• Please review our updated Terms and Rules here

Intermittent CBM 8032 Power on Issue

daver2 said:
There is, but the rub is it is wrong!


Pin 4 is missing to the positive side of the capacitor.

We have already confirmed that the two (2) wires are connected to J8.

Dave
Click to expand...
The J8 connector and soldering needs to be checked and the AC inputs and DC outputs of the rectifiers.

I think they just forgot to label the capacitor + terminal going to Pin 2 and Pin 4 as shown with the pink lines on post #60

I guess it is also possible that some PET variants may have pin 2 and 4 connected on the PCB, but I am not sure about that.

Either way its hard to make any sense of it without the complete diagram.
 
Last edited:
There is one thing about this sort of circuit that can present problems for repairers and give odd results. Although the main rectifier assembly (top one in the schematic) is drawn as a bridge rectifier, it does not operate as one, it is operating as two diode pairs, because of the transformer's centre tap. If the center tap, for any reason loses its connection to ground or that goes high resistance via a defective connector or defective soldering, it can cause unexpected and confusing voltage readings.
 
Last edited:
Can I make a suggestion? Make a temporary bypass connection from the +ve end of the +9V rectifiers to the +Ve of the offboard capacitor. Leave all existing wiring in place. The aim is to bypass the JP8 socket / pin connection #4.
 
With J8 connected providing power to the board I see the expected VAC across the transformer secondary outputs.
VR1 -5.1 VDC
CR5 20/12 VDC
C10 2 VDC
C11 .5 VDC
 
The VR1 and CR5 readings are exactly what I would have expected.

The fact that the voltage across C10 is 2V will mean that the +5V voltage regulator (VR4) is starved of voltage and this will result in the voltage across C11 to be very low.

Now, the question as to why...

This is nothing to do with an increased current flow through VR4. If this was to occur, the voltage regulator would shutdown. This would not affect the voltage across C10. That should rule out the PET logic itself.

VR4 could be faulty - shunting all of the current from the input pin to ground and dragging the input voltage down. I would have expected something to 'smoke' in this case (or the diodes forming CR2 to get hot). Is CR2 getting hot? Remember, hot = burns...

I would be inclined to measure the DC voltage at pin 1 of CR2 (or at the PCB forming the connection between CR2/1, C3, C4 and J8/4).

Just checking - you have nothing plugged into J10, J11 or the cassette ports have you?

I would have to assume there is a poor connection somewhere before assuming a short circuit. Have you tried doing what Sirius suggested in post #64?

Dave
 
With regards to post #63, test with the meter between the center tap of the transformer (at the transformer) and ground on the pcb, there should be near zero detectable voltage if that connection is good.
 
Well don't I feel stupid :rolleyes:
I have examined the board several times for any anomalies but missed it completely. I found a bad solder joint under J8, bet you can guess the pin.
So I have repaired the pin 4 solder joint and have PETESTER running now. The voltages check out good now. I feel like an apology is due to everyone for my poor diagnostic effort.
I will let the board run for a while.
 
One takeaway for me is to use a higher magnification!

Do CR2 and CR4 normally run hot? Not too hot to touch but quite warm.
 
>>> Well don't I feel stupid.

Don't worry, we've all been there at one time or another!

I thought there was something around there that was wrong! We just had to find it...

Yes, CR2 and CR4 do run warm/hot to the touch. If you can't touch them that is when they are too hot...

So how far is PETTESTER getting now?

You should get good checksums for the ROMs, be able to test the keyboard, and get about 30 good test passes of the DRAM test. Put a film on and get a large glass of wine for this stage...

Dave
 
So I have repaired the pin 4 solder joint and have PETESTER running now.
Click to expand...

I'm glad you got to the bottom of it, but honestly, the next time a PET presents with +9V at the rectifiers and nearly nothing on the offboard capacitor +Ve, go straight for the JP8 / pin / socket connection #4. It is a common / recurring problem.
 
Dave, thanks for your help troubleshooting thisi ssue. I have left PETTESTER running through hex 30 cycles and will now put things back together, adding the 6809 board to see where things stand.
 
SiriusHardware said:
I'm glad you got to the bottom of it, but honestly, the next time a PET presents with +9V at the rectifiers and nearly nothing on the offboard capacitor +Ve, go straight for the JP8 / pin / socket connection #4. It is a common / recurring problem.
Click to expand...
It makes sense too, because the capacitor's charging currents on alternate half cycles of peaks of conduction from the diodes are much higher than the average current consumption and the current via each diode separately, so the connections/plug etc for pin 4 probably were a little inadequate from the start, explaining the high failure rate.
 
>>> Dave, thanks for your help troubleshooting this issue.

No problem. I just had to keep your nose to the grindstone :)!

A bad connection should appear as a resistance. So checking the resistance from one point to another should give you a clue. However, once a current is drawn, there should be a voltage that builds up across the bad connection. You then use a differential voltage measurement technique to find where the fault is. Not only are PCB solder connections a problem - but also cables. Most of my simple 'fixes' are bad connections - either in the solder joints of the power connectors or in interconnecting cables.

Get a few repairs under your belt and you will start to build up some useful experience. However (a word of warning) once you have repaired a few machines, you may start to get complacent and concentrate on things that you now know fail. This can lead you astray. You really should be like a detective and follow the clues based upon a test methodology.

Also, document each step in a notebook. You will need this later when you get lost and need to backtrack. Also, once you have found the fault(s) you can re-read your notebook to see how you got there - or more likely, where you got distracted in your testing.

On a small number of occasions I have found in my notebook something that couldn't be accounted for by the fault I had found. Unknown to me, the first fault I was chasing when I started out mysteriously fixed itself and was replaced by another fault! I had found (and fixed) the second one... I hunted a bit more and was happy to find (and fix) the original fault in the end. It would have only come back to haunt me in the long run!

Dave
 
Just a cautionary note, I had one case of this where, when measured with the power off, there seemed little or no resistance between the rectifier +VE output and the offboard capacitor +VE, but when powered on there was a huge voltage difference between the rectifier +VE and the offboard capacitor +VE. This was obviously an impossible combination of measurements, with no resistance between the two points there can not be a large voltage difference between them.

The question was only resolved when the J8 pin 4 connection was bypassed altogether by connecting a bypass wire directly between the rectifier +VE and the capacitor +VE, which brought the machine back to life. That was why I suggested it as a diagnostic step earlier.

Although in miraco's case it was a cracked solder joint on the J8 pin 4 connection to the PCB tracks, often it is a failure of the connection between the pin and socket but it doesn't necessarily just fail high-resistance in an obvious way, it can fail in an odd non-linear fashion which only manifests itself when you try to pass a large current through the connection.

In the case mentioned, thorough cleaning of the JP8 / #4 pin and corresponding female socket contact in the plug sorted the problem out.

My theory about why it is always that particular pin / socket connection is that at initial power-on, the inrush current into that huge offboard capacitor through the J8 / 4 connection must be enormous.
 
I was preparing to solder a bypass wire when I decided to check all J8 solder points again with a stronger magnification. I have an idea as to how this failure point occurred in this case. About 12 years ago I had a short on the 6809 board that ran down the J8 power strap on pin 4 to the 8032 board. I replaced a capacitor on the 6809 board and was able to get the system back up and running, with an apparently undamaged 8032. I believe the solder joint was weakened then but not enough to cause an issue until now. Another lesson learned is to be more thorough in researching the after effects of a failure.

I really appreciate the help from all of you. Your knowledge and experience are off the charts.

Now to get the rest of this SuperPet working …
 
I am envious of your SuperPET board by the way. I keep looking for one for my 8032...

Do you have the two or three board SuperPET?

Are you using my 'de-dongled' Waterloo SuperPET software disk? Just in case your 6702 dies...

I was involved in the 'de-dongling' of the software and reverse-engineering of the 6702 for the SuperPET implementation for VICE (although my name doesn't appear in the list of contributors for some reason).

Dave
 
I have the “three board” configuration SuperPet. Yes, I have downloaded the patched Waterloo disks from Mike Nebrezney’s site. I learned about your efforts to crack the 6702 while working with him on some document and software archiving work a few years ago. That was some impressive work you did there!
 
You must log in or register to reply here.
Back
Top