• Please review our updated Terms and Rules here

Commodore Pet smoke from PSU? (With photos)

wallaby said:
I'm not sure what you mean by the first part of the line "/SEC C on"
Click to expand...

SEL C is the name of the signal on D2-pin14. The slash(/) in front implies it is a negative going pulse. This signal goes low (enabling the D6 ROM) during the address range of C000 to CFFF. When in doubt, check the schematics to clarify instructions.

14, 15, and 16 all stay high and don't pulse.

17 pulses once (from high to low)

CPU pulses on 3, 39, 37 (input and output?)
Click to expand...

This is good information. I think it is telling us that either the kernel ROM (D9) is bad as the address seems to get stuck there, or perhaps the low RAM is bad. If zero page or the stack space on page one ($0100-$01FF) is bad, the machine will not boot. Get some good 4116 RAMs and perform a piggyback test on the lower (J row) RAMs. In the mean time I can send you a D9 ROM. Provide me a mailing address by private message on this Forum. Things may be looking up.
-Dave
 
dave_m said:
This is good information. I think it is telling us that either the kernel ROM (D9) is bad as the address seems to get stuck there, or perhaps the low RAM is bad. If zero page or the stack space on page one ($0100-$01FF) is bad, the machine will not boot. Get some good 4116 RAMs and perform a piggyback test on the lower (J row) RAMs. In the mean time I can send you a D9 ROM. Provide me a mailing address by private message on this Forum. Things may be looking up.
-Dave
Click to expand...

Thank you for that. I'll send a PM. I ordered about 10 4116 RAM chips a few days ago and expect them to turn up in a few days. The IC removal tool should arrive soon too.
 
wallaby said:
Thank you for that. I'll send a PM. I ordered about 10 4116 RAM chips a few days ago and expect them to turn up in a few days. The IC removal tool should arrive soon too.
Click to expand...

OK, we have a plan. I'll program a 2532 EPROM which is pin compatible with the 2332 ROM and pop it in the mail. The EPROMs I've sent to PET guys in Europe took about ten days. I wonder how long it takes to get to Down Under? :D
 
dave_m said:
OK, we have a plan. I'll program a 2532 EPROM which is pin compatible with the 2332 ROM and pop it in the mail. The EPROMs I've sent to PET guys in Europe took about ten days. I wonder how long it takes to get to Down Under? :D
Click to expand...

Thanks, the ROMs arrived today. I tried replacing the originals one by one but there was no change at all. My logic probe no longer works for some reason but I measured the voltages and the same pins are staying high as before. I tried piggybacking some ram, but that's a very inexact science, and nothing made even the slightest change.

Could it be the ram? Is there any way to test them without having to remove them from the board? They're the only thing without a socket.

I don't have a de-soldering iron, and I'm getting more reluctant to buy tools for this Pet.
 
wallaby said:
Could it be the ram? Is there any way to test them without having to remove them from the board?
Click to expand...

Yes, there is a program called Pettest that fits in an EPROM in the D9 kernel socket. It will test the RAM chips. You would need the 4K version that needs a 2532 EPROM.

Do you know someone down-under who has a PROM Programmer and could program one of the 2532 EPROMs? A good candidate to use would be the D6 EPROM. It would have to be erased using a UV light first.

Lacking that capability, you should:
1. carefully examine (using a magnifying glass) the ROM sockets for bad connections like cold solder joints or solder splashes. An intermittent or bad connection could be the cause of your problem. If in doubt, touch up the solder connections.

2. Carefully re-try the RAM piggyback test making sure all 16 pins are in contact. Tack solder two opposite corner pins if necessary. Power up after you have piggybacked a part looking some sign of life. It is only necessary to do the lower eight RAM chips (I row).

3. Replace the lower eight RAM chips.

Without a scope there is little else you can do to troubleshoot beside perhaps blindly replacing other chips. It is kind of bad luck that a bad ROM wasn't the problem. I forget, besides pulling the 6520s and the 6522, did you try new 6502 CPU? They rarely fail but who knows.
-Dave
 
In the past I had a similar problem with my 3032 (2001N/32). I'm sure Dave will remember, since he and Mike_S helped me a lot with it.
I had a lot of problems... bad ICs, and a broken trace. I had the PET ROM/RAM board (http://vic-20.de/x1541/hardware/petram.html) that helped me a lot to discovery what wasn't working, and an EPROM programmer that was absolutely the ultimate magic.

I discovered that, due a broken trace, the BASIC 2.0 ROMset simply didn't work. I tried the BASIC 4.0 ROMset and the computer booted with no problem. So, after thousand tests and ROM comparisons, I found the BASIC 2.0 hex code (few bytes) who prevent the computer from booting. Somebody here in the VCF told me what to phisically check, so I found the broken trace.

Before to ask help here, I spent about 10 years trying fixing it... I replaced *all* chips, and it still didn't work... But I needed some tools: the RAM/ROM board, and an eprom programmer. I bought the G540 Genius, a cheap, chines programmer who worth every single cent I spent. Currently it costs about 40 euros on eBay:
http://www.ebay.it/itm/New-Genius-G...986290?hash=item1c4531f572:g:tnwAAOSw3xJVXpdv

If you choose to buy an eprom programmer, don't forget to buy also an eprom eraser, they're cheap, about 10 euros.
http://www.ebay.it/itm/Ultraviolet-...632875?hash=item462aca9f6b:g:0pAAAMXQydtTOL7F


Too bad Nicholas Welte no longer offers its products (at least until some years ago he didn't answer emails), but it seems this guy is offering the same product:
http://blog.tynemouthsoftware.co.uk/2015/09/commodore-pet-rom-ram-replacement-boards.html

This board will replace all system ram and romset, so you can easily exclude two big source of problems. If you have some problem with ram or rom, the computer will boot. You can then choose to leave the board in place, or to chase the issue as I did.

You will need the eprom programmer to program the flash memory; in this case, I can send you the rom image ready to be programmed, I should have it somewhere, just post me in PM.


About wiring on the solder side: I don't know about 3032, but one of my two 8032 has some wiring, due to revision on the board. When I had to fix the other one, replacing the ROMset with the ROM image from the zimmers.net site, I had to make the same wiring mod, because it refused to boot with the new ROMset.

You could be in the reverse position: the wiring mod could be prevent the computer from booting with the new romset.
Take a look to that wires and try to see if they were made by somebody who needed to fix some bad traces, or if they were made at factory. In this case, maybe the new ROMset could be incompatible with that wiring mod. It's just a theory, but I think it must to be investigated.


At least, you have Dave on your side! LOL

Good luck.
 
Last edited:
dave_m said:
Yes, there is a program called Pettest that fits in an EPROM in the D9 kernel socket. It will test the RAM chips. You would need the 4K version that needs a 2532 EPROM.

Do you know someone down-under who has a PROM Programmer and could program one of the 2532 EPROMs? A good candidate to use would be the D6 EPROM. It would have to be erased using a UV light first.

Lacking that capability, you should:
1. carefully examine (using a magnifying glass) the ROM sockets for bad connections like cold solder joints or solder splashes. An intermittent or bad connection could be the cause of your problem. If in doubt, touch up the solder connections.

2. Carefully re-try the RAM piggyback test making sure all 16 pins are in contact. Tack solder two opposite corner pins if necessary. Power up after you have piggybacked a part looking some sign of life. It is only necessary to do the lower eight RAM chips (I row).

3. Replace the lower eight RAM chips.

Without a scope there is little else you can do to troubleshoot beside perhaps blindly replacing other chips. It is kind of bad luck that a bad ROM wasn't the problem. I forget, besides pulling the 6520s and the 6522, did you try new 6502 CPU? They rarely fail but who knows.
-Dave
Click to expand...

Hi Wallaby,

where in Oz are you ? I am in Melbourne so may be able to help some, as I have desoldering station, Oscilloscope and eprom programmer. Just have a bit of limited time at moment .

regards
David
 
Aaaaaaaahhhh, one more thing:

In many different circumstances, with almost all of my PETs, I have had *many* problems with Commodore IC sockets. Now, when I have some kind of trouble with a C= board, first of all I replace all the original sockets with fresh ones (possibly high precision sockets).

To remove the original CBM sockets is easy: *CAREFULLY* insert a small screwdriver under the plastic housing (beware not to break any trace!) and GENTLY rise it. The socket pins will remain soldered on the board, and you can remove them once at time.

After you remove the socket, and before to put a new one, double check you didn't cut any trace.

I fixed some issues simply replacing the sockets. Simply cleaning them often doesn't solve. A replacement was the only way to fix the issue.
 
giobbi said:
In the past I had a similar problem with my 3032 (2001N/32). I'm sure Dave will remember, since he and Mike_S helped me a lot with it.
Click to expand...

Giovi,
Good to hear from you! You had every problem in the book, but never quit.

I hope wallaby doesn't run into all those problems....


I discovered that, due a broken trace, the BASIC 2.0 ROMset simply didn't work. I tried the BASIC 4.0 ROMset and the computer booted with no problem. So, after thousand tests and ROM comparisons, I found the BASIC 2.0 hex code (few bytes) who prevent the computer from booting. Somebody here in the VCF told me what to physically check, so I found the broken trace.
Click to expand...

As I remember, the 'Video ON' signal to the VIA C5-pin15 was open and the BASIC 2 code would wait forever in a tight loop for that signal to go low.
-Dave
 
dave_m said:
Giovi,
Good to hear from you! You had every problem in the book, but never quit.

I hope wallaby doesn't run into all those problems....
Click to expand...

Hi Dave! Nice to "meet" you again. Yep, I got the whole "PET issues" season: everything can be wrong in a PET, went wrong in my 3032... ^_^

dave_m said:
As I remember, the 'Video ON' signal to the VIA C5-pin15 was open and the BASIC 2 code would wait forever in a tight loop for that signal to go low.
-Dave
Click to expand...

I think you're right, it was something related with the VIA chip and a broken trace. I saved the whole thread on another computer... tomorrow I will take a look. However mine was quite a specific issue... somebody removed a chip using a big screwdriver... or a crowbar, maybe ^_^

however many issues I had with 3032 and 8032 were related with sockets. I solved many problems simply replacing the old, CBM cheap sockets with high precision ones. This left me astonished, I didn't know it would be such a problem. But it was. Ok, I'm quite a different child, everything strange happens with me, LOL, but my suggestion is to replace every socket at first. It's easy, fast, and it worth a try.

Of course the RAM/ROM board helped me a lot, and the eprom programmer saved me, in the end... for not to mention you and Mike, of course.
 
i tried to read everything here so hopefully I didn't miss it, but did you check to make sure RESET goes low momentarily upon power up? Also check to make sure that IRQ and NMI arent't held low.
 
Linear power supplies are dirt simple to fix and very reliable when they are working. I wouldn't worry about power supply problems. If there are indeed power supply problems, fix them.

But unless I missed something, everything I've read here indicates that the power supply is working fine.
 
inotarobot said:
Hi Wallaby,

where in Oz are you ? I am in Melbourne so may be able to help some, as I have desoldering station, Oscilloscope and eprom programmer. Just have a bit of limited time at moment .

regards
David
Click to expand...

I'm in Tasmania. I could probably post the board to you to look at if you're available. I had it on my "project" table for months and I couldn't figure it out. Since it's so heavy I was reluctant to take it down, but I have other computers with issues now too (Australian BBC Micros, 0 for 3!)

I looked to purchase a PetVet, but there was no "Add to Cart" button on their website.

So the condition remains the same. It powers on and displays the random characters on the screen. For whatever reason, my logic probe stopped working and the IC remover I got was about as elegant as taking an axe to it, and I started to wonder if this really was all worth the trouble.

In the meantime, I tried my hand at desoldering a battery on a rare 486 board and ended up destroying it so that put the brakes on me trying to desolder the ram in the Pet and add new sockets.

If you can help, I'm very open to outside help at this time! Haha!
 
I read through some of the posts and one thing stood out to me.
You mentioned the 5v rail measured around 4.8v and the reading was unstable. An unstable 5v rail reading that low may be why its not booting. Its a fair indication that the huge blue off board capacitor has degraded over time.
 
4.8V isn't bad, but yes, 4.8V + a lot of ripple is. At this point, we don't know how much ripple there is.

Wallaby, have you checked what I posted in #52?
 
KC9UDX said:
4.8V isn't bad, but yes, 4.8V + a lot of ripple is. At this point, we don't know how much ripple there is.

Wallaby, have you checked what I posted in #52?
Click to expand...

I'm not sure if the ripple was my multimeter or not. The last thing I want is that giant blue capacitor to explode.

I'll set it up again and see what I can find with the reset.

Where is the best place to connect a logic probe? I had it connected to the expansion port power and ground but the clips can slide and I wonder if that's what destroyed the probe.
 
I wouldn't worry too much about that exploding. It's probably drying up.

I'd expect damage to the PET before damage to your probe if something gets shorted there.

If you want to really be safe, you can connect your probe to a 6V "lantern"battery, and connect the minus lead on that battery to the PET chassis.
 
KC9UDX said:
If you want to really be safe, you can connect your probe to a 6V "lantern"battery, and connect the minus lead on that battery to the PET chassis.
Click to expand...

6V or 5V? I connected the lead to a 5V power adapter and the negative to the Pet chassis. It didn't report anything. I suspect the probe is dead. It's a BK DP-21. No idea if that is a good brand, but after some research, I'm getting a new one sent (a different brand.)
 
You must log in or register to reply here.
Back
Top