9th Jan 2021, 4:23 pm | #21 | |
Octode
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,440
|
Re: Commodore PET 2001
Quote:
I found that if you then try selecting on 'Commodore' (or 'PET') label, then you get a reverse-order view of the blog posts tagged with this - but only the last 3 (with last one covering a repair fault-finding investigation). Although selecting 'Older Posts' doesn't work on this. See: https://web.archive.org/web/20200618...arch/label/Pet I've done a search on what I've saved on my current PC's HDD, and found a few I'd save from 2017 - hopefully attached. |
|
9th Jan 2021, 4:42 pm | #22 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001
I must admit I had hoped when I get access to my 2001 to get a IEEE488-SD interface and possibly the diagnostic board, so I hope Mr Curran brings back this stuff or open sources it. Getting the Micro-SD to work took a lot of playing about apparently. I suppose the diagnostic wouldn't be too hard to re-engineer.
|
9th Jan 2021, 5:54 pm | #23 |
Octode
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,440
|
Re: Commodore PET 2001
From what he's posted on his website, he's only taken it down for the time being to save getting lots of enquiries when he's currently closed as he says he can't get things sent out in the way he thinks is safest / OK at the the moment.
But that it was his main source of income, so he does intend to re-open again. Attached are some info I found n PET diagnostics etc. on the 'net about 8 years ago and saved for future reference. So might have some useful info in it, in case websites aren't there anymore. |
9th Jan 2021, 6:11 pm | #24 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001
While we're gathering resources, I've found this online: Nick Hampshire's PET REVEALED book - I had this book back in the day and it was an invaluable reference to the hardware and software the PET. The linked document is the 2nd edition (I had the first) which covers both "old" and "new" ROM versions of the 2001. http://primrosebank.net/computers/pe...T_revealed.pdf
The site also has a lot of info about the PET. http://primrosebank.net/computers/pet/pet.htm |
9th Jan 2021, 6:16 pm | #25 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
Lots more useful information for me to study, thank you once again. I've had some more fun & games with the PET today and will provide an update on progress (or lack thereof) later. The wretched thing's got me hooked, for the moment anyway.
Alan |
9th Jan 2021, 6:33 pm | #26 | |
Octode
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,440
|
Re: Commodore PET 2001
Quote:
I haven't got too many books on the PET, as very rare to see any at radio rallies etc, where I've picked-up most others, but I do recall getting some A4-sized Technical Manuals (Probably by Commodore themselves), so will have to see what I've got. |
|
9th Jan 2021, 8:09 pm | #27 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001
The PET I had came with an A5 "start guide" ( http://www.commodore.ca/wp-content/u...r_Manual-2.pdf ) and a much less flashy A4 sized user guide which contained a listing of the monitor program that you could get on tape (I think if you asked nicely they'd give it to you, I got mine by visiting the Commodore sales office in London). Later "New Rom" pets had the monitor built in to the ROMs, so a SYS(1024) would get you there. I've been searching for years for this manual online and eventually found it here: https://www.manualslib.com/manual/82...re-2001-8.html
|
10th Jan 2021, 12:19 am | #28 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
Removed the motherboard and was pleasantly surprised to find it in very clean condition so I just brushed/wiped away the loose dust for the time being. Importantly there are no signs previous work to the board. The attached ‘photos give an idea of the general condition. As there are 30 multi-pin socketed ICs I decided to leave the video and PIA related chips alone for now leaving 24 for contact cleaning. Carefully removing and replacing a single IC is really quite time consuming if damage is to be avoided so I thought I’d just concentrate on the processor and the four critical RAM chips to start with. I was pleased to find nice bright shiny pins but cleaned them anyway before reseating each IC. Refitted the board and switched on. The result was no change with the same random character display. Having removed the board a second time I thought I may as well replace the potentially dodgy 1µF tantalum capacitor in the re-set timer circuit. Out of curiosity I used a 2.2µF aluminium electrolytic which in theory should roughly double the re-set pulse length. Replaced the board again and switched on to be greeted with a blank screen although I could just hear the transformer humming. In something of a panic I removed the board, put back the original capacitor and tried again only to find that the screen remained blank. Made another coffee at this point without knowing whether or not the higher value capacitor would have made a difference. Eventually I discovered that flexing the board near the PSU’s Molex connector brought the machine back to life. At present I don’t know whether or not this is a dry joint, broken track or just that the connections need a more thorough clean. The possibility of broken tracks in general concerns me as it’s impossible to unclip the board from the stand-offs without flexing it to some degree. Otherwise removing the board is quick and easy which is probably just as well. Bottom line is that there’s been very little progress despite all these words. Next instalment to follow, probably tomorrow, although I might be distracted by all the interesting reference material highlighted in this thread. Alan
|
10th Jan 2021, 1:27 am | #29 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001
Those Molex connectors are subject to a lot of thermal stress in operation, and mechanical stress too if the machine has been serviced. Its entirely possible the joints have cracked due to one of these causes over the years, and removing the board was just the final straw. Reflowing the joints, potentially with a touch of new lead based solder should sort it out.
The PCB is likely to be a little delicate, so be careful when soldering/desoldering, use a temperature controlled iron and keep the temp as low as possible. I'm probably teaching Granny to suck eggs here, but its worth saying. Over the years the glue sticking the tracks to the board may well have deteriorated, and I imagine wasn't as good as modern boards anyhow. That said, the bits of the MB you photo'd look in good condition with little corrosion, so thats a good start. |
10th Jan 2021, 10:34 am | #30 | ||
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
Quote:
Quote:
Alan Last edited by ajgriff; 10th Jan 2021 at 10:41 am. Reason: Re-phrased |
||
10th Jan 2021, 4:07 pm | #31 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
Just a quick update. Re-flowing the PSU connector joints cured the flexing problem and re-fitting the 2.2µF electrolytic in place of the tantalum makes no difference. Back to cleaning pins and re-seating ICs (more in hope than expectation).
Alan |
10th Jan 2021, 4:35 pm | #32 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001
Glad you sorted the power problem, thats one issue you don't need to have when diagnosing the PET.
I suppose what comes next depends on what tools you have available. Given you have established power is good and reset seems to be working then the next step is to check the CPU for activity - a scope or one of those cheapie logic analyser things would be useful. If the CPU is wiggling its address lines then y next I would swap the RAMs around to see if anything changes, then after that my thought would be that ROMs may be suspect. I don't know of any devices that can read 6540s but maybe you could rig something up...... |
10th Jan 2021, 5:17 pm | #33 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
I do have a couple of decent analogue 'scopes (not storage). However, because I've never tried to fault find logic circuitry to component level before I don't know which pins to probe and what to look for. Will do some research. I do also have a couple of VIC20s which were functional when last used twenty years or more ago. Unfortunately they're not readily accessible at present but presumably I could try checking the PET's processor by substitution?
My next step, apart from pin cleaning, was going to be checking RAM by swapping chips around. Incidentally I've found this web page useful in terms of getting me started on this adventure: http://www.dasarodesigns.com/project...dore-pet-2001/ If I get to the point of believing that faulty ROMs are at the root of the problem I'll need to consider the options and decide whether or not it's worth going any further. Alan |
10th Jan 2021, 5:28 pm | #34 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001
If you think it might be the ROMs or RAM then perhaps this $30 RAM/ROM and diagnostic board might help #https://bitfixer.com/product/romulator/
Last edited by Slothie; 10th Jan 2021 at 5:43 pm. |
10th Jan 2021, 6:40 pm | #35 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
Both the 'bitfixer' emulator and the Tynemouth ROM/RAM replacement board are possible options if I get to that stage although I might leave a future owner to decide on the best way forward for them. This is assuming of course that Tynemouth returns to trading in the fulness of time. For now I'll press on with my efforts to diagnose the fault(s).
Alan |
10th Jan 2021, 10:52 pm | #36 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
|
11th Jan 2021, 12:14 am | #37 | |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001
Quote:
Common faults on ROM,RAM and buffer chips can cause address or data lines to be shorted low or high, so if there is no activity on an individual signal on the bus then the fault can be with either the CPU or the peripheral. If there is no clock signal then there will be no bus activity too, but since the CPU clock is derived from the video circuitry and that all seems OK then you should have a clock (unless theres a broken track etc). Code:
Vss 1 |_| 40 ~RESET RDY 2 39 Phase 2 OUT Phase 1 OUT 3 38 SO ~IRQ 4 37 Phase 0 IN (NC) 5 36 (NC) ~NMI 6 35 (NC) SYNC 7 34 R/~W Vcc 8 33 D0 A0 9 32 D1 A1 10 31 D2 A2 11 30 D3 A3 12 29 D4 A4 13 28 D5 A5 14 27 D6 A6 15 26 D7 A7 16 25 A15 A8 17 24 A14 A9 18 23 A13 A10 19 22 A12 A11 20 21 Vss |
|
11th Jan 2021, 9:35 am | #38 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
That's brilliant, thank you. I've struggled to find anything helpful via Google. Will check out the CPU as you've described.
Alan |
12th Jan 2021, 3:40 pm | #39 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001
Have now 'scoped the CPU pins and the clock is clearly running fine at about 1MHz. All the data bus pins have sensible looking signals with similar amplitudes and waveforms. All the address pins have confidence inspiring Hi-Lo signals with the exception of A12 (pin 22) where there seems to be little, if any, activity. A12 connects to the 'A' input of the SN74154 decoder.
Can anyone advise me as to the implications of these findings? The relevant schematic is attached to post#17. Alan |
12th Jan 2021, 4:27 pm | #40 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001
No activity on A12 does seem very suspicious, because as the lsb of the page number it should be doing stuff as the CPU tries to access ROM, I/O, RAM etc.
According to the schematic A12 only goes from the CPU to the page decoder, so there are a limited number of scenarios, at risk of stating the obvious, it seems to me there are 3 possiblities: 1) The A12 Pin iof the CPU is damaged/not makeing contact in the socket (Unlikely since you are likely measuring direct from the pin). 2) The CPU is toast (Swap in your VIC-20 CPU?) 3) The 74154 input is shorted pulling the pin high/low. (Remove CPU & Check with multimeter?). (3) is not completly unlikely. If A12 is always high, it might be worth checking the PCB because A12 which is pin 23 on the 74154 is right next to Vcc (pin 24) and it would only take a tiny solder ball or bit of conductive gunge to short them together. Watching videos of retro repairs I've seen buffers with inputs that short to ground when they fail, so I suppose a decoder could do the same. I suppose there is little risk in swapping the CPU as "vanilla" 6502s aren't hard to find. The specialist variants in the C64 and later commodores or the Atari 8 bits are harder to find. I would imaging the new CMOS 6502s from WDC would probably work fine in a PET. I would probably at this point pull out the CPU, check that the A12 pin isnt shorted to ground or Vcc with a multimeter, then try swapping the CPU. |