Page 6 of 6 FirstFirst ... 23456
Results 51 to 53 of 53

Thread: Add another PET to the emergency room

  1. #51

    Default

    Thanks for writing it all up.

    It's been bothering me a little how the second board's problem was the borked A15 address line given the PETTESTER ROM ran, but... it occurred to me why that's the case. (I double-checked my code to make sure.) The non-intuitive thing you have to remember is when a line is broken in these things it usually causes that line to be considered as "stuck on" downstream, not "stuck off".

    (If the reverse was true the PET wouldn't of been able to see its ROMs at all, as the entire upper 32k of memory space would of been cut off.)

    What was actually happening was that any access to the lower 32k was being offset upwards 32k. So when you had the PETTESTER ROM plugged in and it ran the memory test it was *actually* testing the video RAM. It wouldn't of shown as bad because the test pattern is written in one loop, while another loop does the test read followed by the status write to the screen. That would of course immediately clobber the test value, but the program wouldn't care. And of course the video character set display wouldn't care either.

    With the BASIC ROMs plugged in you'd end up with this bizarre situation where VRAM would essentially be mapped as the CPU's zero page and stack. I'm sure what was going on when it crashed was it set up a few vital things, called the "clear screen" routine, and then wet its pants shortly thereafter as it blew away all its low memory variables. :^)

    PETTESTER definitely needs an extended-test mode where it does things to look for stuck address lines.

  2. #52

    Default

    Quote Originally Posted by tezza View Post
    And here is Part 2.

    The PET has been exercised and is working just fine with both boards. Thanks again for your help and support guys.

    Tez
    Nice repair job Tez!Thanks for sharing your experience with us!

  3. #53

    Default

    Quote Originally Posted by Eudimorphodon View Post
    Thanks for writing it all up.
    Quote Originally Posted by angel_grig View Post
    Nice repair job Tez!Thanks for sharing your experience with us!
    Thanks for that. Knowing that people actually get something out of these things gives me motivation to write them up.

    Quote Originally Posted by Eudimorphodon View Post
    What was actually happening was that any access to the lower 32k was being offset upwards 32k. So when you had the PETTESTER ROM plugged in and it ran the memory test it was *actually* testing the video RAM. It wouldn't of shown as bad because the test pattern is written in one loop, while another loop does the test read followed by the status write to the screen. That would of course immediately clobber the test value, but the program wouldn't care. And of course the video character set display wouldn't care either.
    Ahh, yes I see what you mean. I'll alter the article to reflect that (or maybe just point to your post for those that want an explanation).

    If the PET TESTER was extended it could indeed be a very powerful tool.

    Hopefully you'll be able to make progress on that 4032.

    Tez
    Last edited by tezza; July 17th, 2011 at 03:32 PM.
    ------------------------------------------------
    My vintage collection: http://classic-computers.org.nz/collection/index.htm
    My vintage activities blog: http://www.classic-computers.org.nz/blog/
    Twitter: @classiccomputNZ

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
  •