Commodore PET 3032 Boot Problem

[dansretropod]

Hi Everyone,

I've got this 3032 PET that wont start up. When I turn it on I get the odd random character showing in random places, rather than a screen full of garbage. Sometimes I see the ### COMMODORE BASIC ### line, and then maybe a random character below.

I've removed both 6520's as one of these was faulty and I believe the PET should boot without these anyway, I just wont have a cursor.

The static RAM has been tested and is working and the 4116 chips seem to be good having removed them and tested them in another PET.

Which leads me to thinking its maybe a ROM problem.

The ROM's are in the following sockets:

UD9 - 901465-03
UD8 - 901447-24
UD7 - 901465-02
UD6 - 901465-01

Firstly is there a circuit board diagram for the 3032 that I can use to check the ROM's are in the correct sockets?

I own a 4032 which has BASIC 4. I assume this 3032 has BASIC 2 or something. I was thinking of downloading the BASIC 4 ROM's and putting them in there, but cannot find any references that tell me which ROM should go in which socket?


Thanks in advance!

 

[dansretropod]

Here's a pic of when I turn it on and manage to get something like a normal start-up screen, most of the time I get a blank screen, or, more usually a mostly blank screen with occasional random characters in various places.



I'm going to try my hand at some assembler to see if I can see if accessing RAM works.

 

[crock]

Firstly is there a circuit board diagram for the 3032 that I can use to check the ROM's are in the correct sockets?

Here you go:

http://zimmers.net/anonftp/pub/cbm/schematics/computers/pet/2001N/index.html

Although if the ROM's were in the wrong sockets you would never see the startup screen. I assume you have already tried reseating the ROM's and checking the voltages on the DRAM's? If your DRAM's have already tested good, your problem may lie in the RAM subsystem logic, although that would not explain the random character's on the display. I assume they are stable once they appear and don't flicker on and off?

Rob

 

[dansretropod]

Thanks for the info Rob. If I leave it the characters do occasionally flicker on and off, and sometimes new characters will appear at random positions on the screen. Sometimes I get the solid cursor appearing that flashes like mad! I would have thought that I would get the same result every time I turn the machine on, but no, it does something different each time!

I've been using the pet tester bin I found on another thread and that works and suggests my RAM is okay, though I'm now trying to write some assembler to try to read the ROM contents. I've gone ahead now and replaced all the ROMs with 2532 EPROMs with BASIC 2 on them but still the same results.

Many thanks for your help!

 

[dansretropod]

I've finally gotten some time to write some assembler to test the ram. I did a simple test to write the values 0-255 into the 1st 1k of ram 4 times. I then fill the rest of the 32k of ram with 0's.

I then, in an infinite loop write out the first 1k of ram to screen ram and this is where I see what the problem is. Most of whats rendered is correct, except for seemingly random characters that are wrong and come out as @ characters (Since I set all the other ram to 0 I presumed). To prove these random characters were coming from RAM locations above 1K I changed my code to clear ram above 1k to 255 and instead of random @ characters I got chequered flag pattern (255).

Here's a screenshot to help show what I'm doing:



It always seems to be the same memory locations that receive characters from the wrong place. Other characters will also randomly change and flicker.

So there's something wrong with the RAM or associated chips.

I've double-checked all 4116's, taking them all out and putting 16 others in from my 4032.
I've swapped the 74LS244's.

Am I right in thinking the problem may lie with the address bus? Has anyone had a similar problem with seeming random memory locations being changed by writing to other memory locations or randomly changing/flickering memory?


Many thanks!

 

[MikeS]

I'm not quite following this; if you fill video RAM with $FF (or any value) do you have the problem?

 

[dave_m]

Am I right in thinking the problem may lie with the address bus? Has anyone had a similar problem with seeming random memory locations being changed by writing to other memory locations or randomly changing/flickering memory?

It could be an address problem. You may need to run a NOP Generator test. But first, have you checked the two static video RAM chips (2114)?