6th Apr 2021, 7:59 pm | #1381 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: Non-working Commodore PET 3016
Brilliant, thanks, Slothie. I did consider butchering the code myself but while 64tass assembles the code OK, I still can't get it to produce a binary output file! Good job we have you around.
|
6th Apr 2021, 8:47 pm | #1382 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Non-working Commodore PET 3016
The critical flag is the '-b' one! The Makefile in the zip has all the commands but you need the right tools installed. Cygwin has a good gnu make setup and the tools if you're using windows.
|
6th Apr 2021, 8:51 pm | #1383 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: Non-working Commodore PET 3016
Quite a lot of the above was pure Greek to me but don't worry, I'll Google it as and where necessary. As you have kindly provided an assembled binary, I am about to program it into a 2716.
|
7th Apr 2021, 11:13 am | #1384 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: Non-working Commodore PET 3016
What happened to the idea of swapping banks?
|
7th Apr 2021, 12:02 pm | #1385 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: Non-working Commodore PET 3016
It's still on the 'can do' list.
I've just (Wednesday lunchtime) posted an EPROM with Slothie's revised code, hopefully should arrive tomorrow. As with the earlier 'Slothie' EPROM it will need to be used with the adaptor if Colin is still using that, or modified slightly so that it does not need the adaptor. Hopefully this will give us a constant stream of write to / read from system RAM activity which will allow us to verify that all the write signals, read signals, chip enables, etc are all happening at the right time. |
7th Apr 2021, 5:24 pm | #1386 |
Dekatron
Join Date: Sep 2005
Location: Seaford, East Sussex, UK.
Posts: 5,997
|
Re: Non-working Commodore PET 3016
It would be an easy test just lifting one end of R41 & R42 (27 ohms) and linking the lifted end across to the opposite resistor hole.
|
7th Apr 2021, 6:18 pm | #1387 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: Non-working Commodore PET 3016
By all means give it a go while the EPROM is on its way.
|
7th Apr 2021, 6:45 pm | #1388 | |
Octode
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
|
Re: Non-working Commodore PET 3016
Quote:
However, they've already said they don't want to be confused by requests to try many other options at this stage, but to try and work through it in steps from one person. Having to post updated EPROM's can delay progress a bit, due to not having direct-access to their own IC-programmer (An Arduino / RPi 5V-only one could be quite low cost, if got these) and some DIL 5V parallel FLASH / EEPROM IC's (as it would have been quite easy to have attached various customised diagnostic routines here, to update these with) - It's a pity that the Dave Roberts PETTESTER V04 EPROM appeared to be staying in its screen-RAM Test routine's repeat whole test in an endless loop on readback-failure (seemingly because it couldn't read this back correctly, even though the first 1000bytes on screen looked at). As this did seem to have a good, full-reporting test of Zero-page etc. Main-RAM from looking through the source code, that would enable you to wrk out exactly which RAM IC(s) weren't working. So maybe a future V05 (it seems it has been updated a few times recently) of that could time out on screen test (especially if it seems to be mostly working) and try the main-RAM as well. And I presume you can't rely on keyboard / other Inputs working, to be able to control it via a menu etc. If anyone's in touch with him on the vcfed etc. forum, then maybe could make the suggestion to him, rather than potentially hacking the source and re-assembling ourselves to produce different versions to sending to others. But once all the faults have been identified with Colin's PET (It does appear there's an screen RAM fault lurking somewhere, that doesn't seem to show on the screen - other than now gone? '+32' issue. But it will be much easier to find this from a BASIC program etc., once main-RAM works enough to boot into BASIC), then knowing the exact faults could be useful for determining how future diagnostic versions will work best. Last edited by ortek_service; 7th Apr 2021 at 6:59 pm. |
|
7th Apr 2021, 7:01 pm | #1389 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
s there one of the ends of R41/R42 I should swap in particular? or won't it matter as long as I choose the same end of each one?
Colin. |
7th Apr 2021, 7:02 pm | #1390 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
And while I'm waiting for the EPROM, is it possible for someone to either explain to me how the code on the EPROM is called at boot? or is there a website I should go look at?
I'm interested in how the code is called in the first place. Colin. |
7th Apr 2021, 7:11 pm | #1391 |
Octode
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,265
|
Re: Non-working Commodore PET 3016
Same end of each one, but thats same electrical end which might not be the way its laid out on the pcb. Check which end of each resistor connects to the output of the gates driving them.
|
7th Apr 2021, 7:17 pm | #1392 | |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
R41/R42 are on page 6 of the schematics.
If I have this right, they are both driven by G7 (pins 12 and 6 respectively). Do I have that right? Colin. Quote:
|
|
7th Apr 2021, 7:19 pm | #1393 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: Non-working Commodore PET 3016
I don't think it's a bad idea to try the CAS0 / CAS1 swap, it's just that I try not to leap from one thing to another - at the moment we are focused on trying to get useful results from running test code but that can't proceed until the new test EPROM actually gets there, so it would, as has been said, be a simple thing to try in the meantime.
I did notice Owen's suggestion that the D2 test code is not proceeding because it is not reading what it expects to see back from the screen RAM, so my initial thought was that as we have the author of the other test code here we could ask him to make his do what D2's code currently does not seem to, which is to proceed beyond the screen test, and if the Zero Page check fails, continue to attempt the Zero Page check so that we have continuous activity to follow on the scope - and Slothie has kindly done that. As to why the screen RAM might not be reading back properly, there are two sets of buffers between the screen RAM and the CPU, going in the read direction. UE7 and UE8 isolate the display data bus from the buffered data bus BD0-BD7, on which the system RAM also sits. UI10 and UI11 separate BD0-BD7 from the CPU, so if we are getting bad reads from both the screen memory and the system RAM, then the UI10 / UI11 buffers are going to be quite high on the list of suspects. I have been inclined to replace those for a while but, to date, have never felt we had sufficient evidence to justify it. If Slothie's test code works as hoped we should be able to see the data passing (or failing to pass) through one or more elements of the UI10 / UI11 pair in the read direction. |
7th Apr 2021, 7:19 pm | #1394 | |
Octode
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,265
|
Re: Non-working Commodore PET 3016
Quote:
Described on page 8 of https://archive.org/details/mos_6500...ge/n7/mode/2up |
|
7th Apr 2021, 7:24 pm | #1395 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Non-working Commodore PET 3016
The EPROM is mapped into memory from addresses F800-FFFF. When a 6502 processor is reset, it looks in locations FFFC and FFFD for the address of the boot code. The EPROM has these locations programmed to the start of the diagnostics at F800! The original kernal ROM obviously had different code locations and therefore values in FFFC & FFFD but the principle is the same.
|
7th Apr 2021, 7:25 pm | #1396 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: Non-working Commodore PET 3016
Quote:
Disconnect the end of R41 currently connected to UG7 (12) and connect it to UG7 (6) Disconnect the end of R42 currently connected to UG7 (6) and connect it to UG7 (12). |
|
7th Apr 2021, 7:27 pm | #1397 | |
Octode
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,265
|
Re: Non-working Commodore PET 3016
Quote:
After reading Sirius last post it makes sense to suspect the buffers in the read direction from video ram, so I wonder if it would be better to wait for the updated test code, though it should be low risk to lift the end of the resistors and cross link them. |
|
7th Apr 2021, 7:31 pm | #1398 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: Non-working Commodore PET 3016
It will be fun to do, and may even surprise us.
|
7th Apr 2021, 7:31 pm | #1399 |
Dekatron
Join Date: May 2008
Location: Derby, UK.
Posts: 7,735
|
Re: Non-working Commodore PET 3016
There is a YouTube channel with a very interesting series (still in progress) in which someone builds a complete working computer system from scratch, on solder less breadboard, using a 6502 processor at 1MHz, here: https://www.youtube.com/user/eaterbc
__________________
If I have seen further than others, it is because I was standing on a pile of failed experiments. |
7th Apr 2021, 7:38 pm | #1400 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,484
|
Re: Non-working Commodore PET 3016
Quote:
|
|