Image Map Image Map
Page 1 of 2 12 LastLast
Results 1 to 10 of 18

Thread: PET 4016 with display issues

  1. #1
    Join Date
    Sep 2019
    Location
    Hornslet, Denmark
    Posts
    26

    Default PET 4016 with display issues

    Hi

    I am in the process of fixing a CBM 4016 (PAL) that I recently bought as defective, and I need a little help on how to proceed from where I am now

    Machine has been stripped and cleaned, visually inspected and while some legs on especially the TTL-chips were corroded, no visible damage has been found.

    Board has at some time been populated with another 16K of RAM, making it a 4032.

    The machine was described as no picture, but sound on boot. And yes, that was correct. It chirped away on every boot

    After unpluggin the CRT, I measured a lot of voltages on board, the AC block is fine, pin 9-11 was ~20.5V AC, there was -5V, 12V and 5V on the three voltage regulators V1 - V3. There is a steady 5V on the 6502 as well on both datasette ports.

    J7 (video) on the board was looking a bit weird with low voltages on V- and H-SYNC, so I decided to swap the CRTC with a new one, having read that this has a tendency to go bad.

    Having done this (an checked the CRT board visually), I reconnected the CRT.

    While all of the above went well, it really made no difference. Screen was black, but the neck lit up (and I think it did that from the beginning). I couldn't hear any hunning from the flyback and felt no high voltage on the front of the CRT. I have replaced the 6502 with a known working and socketed it while I was at it.

    Well, searching the web for clues, I read a few threads, where a black screen with no high voltage could be caused by a bad resistor, which may have been used as a kind of fuse. Threads mentioned R752 and R753 on the CRT PCB. And yes, one of them showed an OL, while the other was ~63 Ohm. Schematics says 56 Ohm.

    Swapping both of the resistors with new ones, I got high voltage and a picture

    From the beginning I knew that i had to get help from the outside, so I cashed out and bought this; https://www.tindie.com/products/tyne...commodore-pet/ which should replace both ROM and DRAM on the board. When I get the PET running, I will fix onboard RAM/ROM issues, but not now.

    I'll attach a few not-so-good-pictures of the screen, but I think you can make out that the display shows something on the screen, but something else where it should have said

    ### commodore basic 4.0 ###
    31743 bytes free
    ready.

    When I enter text and/or spaces, the screen changes where the cursor is/or should be.

    The speaker chirps when I get to column 40.

    So, the machine is definately running and accepting input.

    When I pull the editor ROM, the screen shows all solid green in a nice square box.

    And now the question; does all of this point to the 2114 SRAM? Or some glue logic close by? Or what should I do from here? I don't feel like just swapping bits and pieces, I would much rather find the culprit and work it out with you guys

    I'll answer any question as best as I can!
    Attached Images Attached Images

  2. #2

    Default

    If the video RAMs are on sockets, you can create your own characters with jumper wires to ground and maybe some pullup resistors ( 10K are fine ).
    That will help you separate what side the problem is on. It looks s little to me like the character ROM but that is an all out guess. It could be a address mux always on the data bus as well. You need to separate where it might be.
    Dwight

  3. #3
    Join Date
    Jun 2012
    Location
    UK - Worcester
    Posts
    3,128

    Default

    I agree with Dwight here. You may have more than one fault present simultaneously.

    I first thought character generator, but that wouldn't (on its own) resolve the fault you are seeing.

    It also looks that you have a problem in the master timing circuit somewhere (at a guess).

    Where are you based (can you add your rough location to your profile). You may have some VCFED members nearby.

    By the way (as this is your first post) welcome to VCFED!

    Do you have an oscilloscope? Can you measure the H and V synchronising frequency on the video connector of the PET main logic board please.

    Dave

  4. #4

    Default

    This is more likely a bad CHARGEN ROM.
    In one of my videos: https://www.youtube.com/watch?v=PTKSLsCMRos
    there's a little explanation on how the characters images are formed on the PET (and many other similar designs).

    Frank IZ8DWF

  5. #5

    Default

    Quote Originally Posted by iz8dwf View Post
    This is more likely a bad CHARGEN ROM.
    In one of my videos: https://www.youtube.com/watch?v=PTKSLsCMRos
    there's a little explanation on how the characters images are formed on the PET (and many other similar designs).

    Frank IZ8DWF
    Wow, the guy has a HP current probe. Not every one is that lucky.
    Dwight

  6. #6

    Default

    Quote Originally Posted by Dwight Elvey View Post
    Wow, the guy has a HP current probe. Not every one is that lucky.
    Dwight
    The one he had is:

    http://kripton2035.free.fr/Resources...r%20Manual.pdf

    HP also made another one that used contacts to measure current through a pin. I recall using one of those as well.
    The HP547a uses an inductive coil pickup. As you can see he used this to detect that the pin was conducting excess current. Being a coil type, the current must be pulsing. These probes are rare and those that have them expect to take them to the grave.
    Dwight

  7. #7
    Join Date
    Sep 2019
    Location
    Hornslet, Denmark
    Posts
    26

    Default

    Thanks for the input - I will be looking into the suggestions very soon, but I am away for the weekend. The SRAM is not socketed, but I think I will do just that, to be able to test them. I have a spare 2114 and a couple in the mail on it's way to me

    Will get back to you!

  8. #8

    Default

    Quote Originally Posted by pilehave View Post
    Thanks for the input - I will be looking into the suggestions very soon, but I am away for the weekend. The SRAM is not socketed, but I think I will do just that, to be able to test them. I have a spare 2114 and a couple in the mail on it's way to me

    Will get back to you!
    Again: those images are not looking like bad video RAM, I would first try burning a new character generator ROM (or checking the original one on a programmer) and if it looks ok, probably try to substitute the parallel to serial shift register.
    Do you have any test equipment? An oscilloscope can really help diagnosing these faults.
    Good luck anyway

    Frank IZ8DWF

  9. #9

    Default

    Once the video RAMs are off, one can easily check with a few scraps of wire. I agree, it looks like the video ROM but it also looks a little like the processors instruction data being displayed as characters. This puts it on the over side of the video RAM. I don't think the video RAM itself is the issue.
    Dwight

  10. #10

    Default

    Quote Originally Posted by Dwight Elvey View Post
    Once the video RAMs are off, one can easily check with a few scraps of wire. I agree, it looks like the video ROM but it also looks a little like the processors instruction data being displayed as characters. This puts it on the over side of the video RAM. I don't think the video RAM itself is the issue.
    Dwight
    Ok, the pictures are not really of enough good quality, looking back again at them. Out of focus/overexposed. These might actually be legitimate graphic characters, that would imply that the VRAM(s) might be involved in the fault, but I would first check the 74LS373 (UB3) in that case.

    Frank

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
  •