21st Oct 2021, 10:51 pm | #161 | |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Quote:
Thanks again for your help with this. Tim. |
|
21st Oct 2021, 10:58 pm | #162 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,559
|
Re: Commodore PET 2001 voltage regulators
Duh, apologies, I have been mistakenly referring to the 6332 PROMs as 2332s, including in the drawing. All other details are correct though.
|
21st Oct 2021, 11:53 pm | #163 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,559
|
Re: Commodore PET 2001 voltage regulators
If you get geared up for another attempt, the UD6 PROM would be a good one to try as it has quite a bit of plaintext in it near the beginning (BASIC keywords) - although you will notice that the last letter of each keyword is 'missing' - that is normal even when looking at the code in a hex editor.
|
22nd Oct 2021, 5:07 pm | #164 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,559
|
Re: Commodore PET 2001 voltage regulators
The test PROMs have gone in the post this afternoon. In the meantime it would be good if Tim could manage to read and save the code from UD6 through UD9 and also UD5 if that can be managed.
The original code content for UD6-UD9 is, unsurprisingly, on the Zimmerman site in this section: http://www.zimmers.net/anonftp/pub/c...pet/index.html The relevant files there are: Code:
basic-2-c000.901465-01.bin (Code for UD6) basic-2-d000.901465-02.bin (Code for UD7) edit-2-n.901447-24.bin (Code for UD8) kernal-2.901465-03.bin (Code for UD9) |
22nd Oct 2021, 9:51 pm | #165 | |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Quote:
Okay, I have built the PROM adapter mark II and we have some success. The PET BASIC toolkit read fine, so I saved the binary file. As this seemed to work okay, I've carried on with the other PROMs (UD6-9) and saved the code for each. Every read/save was successful. I then downloaded the binary files that were listed (as referenced by yourself from Zimmerman), then performed the load and verify against the PROM; each of UD6, UD7, UD8, and UD9 verified correctly. As an additional check I also did a binary compare on Windows using the 'fc /b' command - it, again, found them to be identical (the downloaded binary file and the saved binary file that I read from the PROMs), with the exception of edit-2-n.901447-24.bin and UD8 as my saved file is 4K, and the edit-2 is 2K in size. The only one I haven't tried is the PROM in UF10 (which is also socketed) - the character generator. Can the adapter be used for this in order to check it's validity? Tim. |
|
22nd Oct 2021, 10:27 pm | #166 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001 voltage regulators
|
23rd Oct 2021, 10:33 am | #167 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,559
|
Re: Commodore PET 2001 voltage regulators
The reason for the size discrepancy with UD8 is that it really is a 2K device but as you are reading all of the devices as though they are 4K, that's how much /how many addresses the reader reads in, regardless of how many bytes the device actually contains.
If you look at the file you read and saved from UD8 you will find that only the lower half contains actual code, or only the upper half contains actual code, or both halves contain the same code. Whichever it is, if you use the block copy feature of the programmer's hex viewer / editor to move one copy of the code to the bottom half and save only the bottom half as a .bin file, that should give you a file which exactly matches edit-2-n.901447-24.bin (Code for UD8). I would say, like AJ, that the character generator PROM UD10 is 99% likely to be OK due to the fact that your random characters on screen are perfectly formed. |
23rd Oct 2021, 10:41 am | #168 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001 voltage regulators
When you get the test PROMs from Sirius they will show you if the character generator is OK, because they will get you a stable display on screen, but as AJ says the random characters show that its mostly OK. I suppose they could have the odd byte wrong but that is unlikely. As far as I am aware they used the same type of PROM in the character generator so the adaptor you have should work OK for the character generator if you do want to check the entire contents against a reference.
|
23rd Oct 2021, 10:54 am | #169 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,559
|
Re: Commodore PET 2001 voltage regulators
I should caution that at the moment, I do not think either of the test PROMs will work due to the fact that the machine is not clearing the screen, so it is (apparently) not able to execute the code in the original UD9 and UD8.
Now that Tim has successfully read the PROMs it appears we can rule those out as the cause of the problem so we are probably looking at a problem on the address or data bus (or both) with the buffers likely to be high on the list of suspects. Last edited by SiriusHardware; 23rd Oct 2021 at 11:01 am. Reason: spleling mitsake |
23rd Oct 2021, 12:02 pm | #170 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001 voltage regulators
If the PROMs are good, it may not necessarily clear the screen. When I was writing the diagnostic code I looked at the dissassembly of the original V1 ROMS, and the reset vector just initialises to the stack and jumps into the editor ROM at $E1E1. This is why the other guys diagnostic usually works; only the 10 or so bytes of the F000 PROM that it reads before jumping into the editor ROM need to be correct, so even if there is widespread corruption of parts of the Fxxx ROM then its possible the diagnostic will run. And in my experience of the 6540 mask programmable ROMs tend to fail on individual bytes rather than whole blocks., and I would imagine the 2332 /2316 types are similar.
Anyway, my point is that in the PETs initialisation code, it initialises all the hardware and some of the RAM locations and does quite a lot of stuff before trying to clear the screen, using the zero page RAM in particular. Its possible there is a RAM problem that is causing the initialising code to loop and preventing it getting to the screen clear code. My diagnostic clears the screen before trying to access any RAM, so it might actually work and indicate a RAM failure even if the normal ROMs don't clear the screen. These observations are based on the dissassembly of the PET V1 ROMs, but I imagine later versions are similar. Disassembly: http://www.zimmers.net/anonftp/pub/c...pet/rom-1.html |
23rd Oct 2021, 12:09 pm | #171 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,559
|
Re: Commodore PET 2001 voltage regulators
Don't get me wrong, I'm not suggesting that the test PROMs should not even be tried: I just don't want anyone to be unduly disappointed if they won't run in the machine as it stands at the moment.
|
23rd Oct 2021, 4:10 pm | #172 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,807
|
Re: Commodore PET 2001 voltage regulators
Hi - back off holiday now if you want any comparative testing done on my PET. It has the same motherboard but only 16k RAM.
Colin |
23rd Oct 2021, 8:25 pm | #173 |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Hi,
Just to let you know that the test EPROMs arrived today - many thanks again to SiriusHardware for sorting these out for me. As I now had these on hand, I swapped over UD8 for the Daver test. On powering up the PET we still had the garbled screen. I then replaced UD8, then removed UD9 and inserted in the Slothie EPROM; again, just the garbled screen. I had previously mentioned to SiriusHardware in a message that I had noticed that the VIC-20 and the PET used the same processor. As I had a VIC-20, I wondered if it was worth swapping this over to see if there was any change in behaviour. So after desoldering the processor from the VIC, I then inserted it into the PET. On powering up I was greeted by a Commodore BASIC screen! The socket for the processor does seem a bit loose, and the processor from the VIC did need to be pressed quite hard to seat it, but seems to work. I swapped them back over again to see if the original behaviour reverted, and it did; a garbled screen. So with the new processor in, the PET gets a clean screen (picture attached). To test it the other way, I tried the PET processor in the VIC-20 and it will not start. So this post isn't too long, I'll add the further testing next. Tim. |
23rd Oct 2021, 8:28 pm | #174 |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Forgot to upload the picture...
|
23rd Oct 2021, 8:31 pm | #175 |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
One thing I did notice when the PET booted was that there was no flashing cursor. I did try and swap the 6520s around in UC6 and UC7, but on doing this (after powering on the PET) I weas greeted by a blank screen. On swapping the 6520s back again, I had the boot screen and no cursor.
The next test was to run the Slothie EPROM in UD9. On powering up the PET it started to run, but gave the following error (picture also attached): KB : E0 ! So I guess this is down to the potential issue with the 6520? Tim. Last edited by retromit; 23rd Oct 2021 at 8:46 pm. Reason: Missed out the testing of the 6520s |
23rd Oct 2021, 8:36 pm | #176 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001 voltage regulators
Great news and well done! Can't comment on KB : E0 !
Alan |
23rd Oct 2021, 8:42 pm | #177 |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Thanks Colin, that's very good of you and hope you had a good holiday!
I had a read through your repair blog (with pictures), as well as the thread on here. It looks like it was quite the journey! The end result of the PET looks great. Tim. |
23rd Oct 2021, 8:48 pm | #178 |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
It was a bit of a relief, Alan! One thing I did miss out, and have edited my earlier post, was that although the boot screen came up there was no flashing cursor (and the keyboard unresponsive). I'm wondering if there is an issue with one (both?) of the 6520s?
Tim. |
23rd Oct 2021, 9:04 pm | #179 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001 voltage regulators
PIA UC7 would be my favourite as it interfaces with the keyboard and is prone to failure. Try swapping the two PIAs.
Alan Last edited by ajgriff; 23rd Oct 2021 at 9:18 pm. Reason: Correction |
23rd Oct 2021, 9:06 pm | #180 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,363
|
Re: Commodore PET 2001 voltage regulators
That is great news! No E0 idea either sorry
|