Image Map Image Map
Page 41 of 43 FirstFirst ... 3137383940414243 LastLast
Results 401 to 410 of 424

Thread: Commodore pet 3032 screen full of lines

  1. #401

    Default

    Hi Daver2,
    i changed only 164 and l04 chips...and some ram chips...

    whit the pettester in ud8 socket, i ve always same checksum for each rom.
    Here you can see a little video that showing pettester boot:

    https://youtu.be/BKjYLLZHnzc

    20191019_155846.jpg

  2. #402

    Default

    I tested rams for many passes..... 0 to 9 and afer again...

  3. #403
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,155

    Default

    >>> whit the pettester in ud8 socket, i ve always same checksum for each rom.

    Yes, but is the value for each ROM consistent?

    >>> I tested rams for many passes..... 0 to 9 and afer again...

    But for how long - how many passes? It counts the passes - how many have you run in the session?

    Dave

  4. #404

    Default

    Quote Originally Posted by daver2 View Post
    >>> whit the pettester in ud8 socket, i ve always same checksum for each rom.

    Yes, but is the value for each ROM consistent?

    >>> I tested rams for many passes..... 0 to 9 and afer again...

    But for how long - how many passes? It counts the passes - how many have you run in the session?

    Dave
    Yes always same values:
    b=7800 c=3038 d=506a f=7c98

    can i try to run pettester again for 1 hour....

  5. #405
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,155

    Default

    We can also let the PET start-up (with - say - 264 byte free) and try and enter the machine code monitor (TIM) and have a look at the memory around where the PET thinks there is a memory fault. We may be able to work out from that what is going on.

    But, before we get there, we need to know that the ROMS, RAMS and KEYBOARD pass my diagnostics.

    Dave

  6. #406
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,155

    Default

    >>> can i try to run pettester again for 1 hour....

    Yes please, the longer the better... Also, close the PET up so that the temperature has a chance to build up within the case. Temperature is your enemy!

    Dave

  7. #407

    Default

    Quote Originally Posted by daver2 View Post
    >>> can i try to run pettester again for 1 hour....

    Yes please, the longer the better... Also, close the PET up so that the temperature has a chance to build up within the case. Temperature is your enemy!

    Dave
    Daver2 now i have error on the first pass!
    4000 10 ef
    20191019_165658.jpg

  8. #408
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,155

    Default

    Test 0 is where the firmware fills all of memory with zeros and checks that zeros are there after the fill (i.e. mainly checking for a stuck '1'). This test is passing OK.

    Test 1 is where the firmware fills all of memory with ones and checks that ones are there after the fill (i.e. mainly checking for a stuck '0'). This is the test that is failing.

    The address is 4000 (the first byte of bank 1 of the PET memory).

    The bit mask in error is '10'.

    The data byte found in the memory was 'ef'.

    Checking:

    Data value . EF = 1110 1111.
    Bitmask .... 10 = 0001 0000.

    It is indeed true that the addressed bit (data bit 4) is zero when it should be a '1'.

    Does this mean that Bank 1 data bit 0 (UJ5) is faulty? It could be. But it could also be that the databus buffer for this data bit (UI11) could also be faulty. It could also be that the address multiplexers (UE3 to UE6) are also faulty so that the multiplexers addressed the incorrect memory cell within the RAM chips. It could also be that the timing circuitry is still faulty in some way.

    We have to identify which of the scenarios it could be before we start swapping things over - otherwise we will get completely and utterly confused.

    My suggestion is to get a notebook (or some paper) and repeat the memory tests about 20 times noting down each error as it occurs (the bit after the 'mem fail' text) and we can then see if anything consistent is revealing itself.

    RESIST THE URGE TO CHANGE ANYTHING AT ALL AT THIS STAGE.

    Dave

  9. #409

    Default

    Great Daver2!
    I checked Ram ic in UJ5 and i saw that ic ram has one folded contact leg....
    i fixed it and now i try to run test!

  10. #410
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,155

    Default

    OK !

    Dave

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
  •