Image Map Image Map
Page 16 of 17 FirstFirst ... 6121314151617 LastLast
Results 151 to 160 of 168

Thread: CBM 8032 Resurrection Project

  1. #151
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,847

    Default

    Usual comment: Is it plugged in and is it switched on sir/madam?

    Well done that man - and all without an oscilloscope. I am proud of you !

    Dave

  2. #152

    Default

    Thanks Dave

    Coming from you that is indeed a compliment.




    -Pip

  3. #153

    Default Whack a Mole PET Faullt

    As this thread is now starting to draw to a conclusion I thought I would add this post about what I called the Whack a Mole issue.

    If you have been following from the binging, you will have seen that as we have been working to fix one fault, another random one would appear.

    At first I put this down to the normal failure curve of components that one would expect for a machine that had been stored in less than ideal conditions for decades. So for the first few faults I was not too phased, but as we got additional problems coming out of the blue when we had solved say the video RAM issue, I was starting to become a little uneasy with things.

    Throughout the project I have been observing proper anti-static handling techniques, using earthing straps and anti static mats and being careful with what footwear and clothing I was wearing. I did this even though we are dealing with standard TTL logic, as one can't be too careful.

    My concern was that even though the voltages on the various power rails were seen to be good, with the lack of a oscilloscope I was unable to check ripple levels, and things like overshoot on power up.

    But...

    During one power up, I had my logic probe next to the motherboard, well away from the monitor. I noticed when turning on the power, it glitched. That is to say it briefly showed toggling activity even though it was not probing anything. Only for a split second but it happened every time I powered up the PET.

    I then realised that it was not doing this on power up par se, there was a short delay and then it did it. The delay coincided with the CRT coming on, as the flickering LEDs on the logic probe synchronized with the slight tell tale static crackle from the CRT as the EHT spooled up. The EHT is switched on by the PET after a short delay as the CPU initalised the CRT controller chip early on in the boot up process.

    I concluded that it was not a power rail insatiability issue, rather something with the monitor itself. I had purposefully kept the CRT well away from the motherboard for safety reasons, being mindful that it contains high voltages.

    As my attention was about to focus on the monitor and it's brightness fault anyway, I thought I would just check the connections anyway.

    J7 on the motherboard was all good, as was the monitor end of the interconnect, however when I looked at the earth connection that goes between the chassis hinge and the monitor frame, I found that at the monitor end of the earthing straps eyelet was loose, very loose in fact. What's worse is that the connection is between the screw head itself and the standoff. This screws down onto the monitor PCB in one corner. The PCB is single sided and has no electrical earth connection on the top, so the connection relies solely the screw-head and its thread into the standoff. Anyhow I tightened this up firmly and then proceeded to powered up. This time the logic probe did not glitch.

    My conclusion is that we have had a poor earth on the VDU chassis.

    The interconnect between the motherboard and the monitor board does contain an earth connection itself, but I am guessing at times this was the lowest impedance path to earth for the monitor.

    It is my belief that on power up with the lack of a good ground to the monitor PCB and chassis, the VDU had been causing transients to pass down the interconnect cable and onto the main PCB. This has been getting into the power supply rails and has been damaging the ICs as we have been powering up the PET during the many tests we have conducted.

    On my version of the PET the upper case is plastic, so this ground strap is crucial for the monitor/VDU to get a good ground. If like me you have opened up your machine to service it or what ever, double check that this earth is intact and secured correctly. Otherwise the VDU could cause problems back down on the motherboard, particularly during power up/on.

    -Pip
    Last edited by Pipcicle; May 6th, 2020 at 12:48 PM.

  4. #154

    Default

    Quote Originally Posted by daver2 View Post
    You'll be starting a thread on the disk drive next...
    Dave
    Yep when this baby has been put to bed. I have ordered that new VIA chip, so I can return the borrowed one back into the 4040

    But in between I have another project in mind that involves a ZX81

    Quote Originally Posted by daver2 View Post
    I am just wondering if (when we get the CV-19 issue out of the way) there would be sufficient interest in the UK for a "Commodore Convention" - or some other aptly named get-together on a Bank Holiday? Or open it up to general Vintage Computers? VCF UK anyone? Perhaps a separate thread topic...
    Dave
    Yes who knows or at least a retro convention.
    Go for it.

  5. #155

    Default

    Quote Originally Posted by daver2 View Post
    It's actually a very bad idea Mike !

    Use an EPROM programmer for now - the checksum routine in an EPROM programmer will give you the same checksum (it is just the modulo 16-bit sum of all of the bytes)...

    But I can't fault your idea. Maybe a Version 5...

    4K ROM space verses 2K EDIT ROM. May have to checksum only half of the ROM area? Upper or lower half me wonders?

    V5 for the chirp sound that dave_m wanted...

    Also have to think about testing interrupts...

    Testing PIA and VIA devices...

    Anything else? Perhaps this ought to go on my own thread?!

    Dave
    How about a loop in the VDU test section that switches the character set between Graphics and Text mode, akin to POKE 59468,14 then delay loop for 3 or so seconds then POKE 59468, 12.

    One would be able to see and check both character sets in the character ROM. It will also demonstrate that this part of the ROM memory addressing via the VIA (pun intended) is working correctly.

    A sounder test would also be a good thing to have. It would have helped me during this restoration.

    Quote Originally Posted by MikeS View Post
    Here's a thought: why not add a checksum of $A000 to $A7FF to the PETTESTER program? You could put the $E edit ROM in there...
    I still like MikeS's idea of moving the edit ROM up into the spare socket at $A000, when putting the PETTESTER in to $E000. 1) its a good place to stow it, 2) one could see if the edit ROM is OK or not.

    During this project we suspected a faulty edit ROM, but were only able to check this with the EPROM programmer. If you haven't got one of these then you would be stuck

    Not sure how you would cater for the variations of different editors. Maybe report the checksum and the type of editor present, if not a know ROM image then state "unknown /fail".

    -Pip
    Last edited by Pipcicle; May 7th, 2020 at 06:09 AM.

  6. #156
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,847

    Default

    I've got the 'chime' on the list. I want to make sure that if the VIA is not working (or is removed - ditto for the PIAs) that this won't mess up the rest of the testing.

    I think what I will do is to checksum the entire $Axxx area - despite the fact that the EDIT EPROM is only half of it (2K). I can then check the correct EDIT ROM checksums out with VICE (I can load them all into the 'Axxx' socket I think. This may possibly give you twice the checksum value as obtained with an EPROM programmer (two images in the A socket verses 1 in the EPROM programmer). I will put both checksums in the documentation for completeness. I will just need to checkout the 2716 EPROM pinout and the 8032 Axxx socket wiring to make this strategy will work.

    I will also need to think about testing the PIAs.

    One thing that I did think about is that the only PET key that seems consistent across all of the keyboards is the RUN/STOP key. I think (when I looked) that this is on the same ROW/COLUMN on all keyboards. If it is, I can use this fact to activate the additional tests. Can someone please check out my thinking please?

    The other thing that was on the list was a VDU testcard - for setting up the monitor...

    Dave

  7. #157

    Default

    Just for completeness, and as promised, I have posted a new thread to document how I did the keyboard repair during this project.

    You can find it here: CBM8032 and other PET keyboard repair

    -Pip

  8. #158

    Default

    I must be getting OCD or something.

    When we fixed the monitor brightness issue, you may remember that I reported the the focus pot didn't do anything, although the screen in nice and sharp so not really an issue.
    I thought while we are waiting for the replacement VIA and PIA to arrive before starting on the floppy drive I would look around the focus circuitry on the VDU board and see if I could spot any problems.

    I checked:
    R253 1MΩ variable, looks OK
    R254 220kΩ got 240kΩ close enough
    R258 470kΩ got 247kΩ looked good
    R259 10kΩ got 10kΩ spot on
    R762 1.2kΩ got 1.18kΩ OK

    C252 10nF 630V still looks OK
    C761 10nF 50V unknown as I am unable to test this in circuit because of R762

    D761 showed around 0.55 Volts forward bias, but 1.6 volts in reverse. That is in circuit so not sure we can deduce that it is leaking from that?

    Looking at the voltages on Q761 we get the following readings:

    Collector = 138.6V , should be 129.7V
    Base = 17.7V =, should be 12.7V
    Emitter = 16.72V, should be 16.5V close enough

    The voltages around Q761 don't seem right, which I guess could account for the non functional focus pot.

    Any ideas?


    -Pip

  9. #159
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,847

    Default

    Hi Pip.

    So, I am a bit confused about your reading for R258 (should be 470 k).

    One side of this resistor is connected to the tube (a very open circuit) and to C252 (should be a good open circuit) - so I am not sure why you are reading 247k??? There is something not right here...

    The other thing would be to measure the resistance (PET powered OFF and capacitors allowed to discharge) from the other end of R258 (the end not connected to the tube pin 7 and C252) and both E400 and 0V with the focus potentiometer (R253) at one end of the travel and then the other.

    In one case you should get 0R (E400 end) and in the other case you should get approximately 220k (the value of R254).

    I have learnt not to take looking at something as an indication of its health... Always measure!

    Not sure about Q761 at this point - but that focus potentiometer (R253) should do something!

    Dave
    Dave

  10. #160

    Default

    Quote Originally Posted by daver2 View Post
    Hi Pip.

    So, I am a bit confused about your reading for R258 (should be 470 k).

    One side of this resistor is connected to the tube (a very open circuit) and to C252 (should be a good open circuit) - so I am not sure why you are reading 247k??? There is something not right here...

    The other thing would be to measure the resistance (PET powered OFF and capacitors allowed to discharge) from the other end of R258 (the end not connected to the tube pin 7 and C252) and both E400 and 0V with the focus potentiometer (R253) at one end of the travel and then the other.

    In one case you should get 0R (E400 end) and in the other case you should get approximately 220k (the value of R254).

    I have learnt not to take looking at something as an indication of its health... Always measure!

    Not sure about Q761 at this point - but that focus potentiometer (R253) should do something!

    Dave
    Dave
    Sorry Dave, it was a typo.
    Just looked at my notes. R258 470kΩ got 471kΩ looked good.

    My bad sorry for the confusion.

    -Pip

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •