• Please review our updated Terms and Rules here

PET 2001-8 Not Booting

StickyRice

Member
Joined
Dec 4, 2021
Messages
17
Location
London , UK
I found a 2001-8 BS in a rubbish skip and remembering them from my youth decided to take it home and see if it worked.
Well of course it didn't , it had just one line of random characters in the centre of the screen.
It has 2114 RAM and 6540 ROM chips with a 320132 system board.

I built Stephan Slabihoud's Retro Chip Tester which I have found super useful in fault finding this PET.
It showed I had several faulty TTL chips in the video circuit, 10 bad 2114's and 4 bad 6540's in H3,H4,H6 & H7.

I replaced the bad TTL, RAM and now get a full screen of random characters.

Next I ordered some 6540 to 27C64 adapter boards from Retro Innovations. Built and installed them but still get same full screen of random characters.

I probably have others failing chips not located yet BUT I am unsure if I have configured the adapters correctly.
I have concatenated the 2k ROM images four times and burnt the 8K 27C64's EEPROMS.

Should I use all 5 resistors for a 27C64? Can I have a mixture of original 2K ROMs & these adapters with 27C64 8k EEPROMS in H1 to H7?
If anyone has used these in the past or can guide me in the right direction I would be grateful. I would love to get this machine working again.

Daniel
 
Welcome

I can't answer directly, but keep watching. It can sometimes take the experts a little while to get here (what with life etc getting in the way)
 
Hi Daniel,

Very good of you to rescue this PET! I recently went through a very similar repair on my PET 2001-8, only a slightly earlier board rev (had 6550s instead of 2114s).

It showed I had several faulty TTL chips in the video circuit, 10 bad 2114's and 4 bad 6540's in H3,H4,H6 & H7.

Wow! That's a lot of bad RAM. I had a bad H3 and H6 in mine. Plus one bad 6550 that died while I was working.

If you leave the PET on for a while, and it locks up + snows the screen, it is likely more bad RAM. That happened to me once I got mine to boot.

I replaced the bad TTL, RAM and now get a full screen of random characters.

Good. This is a step in the right direction.

Next I ordered some 6540 to 27C64 adapter boards from Retro Innovations. Built and installed them but still get same full screen of random characters.

I purchased the same adapters.

I probably have others failing chips not located yet BUT I am unsure if I have configured the adapters correctly.
I have concatenated the 2k ROM images four times and burnt the 8K 27C64's EEPROMS.

I used 2764s on my PET too. You were right to pad the ROM BINs to 8k before writing. It would not have hurt if you padded them to 16k either; I did this just to be sure.

Should I use all 5 resistors for a 27C64? Can I have a mixture of original 2K ROMs & these adapters with 27C64 8k EEPROMS in H1 to H7?

I made no modifications to the adapter. I did mix ROMs in the adapter and 6540 ROMs in the board. You have no problem there.

I assume you got these ROM images from zimmers? What BINs did you download; 901439 or 901447? You can get the Commodore part numbers off the 6540s themselves, to help you be sure. Here is the map of what was in my PET:

H1 901439-09 6540-019 ROM Basic 1r C000-C7FF
H5 901439-05 6540-012 ROM Basic 1 C800-CFFF
H2 901439-02 6540-013 ROM Basic 1 D000-D7FF
H6 901439-06 6540-014 ROM Basic 1 D800-DFFF
H3 901439-03 6540-015 ROM Basic 1 E000-E7FF
H4 901439-04 6540-016 ROM Basic 1 F000-F7FF
H7 901439-07 6540-018 ROM Basic 1 F800-FFFF

One thing I did get to assist in my testing was a ROM/RAM replacement board, so I could selectively enable ROMs and RAMs, while bypassing the main board. The one I got also has some built-in diagnostics (PETTESTER and VOSSI), which were quite helpful. Here is the link https://www.tindie.com/products/nivagswedna/romulan-pet-ramulator-with-eazy-nopulation/ The seller actually posts here, and I wouldn't be surprised if he jumped into this thread.

Let us know what you find next!
 
Thanks Gary C & DrAM19 that helps a lot knowing I'm doing it right. I was doubting myself .

I did download those 901439-XX images from Zimmer's brilliant site for the 4 bad ROMS I have.

So it would seem I have working ROMS now but another problem 'addressing' them perhaps?

I have made a NOP 6502 so its back to the scope and schematics.

Thanks again!
 
There is a 'standard' procedure for fault-finding a PET with a random display. Have a search of some of the similar threads where I have posted.

Power supply first.

Selected pins of the 6502 CPU next (to see if we are getting a clock, a healthy reset, no stuck interrupts etc. and a 'sign of life' on the SYNC pin).

Yep, the NOP generator will help next.

You can also download my PETTEST image. This fits in a 2K EPROM and replaces the EDIT ROM. You will need a valid Kernel ROM present.

My PETTEST doesn't require any functioning RAM to work.

There are test cards out there containing ROM/RAM mapping, a NOP generator and my PETTEST image. Perhaps purchase one of these?

Dave
 
Hi,

Definitely worth configuring one of the 6540 replacements with a PETTEST image (there is a variant that works at $F8xx) but normally you replace the $E0xx ROM and also have a Kernel ROM $F800... so you need to configure two.

Anyway... Having a full screen of Random stuff proves you have a working monitor, a working clock and at least some video circuitry but you could still have a dead CPU, transceivers etc,... be methodical...

Are you able to read the 6540s in your "Retro Chip Tester" ?

You have a scope and a NOP generator so... You can prove the lots of things but first do the basic checks...

Voltages, Clock, Reset,... then with your NOP generator running... Check BA0.. BA11 (the 7417 buffers and 74LS04 inverters acting as buffers do fail), then the outputs of the 74154...

These machines are surprisingly robust... many failures are due to the simple things... bad sockets, bad ROMs, bad RAM, bad CPU, ...

Have fun!

Regards
Nivag

PS
re: Adapters You should install all the resistors. I think you could probably get away with R3 missing as that is NC on a 27C64 but personally I wouldn't bother if I had one myself. Note that A11 and A12 are pulled high by R1 and R2 so you need your image at the top of the EPROM or for simplicity just duplicate it several times to fit.
 
Last edited:
There is a 'standard' procedure for fault-finding a PET with a random display. Have a search of some of the similar threads where I have posted.

Power supply first.

Selected pins of the 6502 CPU next (to see if we are getting a clock, a healthy reset, no stuck interrupts etc. and a 'sign of life' on the SYNC pin).

Yep, the NOP generator will help next.

You can also download my PETTEST image. This fits in a 2K EPROM and replaces the EDIT ROM. You will need a valid Kernel ROM present.

My PETTEST doesn't require any functioning RAM to work.

There are test cards out there containing ROM/RAM mapping, a NOP generator and my PETTEST image. Perhaps purchase one of these?

Dave

Thank you for answering Dave,
The CPU signals now appear good. But I still get a full screen of random characters even when using your PETTESTER and a kernal ROM.
At this stage I wont go for a test card, I'm not in any rush really and enjoy this challenge ..(at the moment :-D).

Daniel
 
Hi,

Definitely worth configuring one of the 6540 replacements with a PETTEST image (there is a variant that works at $F8xx) but normally you replace the $E0xx ROM and also have a Kernel ROM $F800... so you need to configure two.

Anyway... Having a full screen of Random stuff proves you have a working monitor, a working clock and at least some video circuitry but you could still have a dead CPU, transceivers etc,... be methodical...

Are you able to read the 6540s in your "Retro Chip Tester" ?

You have a scope and a NOP generator so... You can prove the lots of things but first do the basic checks...

Voltages, Clock, Reset,... then with your NOP generator running... Check BA0.. BA11 (the 7417 buffers and 74LS04 inverters acting as buffers do fail), then the outputs of the 74154...

These machines are surprisingly robust... many failures are due to the simple things... bad sockets, bad ROMs, bad RAM, bad CPU, ...

Have fun!

Regards
Nivag

PS
re: Adapters You should install all the resistors. I think you could probably get away with R3 missing as that is NC on a 27C64 but personally I wouldn't bother if I had one myself. Note that A11 and A12 are pulled high by R1 and R2 so you need your image at the top of the EPROM or for simplicity just duplicate it several times to fit.

Thanks for answering Nivag,

Really helpful info. Tonight I checked the 7417's & 04's they were good. The 74154 I have previously found bad and replaced. I seem to have had my fair share of failed chips with this board!

Yes I can read the 6540's on the tester which provides a 32bit CRC and able dump them to a on-board SD card.

I have started checking the ROMS for bad tracks...


Daniel
 
Assuming you have a working CPU you should aim to get PETTEST to work as the ROMs are directly connected to the data-bus and the buffered address (above).

With the NOP generator running you should probe the CS lines of the ROMs and RAMs to see if you are getting expected results.
 
I tried to reply to your message re F800 but i'm not authorised to post or download there, maybe because I'm a newbie.

Still got the garbage screen, perhaps I can post some scope pics of the chip select pins for F8 & E0 this weekend, see what you think?
 
Yes, you will need to get your post level to above 10. Just stick at it and you will get there shortly.

If you search the Commodore forum you will already find examples of what to expect from a NOP generator and the various chip select pins.

I’ll see if I can find one of the threads for you tomorrow if you luck out.

Dave
 
The CS traces should be quite obvious if you scope two signals at once e.g. A15 versus CS on ROM H7. A15 gives you the period of the NOP generator and the F8 should be a small chunk at the end of it.
I had some NOP pictures somewhere but for the life of me I cannot find them now.
Keep posting info and you will soon be unlocked.
 
Strangest thing..Already from last night, had CS1 pin17 H7 Kernal ROM connected to scope. Powered on scope and PET looking at scope screen CS1 it looked ok, then I noticed PETTEST was running!
First time ever.
ROM C=7800 D=7800 F=B688 but ROM B was Dxxx where xxx was changing values spuriously.
After the COUNTDOWN it failed memory test with
3 1 0480 10 FF so I replaced J7
Powered back on and ROM B was changing values only between 7800 & D800 as if replacing J7 had helped.
It failed the memory test again this time
3 1 1100 01 01 so I replaced I4
Powered back on got the old garbage screen...replaced I4 with another 2114 ...garbage screen!!
Checked memory from J7 and fails in my tester doing R/W's & I4 cant be recognised by tester now. This was a working new old stock 2114 I purchased. So is my board killing RAM or is it to be expected of old chips getting switched on and off?
 
I would completely de-populate the RAM for now and try and get a consistent results with just the ROMs. For $B000 obviously all bets are off.. that area isn't mapped.

If you look at the schematic you will see the ROMs are pretty much directly connected to the CPU, the video RAM and main RAM on the other hand are connected via pairs of transceivers (and these do fail). If you have seen good screen updates that would suggest at least two pairs of transceivers and their associated decoding is working. If I was you I would be tempted to remove the two transceivers associated with main memory H8 & H9 as part of the divide and conquer approach.
 
Last edited:
Yes I will try taking out all the memory tonight also I found a copy VOSSI PET diags that runs from F8 , so I might burn that and give it a try too.

I did previously replaced 1 failed bus transceiver in H8 or H9 and I9 was toast. Also both had failed in G5 & G6 and D3 & D6 have been replaced bad.

Appreciate your help.... we'll get this old girl working yet !! :)
 
Burnt VOSSI diags put it in F8 removed H3 and whammy diags up and working. Put H3 back in to see if it checksums and garbage screen! Must be H3 rom I thought ,took it out and same garbage screen again. Several power-on's garbage screen. Data bus levels looked a bit low on CPU. Replaced CPU with a Rockwell 6502A and whammy , powers up every time with VOSSI diags.
:D
 
I burnt some new ROMS that had failed and lo and behold it works... but after about 2 hours working I now have vertical collapse and can just make out the cursor flashing. So near but so far...
 

Attachments

  • photo65241.jpg
    photo65241.jpg
    99.2 KB · Views: 3
  • photo65242.jpg
    photo65242.jpg
    76.1 KB · Views: 3
So if you have H7 working (aka F800 to FFFF) then theoretically you could get the tester I sent you via message working too. I actually like VOSSI a lot, it's the only tester I had to hand when I started out, PETTESTER is a bit more useful though... e.g. you can test the keyboard VIA as well...

anyway... Either the ROM you are putting in H3 is bad or it is selected at the wrong time... either BA11 is duff or... you have a bad socket or bad trace to H3... a lot of these chip selects are active LOW which is quite unhelpful sometimes as unconnected might result in the wrong device being active.

Send some screenshots as you progress to make the thread more fun!
 
Back
Top