above and below the screen content?

There is probably a pot in the video board to control vertical size, but I am not familiar with the schematic. There are more than one schematic. Check Zimmers:
http://zimmers.net/anonftp/pub/cbm/schematics/computers/pet/8032/index.html

 

[MikeS]

Finally I have this fixed now. In the end it turned out UB5 was faulty so after replacing it I have a working 8032 with a working screen.

One last question, when its turned on, is there a way to stretch the rows vertically as they all seem a bit compressed (vertical only, horizontal seems fine), at the moment theres a large amount of black space above and below the screen content?

Congratulations! As usual, our good advice just motivated you to actually find the correct solution ;-)

If it's the most common monitor type there are two trimpots on the PCB directly behind the brightness control, Height and Vertical Linearity; that'll probably let you bring it back to normal height but it shouldn't be necessary, so there's a possibility that something's starting to fail in the power supply or vertical deflection section. Meanwhile, enjoy your success!

 

[magicappstore]

Hi Mike,

If it's the most common monitor type there are two trimpots on the PCB directly behind the brightness control, Height and Vertical Linearity; that'll probably let you bring it back to normal height

That's great, yep I've found it and got the height back to normal.

so there's a possibility that something's starting to fail in the power supply or vertical deflection section

I hope not but will keep it in mind if it starts to shrink again.

Many thanks to you all for the help with this. When I got this PET a month ago it did nothing when I turned it on and now, 4 weeks later its working perfectly with a lot of help from you guys. I've also learnt a lot about circuits, diagnosing PET problems, and 6502 assembler which will serve me well on my next PET repair!