Image Map Image Map
Page 1 of 3 123 LastLast
Results 1 to 10 of 30

Thread: Commodore Pet 4016 won't read memory

  1. #1

    Default Commodore Pet 4016 won't read memory

    I have a Pet 4016 and when it boots up it only says 12 bytes free... what can be causing that? It works fine otherwise, but I can't load any programs or program anything in basic because of it.

  2. #2

    Default

    It's likely that you have some bad memory IC's. You can start by turning it on and then checking the RAM IC's
    temperature with a non contact temperature probe. If any are really hot, they need to be replaced. A few minutes
    with power on should allow them to get past operating temperature, if they are defective.

    I fixed lots of C-64's that way, but I had a better temperature probe.

    Larry

  3. #3

    Default

    I took the machine apart today and I think I found the problem. The board is the 8032090, and it has the extra sockets installed for more ram. Also the jumper is set to 32k, but it never read the additional memory. I think way back when I installed the chips someone on a forum told it it was the chip at UD5 so I tried to remove it, then changed my mind, and tried to resolder it. I bridged 2 pins together. Or are they supposed to be bridged. If they arent whats the best way to get the thin layer of solder I left behind.

  4. #4
    Join Date
    Feb 2009
    Location
    Southern California, USA
    Posts
    2,326

    Default

    I agree with Larry. It may be a bad memory chip (4116). Let us know the type of 4016 board you have. How many 40 pin chips? 4 or 5?

    After that, we will point you to the right schematic and we can try to run the Monitor program (in ROM) and display a memory block from $0400 and perhaps find a bad test pattern. As a power up test, the PET writes an alternating 1/0 pattern ($55) starting at location $401 and it that passes, it writes the other alternating 1/0 pattern $AA. I suspect a problem around memory location $040D or location 1036 decimal (1024 +12).

  5. #5

    Default

    Leave it on for an hour (but be around it), and do check for heat once in a while.

    Then cycle the power and see if you don't get more (or less) RAM. 4116s that go bad aren't very repeatable in my experience. If it's always 12 BYTES FREE, then in my opinion you may have bad RAM as well as one or more open address lines, be it broken traces or poor contact in sockets. (Or like Dave says, one bad location in just the right spot).

    Of course, you may have bad zero page RAM or a ROM glitch causing the RAM test to not execute properly, too.
    Be polite and I may let you live.

  6. #6

    Default

    Quote Originally Posted by ldkraemer View Post
    It's likely that you have some bad memory IC's.
    As Larry says: "likely". I had a C64 with 18 bytes free. It turned out that one of the address multiplexers, 74LS257, was bad. The CBM computers use 74LS244s as multiplexers: a circuit activates the needed 244 at the right moment to output an address towards the DRAMs. If one of those 244s is not functioning properly (or the circuit), you will get errors as well.
    With kind regards / met vriendelijke groet, Ruud Baltissen
    www.Baltissen.org

  7. #7

    Default

    Quote Originally Posted by RuudB View Post
    As Larry says: "likely". I had a C64 with 18 bytes free. It turned out that one of the address multiplexers, 74LS257, was bad. The CBM computers use 74LS244s as multiplexers: a circuit activates the needed 244 at the right moment to output an address towards the DRAMs. If one of those 244s is not functioning properly (or the circuit), you will get errors as well.
    Well, quite right. On PETs the actual RAM (Row/Col) address multiplexers are 74153 chips (4 of them). 74LS244 are used (correctly) as bidirectional buffers, on DATA lines and on Address lines going for example to the expansion connector.
    Inputs on the 74153 are mixed, some unbuffered addresses from CPU and some buffered ones (that's a bit messy, but probably has been done mostly for routing purposes).
    Speaking of the 2001N variant anyway.

    Frank IZ8DWF

  8. #8

    Default

    Quote Originally Posted by iz8dwf View Post
    Well, quite right. On PETs the actual RAM (Row/Col) address multiplexers are 74153 chips (4 of them). 74LS244 are used (correctly) as bidirectional buffers, on DATA lines and on Address lines going for example to the expansion connector.
    Inputs on the 74153 are mixed, some unbuffered addresses from CPU and some buffered ones (that's a bit messy, but probably has been done mostly for routing purposes).
    Speaking of the 2001N variant anyway.

    Frank IZ8DWF
    I just orded new 4116 3N chips from Bolivia off ebay and a brand new 74LS244 I will try to solder them in next week when they come. Whats the best way to unsolder a chip without wrecking it, just so I dont ruin possibly good chips.

  9. #9

    Default

    Quote Originally Posted by Pet4016 View Post
    I just orded new 4116 3N chips from Bolivia off ebay and a brand new 74LS244 I will try to solder them in next week when they come. Whats the best way to unsolder a chip without wrecking it, just so I dont ruin possibly good chips.

    The IC I was doubting is located at UD5 and shows DM74LS00N on the face of the chip. I ordered a replacement on eBay which just showed up today but the chip is SN74LS00N are they compatible? I have 2 other chips with the same label on the board, but I know that doesn't mean they're nessicarily the same. I was going to try replacing it first before I started unsoldering the 4116 ram. Those should be coming soon.

  10. #10
    Join Date
    Oct 2014
    Location
    Pottstown, PA
    Posts
    203

    Default

    I just posted an isolation technique in your earlier post.

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
  •