• Please review our updated Terms and Rules here

Commodore PET 2001-32N not working properly

Tim Locke

New Member
Joined
Aug 19, 2012
Messages
4
Location
Sussex NB Canada
Hello. I purchased one of these machines today but it is not working properly. I am looking for suggestions how to figure out what is wrong with it.

When I turn it on, sometimes the screen is full of forward slashes with the cursor blinking in the top left corner. Other times the screen is full of forward slashes except there are some capital 'O's, question marks and left arrows on lines 4 and 5 with the cursor blinking in column 2 on row 6. The pattern of the characters on lines 4 and 5 do not seem to correlate to the expected boot up messages, although the "messages" seem to be consistent on every boot up.

Either way, typing produces either a forward slash, left arrow, capital 'O' or question mark. Shifted typing produces one of only 4-6 graphic characters. Some keys do not work, (eg. @, Z, X, V, N, M, Space). RVS, Shift, Cursor keys, Enter, Clr/Home and Ins/Del work normally. Pressing Enter displays a question mark and moves the cursor down one line. Typed commands don't seem to change the result of pressing Enter. I plan to clean the keyboard in hopes of getting the rest of the keys working.

Does anyone have suggestions on tests I should perform to determine what is wrong? It seems like a problem with the video chip, RAM or bus to me but it seems like there may be another issues as well. I have and can use a multi-meter but I do not have an oscilloscope.

Thank you for reading.
 
There were certainly inconsistencies in model numbering and of course many mother boards were replaced/upgraded, but a normal 2001 series PET did not have a "video chip" per se; it does sound like it might be a problem with the video RAM though, or related circuitry. For reference, what's the number on the main board and what are the numbers on the ROM chips?

Can you replace one or both of the 2114 video RAMs without too much trouble?

The usual first step is to reseat any socketed chips, being careful not to bend or break any pins in the process.
 
Thanks for the help. I have removed and reseated all of the socketed chips. No change in operation.

The number on the main board is "PART NO 320350 REV E". "135012673" is written in one corner with black marker.

It also says, "COPYRIGHT 1978/COMMODORE INTERNATIONAL". This system was purchased in Canada.

I'm not sure which chips are ROMs but here are the socketed ones:

UC4: MPS6502/2779
UC5: 7842/6522
UC6: MPS6520/4578
UC7: MPS6520/3278
UD3: 901447-?5/2979 (? may be a 2)
UD4: Empty
UD5: Empty
UD6: MPS901465/01/2579
UD7: MPS901465/02/2175
UD8: 01447-24/3979
UD9: 901447-26/5178
UF7: 2114UCB/7904
UF8: 915/MM2114N(or H)
UF10: MP901447/10/2279

I surmise UF7 and UF8 are the video RAM. I have no replacement chips. Are the two 2114s swappable? I ask because the markings are not identical.

I believe this system was originally owned by a university so perhaps some chips were replaced as they failed and are not the same brand as the originals.
 
UF8 is a likely suspect: the screen is normally filled with spaces (binary 0010 0000); a slash is binary 0010 1111 so it looks lik the lower 4 bits are staying high instead of low, and that's UF8. Try removing it to see if that makes any difference, also try removing UF7 and replacing it with the chip from UF8, and finally put the UF7 chip in UF8; that should tell us whether it's the RAM chip or the surrounding circuitry. 2114s are notorious, but also cheap and still readily available.

Something odd in those ROMs: UF9 is normally a 901465-03, and along with the chip in UD3 it looks like you have a couple of non-standard ROMs in there; once you get it working it would be interesting to see what they are since they have Commodore part numbers.

Where are you, by the way?
 
I'm in Sussex, New Brunswick. I've updated my location in my profile. I believe this computer was owned by Acadia University.

Removing UF8 didn't change the operation.

Moving the chip from UF7 to UF8 displays the reversed graphic character on the zero key in each cell. There is no cursor and typing yields nothing.

Placing the chip from UF8 in UF7 causes no display. Only a green dot in the middle for a few seconds.

For completeness I then removed the chip now in UF8. This displayed the reversed graphic on the question mark key in each cell. Again, no cursor and typing yields nothing.

Perhaps the chip from UF7 (now in UF8) is bad?

What are UD3 and UF9 supposed to contain? Do you mean UD9?
 
Well, since removing UF8 makes no difference but the chip from UF7 does, that would certainly suggest that UF8 is defective; I'd order a couple of 2114s and get that out of the way before looking to see what else is going on. Sounds like the keyboard also needs a little cleaning; something to do while waiting for the chips ;-)

Yes, sorry, I did mean UD9 which contains the machine's kernel while UD3 is usually unused. UD9 also contains the startup vector; quite possibly your chip contains a different vector telling the machine to run the code in UD3 before/instead of the usual startup routine.

Let us know when you've replaced the 2114s ;-)
 
I believe that 901447-25 and 901447-26 are the 2k/2316 versions of 901465-03. If the original 4k kernal ROM had failed, it would be possible to replace it with the two roms present, but I think it would need some minor modding of the board. Specifically, SEL_9 would need to be cut and SEL_F need to be routed to UD3. I *think* that might be all because the x000 and x800 2k rom pairs have opposite active states for CS2.

Tim, can you see any sign of modding on the board (it may be underneath?)

Rob
 
According to Martin's site 901447-12 and -14 were Swedish and Japanese character ROMs. I am not aware of any other ROMS in that range.

http://mhv.bplaced.de/cbmroms/cbmroms.php
Thanks for that link! I had misplaced it somewhere; glad to have it again.

I was just wondering what the missing numbers between 10 and 20 were; looks like maybe different national CGs.

Thanks again!
 
Looking at the back of the board, UD8 (NOT socketed) and the two video RAM sockets seem to have darker solder. Also, there are five thin, green wires linking UD8 to UD7 and UD9 as follows:

UD8p01->UD9p01
UD8p13->UD7p13
UD8p15->UD9p15
UD8p17->UD7p17
UD8p17->UD9p17
 
As these are all parallel links I would guess that ud8 was removed at some point and the tracks got broken. I am surprised there is no links to ud3, I cannot explain how the Rom there would work without some missing to the chip select lines.

Rob
 
Back
Top