Image Map Image Map
Page 1 of 17 1234511 ... LastLast
Results 1 to 10 of 167

Thread: CBM 8032 Resurrection Project

  1. #1

    Default CBM 8032 Resurrection Project

    Hi
    As we have a bit of spare time on our hands at the moment, I have decided to dig out my CBM 8032 that was donated to me years ago and attempt to get it up and running again. I have seen a thread about some other people fault finding on the 8032 so I am hoping I can tap into people’s expertise and experience in this forum.

    Anyhow to give a brief update as to where I have got up to:
    • The machine has been completely disassembled, carefully cleaned and reassembled
    • All socketed chips have been removed, legs cleaned and re-seated in their sockets
    • First power up no bang, no escaping magic smoke!
    • Power rails seem to be in range. Not able to check ripple as due to lockdown I don't have to access a scope.
    • The display was working to begin with, but displaying garbage.
    • Checked the 4 x 2114 memory chips and found that 3 out of the 4 were bad.
    • During the replacement of the memory chips the display suddenly went blank

    I am now trouble shooting the display part of the circuitry and I suspect that the 6545 CRT controller has passed away.
    I can see the CPU is doing something as the sync, data and address lines are all toggling, as well as the video memory.
    The clock signal to the controller also seems to be good.

    Looking more closely at the 6545, it seems to be putting out an HSYNC an signal on pin 39, however the VSYNC on pin 40 is permanently low.

    Before I source a new chip and replace this current one, can anyone suggest further tests I can perform on the chip and circuit around it, before declaring it dead.

    Cheers
    Pip

  2. #2
    Join Date
    Feb 2009
    Location
    Southern California, USA
    Posts
    2,908

    Default

    The 6545 is initialized by the CPU once right after power-up. That should start the HSYNC and VSYNC. The 6545 would need a good /RESET so check that the signal is low for 1/2 sec minimum at power-up. I would want to check for proper periodic activity on the data and address lines, but without a scope that may be difficult.

  3. #3
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,790

    Default

    Welcome to VCFED Pip.

    Another 8032 resurrection - Yeh!

    My first question would be: without an oscilloscope - what test equipment are you using to look for HSYNC and VSYNC and what test equipment have you access to?

    Hi dave_m. How are you keeping?

    Dave

  4. #4

    Default

    Hi everyone

    First of all, I have to say that this forum is awesome. I only posted the thread a few minutes ago. Thank you

    The 6545 or 6845 I was referencing from the schematics is actually an HD46505SP on the board. Not sure if that means that it has already been replaced at some point. One of the 2114s has been replaced in the past as it was socketed. The other 3 bad ones were the originals, as there was no evidence of re-soldering.

    As far as test equipment at hand I have.
    DVM
    Logic probe
    Frequency counter
    I looked on the HSYNC and VSYNC using the logic probe. I also lifted pin 40 to check that it was not being held down elsewhere, and it still was showing low.
    I am aware that the reset to the chip is a short burst of pulses at power up, however I have no way of checking this at the moment, the logic probe did not show this, in any case the probe gets it supply from board so is powering up at the same time as the board. If that makes sense.

    Pip

  5. #5

    Default

    Quote Originally Posted by dave_m View Post
    The 6545 is initialized by the CPU once right after power-up. That should start the HSYNC and VSYNC. The 6545 would need a good /RESET so check that the signal is low for 1/2 sec minimum at power-up. I would want to check for proper periodic activity on the data and address lines, but without a scope that may be difficult.
    Hi Dave
    Yes the reset signal looks okay. It did hold low for a time after power up. I also forced a manual reset by pulling pin7 low on UD16 (555 timer).

    In both cases there was no activity on pin 40 of the CRT controller, or at least none that my logic probe would pick up.

    Pip

  6. #6
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,790

    Default

    >>> First of all, I have to say that this forum is awesome. I only posted the thread a few minutes ago.

    We try!

    A frequency counter will be very useful if you need to build a NOP generator...

    I just connect my analogue multimeter between 0V and the /RESET pin of the CPU and power up the PET. The value should stay low for a short period after power-up and then go high.

    You could have a duff CRTC - but the data writes to the CRTC could also be faulty - or the data bytes initialising the CRTC (stored in the EDIT ROM) could also have become corrupted. Just thinking of various scenarios to check.

    Two useful things in your arsenal to fault find are:

    A NOP generator. This is a couple of 40 pin sockets sitting between the CPU and the processor socket 'forcing' NOP instructions ($EA) onto the data bus of the CPU.
    My PETTEST code burnt into a 2716 EPROM to replace the EDIT ROM. I will have to provide a link to this later.

    Dave

  7. #7

    Default

    Quote Originally Posted by daver2 View Post
    >>> First of all, I have to say that this forum is awesome. I only posted the thread a few minutes ago.

    We try!

    A frequency counter will be very useful if you need to build a NOP generator...

    I just connect my analogue multimeter between 0V and the /RESET pin of the CPU and power up the PET. The value should stay low for a short period after power-up and then go high.

    You could have a duff CRTC - but the data writes to the CRTC could also be faulty - or the data bytes initialising the CRTC (stored in the EDIT ROM) could also have become corrupted. Just thinking of various scenarios to check.

    Two useful things in your arsenal to fault find are:

    A NOP generator. This is a couple of 40 pin sockets sitting between the CPU and the processor socket 'forcing' NOP instructions ($EA) onto the data bus of the CPU.
    My PETTEST code burnt into a 2716 EPROM to replace the EDIT ROM. I will have to provide a link to this later.

    Dave
    I did check the reset and did a manual reset, so I am happy that this is okay.
    Even if the CRTC was getting duff data, surely I would still be getting something on the VSYNC output?

    Pip

  8. #8
    Join Date
    Feb 2009
    Location
    Southern California, USA
    Posts
    2,908

    Default

    Quote Originally Posted by Pipcicle View Post
    I did check the reset and did a manual reset, so I am happy that this is okay.
    Even if the CRTC was getting duff data, surely I would still be getting something on the VSYNC output?
    Pip,
    If it happens to be a bad address line, the correct CRTC register may not be loaded. You say the CRTC may have been replaced before. Is it in a socket? That would make things easier. There are safe ways to remove the 40 pin part that will not overheat the PCB although it will sacrifice the suspected part.
    -dave_m

  9. #9

    Default

    Quote Originally Posted by dave_m View Post
    Pip,
    If it happens to be a bad address line, the correct CRTC register may not be loaded. You say the CRTC may have been replaced before. Is it in a socket? That would make things easier. There are safe ways to remove the 40 pin part that will not overheat the PCB although it will sacrifice the suspected part.
    -dave_m
    Luckily all the 40 pin chips and ROMs are socketed, and now the 2114s. I have looked at all the address and data lines and they are all toggling. Previously one of the faulty 2114s IO (0) was indeterminate, the other two faulty chips IOs (0-3) were all stuck low.

  10. #10
    Join Date
    Feb 2009
    Location
    Southern California, USA
    Posts
    2,908

    Default

    You say you did not detect any pulses on the CRTC chip select (UB13-pin25)? You can keep the logic probe or counter (set to TTL) on this line while you quickly ground the RESET line at UB13-pin2, or at the expansion connector J4-pin22 to the pin opposite (all grounds). A wire wrap jumper is good for this.

    There will be about 34 pulses on the line to select 17 register addresses and write 17 bytes of data. If you catch a pulse, then replace the 6545.

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
  •