PET fixing & 6502 pin working

[clm72]

I just bought a faulty PET CBM 3032. I starts with garbage collection screen and stays like this.

I tested voltages which are good, removed 6520 & 6522, and finally looked on the 6502 pins with
a 100mhz fluke scope, clock is ok coming in, and clock ph1 & ph2 signals look ok coming out.
Reset, NMI, IRQ, etc... look OK level, but data bus, address bus are all static around 1,83V no activity.

I so thought the 6502 may be faulty so I placed in a Commodore 1541 unit, and it looks to work normally,
unit boots up normally and checking signals on the 6502 shows normal activity on data & address bus,
so that 6502 looks oK.
Still wondering what's wrong in this pet... I checked also connection of the 6502 socket on the board and it's OK.
I will next test the ROM, reading them with eprom programmer, and next would have to test rams, but since they
are soldered on the board, will have to desolder first and place on sockets.

Now, the question is about 1,83 static signals on the DATA and Address... is that normal or sign of anything that could
help me diagnosing the problem?

[Dr Jefyll]

Welcome, clm72!

clm72 wrote:

Reset, NMI, IRQ, etc... look OK level

What about RDY (pin 2 of the 6502)? Is it high?

-- Jeff

[BigEd]

If the data lines alone were misbehaving, I'd wonder if some device is broken, or some address decoding broken, such that you have a conflict on the bus. But with the address lines misbehaving, I'd have to think of something else. I believe the PET serves video from the main RAM, and so it must have some multiplexing of address sources - either the video system or the CPU. I wonder if those multiplexors are misbehaving, or receiving bad signals.

Have you found and studied a schematic for this model of PET? It would be nice to have one big page, but the first result I find is a series of mini schematics:
http://www.zimmers.net/anonftp/pub/cbm/ ... index.html

If it could be bad RAM, then it's worth checking to see if one or more of the RAM chips is much hotter than the others. Any very hot chip might be a bad one. I read a diagnosis and repair story recently which involved replacing a few chips before the machine came to life.

[Dr Jefyll]

clm72 wrote:

I tested voltages which are good

In this case it almost seems the voltage isn't reaching the CPU -- for example, as if the socket were not making a good contact. It would be a good idea to check the Vcc and Gnd voltages as measured on the actual pins of the CPU itself.

[clm72]

Dr Jefyll wrote:

clm72 wrote:

I tested voltages which are good

In this case it almost seems the voltage isn't reaching the CPU -- for example, as if the socket were not making a good contact. It would be a good idea to check the Vcc and Gnd voltages as measured on the actual pins of the CPU itself.

I checked VCC on the 6502, it is 5V. and GND is zero.

[clm72]

BigEd wrote:

If the data lines alone were misbehaving, I'd wonder if some device is broken, or some address decoding broken, such that you have a conflict on the bus. But with the address lines misbehaving, I'd have to think of something else. I believe the PET serves video from the main RAM, and so it must have some multiplexing of address sources - either the video system or the CPU. I wonder if those multiplexors are misbehaving, or receiving bad signals.

Have you found and studied a schematic for this model of PET? It would be nice to have one big page, but the first result I find is a series of mini schematics:
http://www.zimmers.net/anonftp/pub/cbm/ ... index.html

If it could be bad RAM, then it's worth checking to see if one or more of the RAM chips is much hotter than the others. Any very hot chip might be a bad one. I read a diagnosis and repair story recently which involved replacing a few chips before the machine came to life.

I downloaded the schematics and printed them, I wouldn't work without this.
I agree failure may come from different sources, but just hope this is just one, easier and faster to find and solve.
On the Pet 2001 et 3000 series videos is served from video ram, this can be seen on the schematics and specific logic, no CRTC,
which will appear from the 4000 series from I read on the subject. This explains while the computer itself doesn't work, video works displaying the garbage collection on screen.

Good idea about checking ram temperature, I didn't think about this, will try and report here result.

If the problem comes from other than 6502, 6520, 6522 or Roms, I will have desolder components and place them on chip carriers, tricky and long job. fortunately I have a desoldering machine with air pump.
I have a Commodore 64 where all the chips have been desoldered and place on chip carriers, I got them from friend who was in military police (gendarmerie), in military school police where they had an radio and electronic section. Seems they did the job as an exercice.

[clm72]

Dr Jefyll wrote:

Welcome, clm72!

clm72 wrote:

Reset, NMI, IRQ, etc... look OK level

What about RDY (pin 2 of the 6502)? Is it high?

-- Jeff

think I did, but don't remember, will recheck, good idea.

[clm72]

Dr Jefyll wrote:

Welcome, clm72!

clm72 wrote:

Reset, NMI, IRQ, etc... look OK level

What about RDY (pin 2 of the 6502)? Is it high?

-- Jeff

checked, yes RDY is high! 5V

[Dr Jefyll]

This is very puzzling. What brand is the CPU, and is it an actual 6502 (rather than a 65C02, for example)? Also, do you see any activity on the SYNC output (pin 7) of the CPU? AFAIK it's impossible NOT to have pulses on SYNC (assuming RDY is high, and the clocks and power supply are alright). But please check anyway.

Under these conditions it also seems impossible to have no activity on the address and data buses... unless some other device is loading or driving these lines (in opposition to the CPU). To check for this you could remove the CPU from its socket and carefully bend the pin for A0 outward at a 45 degree angle (approx). Then reinsert the CPU in its socket. The idea is that the A0 pin will NOT engage the socket -- instead it just hangs in midair. Now you can power up the machine and see if there's any activity on the A0 pin. You've ensured that *nothing* connects to it, which means you're free of any possible influence from another device driving this line. Under these conditions you should see activity. (A0 is a very busy pin -- it usually toggles almost constantly. )

Check SYNC first, though. If you don't see any activity there then we need to look elsewhere. For example (and this is not the only possibility) you might have a fault in your test equipment or test procedures.

ETA:

Quote:

next would have to test rams, but since they are soldered on the board, will have to desolder first and place on sockets.

Not a good idea, at least not yet. There's a very real risk of creating new problems. Moreover there's no solid indication the RAM's are faulty. And we can continue diagnosis without desoldering the RAM's.

[Martin A]

Dr Jefyll wrote:

To check for this you could remove the CPU from its socket and carefully bend the pin for A0 outward at a 45 degree angle (approx). Then reinsert the CPU in its socket. The idea is that the A0 pin will NOT engage the socket -- instead it just hangs in midair.

And if you don't fancy bending the pins on the 6502 and you have a spare 40 pin socket, you could remove/bend the A0 pin from the spare, and put that between the CPU and the motherboard socket for the same effect.

[GaBuZoMeu]

Was the inspection of the ROM(s) helpful?
What about a NOP test (hard wire (resistors) databus to $EA) ?

Good Luck!

[barrym95838]

Faulty RESET circuit hysteresis? Can you try to manually strobe the RESET pin after full power-up and see if you get any response?

Mike B.

[BigEd]

It's a good point: the edges of the signals must be suitably fast: that applies especially to the input clock and to the reset.

[cbmeeks]

Did you ever get your PET to work? I'm cleaning my 8032 and I've ordered new capacitors for it so I will be recapping mine soon.

That 4700uF cap from the 80's is about the size of a C cell battery. It's amazing how much smaller the Panasonic I ordered is with the same specs.

Anyway, I'm curious as to the outcome of your PET.