• Please review our updated Terms and Rules here

Stèphane's and Alex' PET ROM checks

carlsson

Veteran Member
Joined
Jul 30, 2003
Messages
6,275
Location
Västerås, Sweden
Yay, another PET thread! ;-)

In this thread, I will summarize the results of checking both Stèphane's and Alex' ROM sets. Both sent me two sets of ROMs:

* Stèphane sent me the Basic V2 ROMs (FR) that were present on the motherboard I sold him + Dave's 2532 EPROM's with Basic V2 (DME) sent before the original motherboard was replaced. I believe Stèphane might've had another set of V2 or V4 ROMs on his original motherboard, but those were not included.

* Alex sent me the Basic V2 ROM + EPROMs (AC, ACE) that were present on the motherboard I sold him + his old Basic V4 ROMs (IT) on the original motherboard.

I am using the Windows XP command fc /b to compare dumped ROMs with downloaded known good BIN files.

Stèphane
901465-01 FR = OK ($C000, Basic V2)
901465-01 DME = OK (2532)

901465-02 FR = OK ($D000, Basic V2)
901465-02 DME = OK (2532)

901465-03 FR = OK ($F000, Kernal V2)
901465-03 DME = OK (2532)

901447-24 FR = Blank?! ($E000, Editor V2 Graphics keyboard, no CRTC)
901447-24 DME = OK (2532, but only first 2K read)

I remember Stèphane's PET has the Business keyboard, so I probably need to program a 2716 with 901474-01.

I'll probably re-dump the 901447-24 ROM that was on the motherboard I sent to France and now has returned to me. It worked earlier this year and as all the other chips have survived two flights, it seems unfortunate if that one got wiped.

Alex
901465-01 AC = OK ($C000, Basic V2)
901465-02 AC = OK ($D000, Basic V2)
901465-03 ACE = Faulty ($F000, Kernal V2) (2532)
901447-24 ACE = OK ($E000, Editor V2 Graphics keyboard, no CRTC) (2716)
901447-14 AC = OK (Swedish character generator)

Again, I should try to redump the 901465-03 EPROM as it looks to be partly correct.

Here comes the most important part, checking Alex' original ROM set to figure out if he had a ROM problem or other motherboard fault which required him to get a new one:

901465-20 IT = OK ($C000, Basic V4)
901465-21 IT = Faulty ($D000, Basic V4)
901465-22 IT = Faulty ($F000, Kernal V4)
901465-23 IT = OK ($B000, Basic V4)
901447-29 IT = OK ($E000, Editor V4 Graphics keyboard, no CRTC)
901447-10 IT = OK (Character generator for both V2 and V4)

After closer inspection, bit 4 (00010000) is stuck on the 901465-21. The 901465-22 is partly good and partly bad. Again it could be worth taking another dump.

It means if Alex wants Basic V4 on his machine, I should burn him two 2532 to replace the bad ROMs. If he wants V2, I could include the 2532 already burned by Dave.

Any questions? Just let me know if anyone wants to see the dumps for themselves.
 

dave_m

Veteran Member
Joined
Feb 2, 2009
Messages
3,481
Location
Southern California, USA
Anders,
Somehow I had missed this message.

Great work! Now you can get both Alex and Stéphane's machines on the air. It would be good for Stéphane to have a working BASIC 4 set also. If I recall correctly, he ran Mike's Sumcheck test and found a bad kernal chip (901465-22) on that set.
-Dave
 
Last edited:

carlsson

Veteran Member
Joined
Jul 30, 2003
Messages
6,275
Location
Västerås, Sweden
I have now re-borrowed the advanced EPROM programmer that reads and writes 2532 (my own USB powered one doesn't handle those).

I re-read and verified all the chips and got slightly different results. All the following chips still verify OK:

Stèphane
901465-01 FR, 901465-02 FR, 901465-03 FR, 901465-01 DME, 901465-02 DME, 901465-03 DME, 901447-24 DME

Alessandro
901465-01 AC, 901465-02 AC, 901447-24 ACE, 901447-14 AC, 901447-10 IT

The following chips still read out as corrupt:
901447-24 FR, 901465-03 ACE (2532)

However when it comes to Alessandro's set of Basic V4 chips, the readout is different than what I got when I dumped them in the PET. I read all the ROMs as they would be 2532 EPROMs.

901465-20 IT = PET-OK, PRG-OK
901465-21 IT = PET-BAD, PRG-OK (?)
901465-22 IT = PET-BAD, PRG-OK (?)
901465-23 IT = PET-OK, PRG-BAD (?)
901447-29 IT = PET-OK, PRG-BAD (?)

I will re-read the last four ROMs a couple more times and compare to known good copies. It bothers me slightly that a chip that was dumped from inside a PET first verified as OK and when re-read in the programmer reads out as bad and vice versa.
 

carlsson

Veteran Member
Joined
Jul 30, 2003
Messages
6,275
Location
Västerås, Sweden
I am using the Windows XP command fc /b to compare dumped ROMs with downloaded known good BIN files.
That means I saved memory dumps to 2031 disk from the TIM monitor, moved the floppy to a 1541-II and read out the files with XM1541. The third step was to remove the loading address for each dumped file, and run fc /b on the remaining file, comparing it to the corresponding ROM file downloaded from Zimmers.

Although I have an own PET with Basic V4, I'm not sure if replacing the questionable ROMs one by one is a better way of fault detection than dumping them in two different devices.
 

MikeS

Veteran Member
Joined
Dec 23, 2005
Messages
7,493
Location
Toronto ON Canada
Ah, OK; I was just wondering how you got them from the PET to the PC.

I read all the ROMs as they would be 2532 EPROMs.
Probably not relevant, but you did remember that the 901447 and 901474 ROMs are only 2K?
 

carlsson

Veteran Member
Joined
Jul 30, 2003
Messages
6,275
Location
Västerås, Sweden
If I recall correctly, Stèphane ran Mike's Sumcheck test and found a bad kernal chip (901465-22) on that set.
sbo said:
I could determine the faulty ROM might be the EEPROM [D8] as the sum does not fit with the reference value.
Hm, unless I'm completely off target, UD8 would hold the Editor slot ($E000) also in Basic V4. I could probably make him two Editor EPROMs for Business keyboard, one V2 and one V4 assuming the other ROMs in the original V4 ROM-set were fine.
 

dave_m

Veteran Member
Joined
Feb 2, 2009
Messages
3,481
Location
Southern California, USA
Anders,
Yes, that was what I was trying to recall. It would be good for him to have two working sets. I think you mentioned that you had extra 2716's?
-Dave
 

carlsson

Veteran Member
Joined
Jul 30, 2003
Messages
6,275
Location
Västerås, Sweden
901465-03 ACE = Faulty ($F000, Kernal V2) (2532)
Hm, one leg had folded. When I unfolded it, the chip verifies OK again.

Alex sent me a PM on November 2 that his new board with my V2 chips booted fine. On November 6, it crashed into the monitor and thus I have the set back at my house to troubleshoot. Perhaps there are other, intermittent faults on the motherboard although when I tested it, I found it to be fine? Could a bad 6520 or 6522 cause the computer every now and then to fail to boot, or should I read the ROMs and EPROMs a fourth time? What about 2114 RAMs, should I have been able to bring it up to 32K on every occasion if there were temporary memory problems? I don't mind a bit of extra work but it would be fruitless if one can't solve the problem.

Furthermore, I find the ZIF socket on this EPROM programmer is a bit picky with how the chip is inserted. After fiddling a bit with it, I can successfully verify all the V4 ROMs except 901465-23. The pins on the ROMs look straight and reasonably clean, so the question is how good contact they make once into the sockets on the PET motherboard. Perhaps that is the real problem, ROMs and EPROMs not fully pushed into the socket so they only make contact every now and then?
 
Last edited:

dave_m

Veteran Member
Joined
Feb 2, 2009
Messages
3,481
Location
Southern California, USA
Anders,
Did you send Stéphane the ROM/EPROM he needed?

As far as Alex, if all his firmware was good, then there must be another problem. Maybe he can send you the mainboard for some troubleshooting?
 

sbo

Experienced Member
Joined
Oct 8, 2010
Messages
98
Hello,

Just to let you know, I have just received the package from Anders. (MANY THANKS !!!).
I am going to test the EEPROMs this week ; I will update this post with the results.

All the best.

Stéphane.
 

carlsson

Veteran Member
Joined
Jul 30, 2003
Messages
6,275
Location
Västerås, Sweden
Unfortunately Alex is still reporting problems. His set of Basic V2 ROMs booted OK once, but then became garbage. His V4 ROMs that I fixed up to a supposedly working state doesn't work at all with the motherboard I sent to him earlier and which checked out OK before I sent it. We are wondering if he might have a bad power supply in his 3032 that causes chips to go bad. He even tried to swap 6502 again. His 6520's and 6522 are expected to be OK. Since the board has booted OK two times, once first he received it and now again after reseating the chips, I would think the big chips should not be damaged?

We'll hope that Stéphane has better luck.
 

tezza

Veteran Member
Joined
Oct 1, 2007
Messages
4,720
Location
New Zealand
Unfortunately Alex is still reporting problems. His set of Basic V2 ROMs booted OK once, but then became garbage. His V4 ROMs that I fixed up to a supposedly working state doesn't work at all with the motherboard I sent to him earlier and which checked out OK before I sent it. We are wondering if he might have a bad power supply in his 3032 that causes chips to go bad. He even tried to swap 6502 again. His 6520's and 6522 are expected to be OK. Since the board has booted OK two times, once first he received it and now again after reseating the chips, I would think the big chips should not be damaged?

We'll hope that Stéphane has better luck.

Just catching up on these PET threads....

Ouch, that's too bad. Yes, it could be the power supply. I had an Atari 800xl with a bad power supply a while back. It kept blowing RAM chips until I finally figured out what was going on.

Tez
 

sbo

Experienced Member
Joined
Oct 8, 2010
Messages
98
Hello,

I have made some testing tonight.
It is 2:15AM here in Paris, and I am working on the PET since almost 3 hours.

I have tried to switch the original EEPROMs with the ones Anders kindly sent me.
Unfortunately, when I do so, I get garbage on screen, and the computer freezes in that state.

I do not know what to try to get it work.

Stéphane.
 

dave_m

Veteran Member
Joined
Feb 2, 2009
Messages
3,481
Location
Southern California, USA
I have tried to switch the original EEPROMs with the ones Anders kindly sent me.
Unfortunately, when I do so, I get garbage on screen, and the computer freezes in that state.

Stéphane,
It looks like we made a small step backwards. At one point, a V2 set only needed a correct "D8" screen editor chip for the business keyboard (901474-01) programmed into a 2716 EPROM. Is that the combination that you are trying?

Before with that set, but using the wrong D8 chip, did not you get to either the machine language monitor or the welcome screen?
-Dave
 

sbo

Experienced Member
Joined
Oct 8, 2010
Messages
98
Stéphane,
It looks like we made a small step backwards. At one point, a V2 set only needed a correct "D8" screen editor chip for the business keyboard (901474-01) programmed into a 2716 EPROM. Is that the combination that you are trying?

Before with that set, but using the wrong D8 chip, did not you get to either the machine language monitor or the welcome screen?
-Dave

Hi dave,

This is the situation :

1) OLD chips on the mainboard : I get the Machine Language Monitor.
2) NEW chips on the mainboard : I get a frozen state with garbage on screen.
3)- OLD chips with the NEW "D8" Screen Editor : same as (2).

Thank you.

Stéphane.
 

dave_m

Veteran Member
Joined
Feb 2, 2009
Messages
3,481
Location
Southern California, USA
This is the situation :

1) OLD chips on the mainboard : I get the Machine Language Monitor.
2) NEW chips on the mainboard : I get a frozen state with garbage on screen.
3)- OLD chips with the NEW "D8" Screen Editor : same as (2).

Stéphane,

If you still have access to the working PET, perhaps you could run another sumcheck on the D8 chip.

Or perhaps there is something wrong with the socket in D8. This is very puzzling.

Keep up your spirits as I feel you are close to fixing the PET.
 

carlsson

Veteran Member
Joined
Jul 30, 2003
Messages
6,275
Location
Västerås, Sweden
Please check carefully that all chips are in the right positions and no legs are bent. I know it sounds trivial but it can be easy to miss a leg bent inwards. Of course the motherboards I sent might be less than 100% operational, just that I got lucky and had both fully working when testing before shipping them.
 
Top