Image Map Image Map
Page 3 of 3 FirstFirst 123
Results 21 to 29 of 29

Thread: Broken Cromemco 8k Bytesaver

  1. #21

    Default

    It is likely a schematic error, as with the pin numbering. I agree a power feed to the tap makes sense. The small turns would then do what is expected to oscillate, as feed back.
    Dwight

  2. #22

    Default

    Quote Originally Posted by Dwight Elvey View Post
    It is likely a schematic error, as with the pin numbering. I agree a power feed to the tap makes sense. The small turns would then do what is expected to oscillate, as feed back.
    Dwight
    Yes I agree.

    The fact on that schematic, that they drew what should be the feedback winding as shorter than the primary, does suggest the feedback voltage would be lower than the collector voltage, which would be required to avoid zenering the transistor's B-E junction. So maybe, as a guess, if the primary winding was assigned a value of 1 for turns ratio purposes, calculating the number of turns, the feedback winding would be assigned 0.25 and the the secondary maybe about 3. Then if the off load peak voltage was say 20V on the primary, it would be about 5V on the feedback winding and about 60V on the secondary winding. Then on load, with the rectifier and capacitor filter damping it on peaks and the feedback control operating, it would assume around 30V.

    So with the limited data we have so far, what to do from the practical perspective ?

    I would forget the Spice simulation.

    I would buy an inductor of the radial lead type with a ferrite core where you can unwind the winding. With an approximate physical size of the original inductor see on the photos of the board.

    Unwind it and discard the wire. Wind a coil (primary winding) with enough turns that it measures about 5 to 10mH on an inductance meter. Use a wire size that fills about 1/3 of the available space.

    Whatever those turns are use about 1/4 that number for the feedback winding and about x 3 that number for the secondary winding.

    Then I think you will be off to a good start, replicating the original transformer.

    Something like this where you can scavenge the ferrite core for a rewind:

    https://au.rs-online.com/web/p/leade...ctors/4890059/
    Last edited by Hugo Holden; October 31st, 2020 at 04:30 AM.

  3. #23

    Default

    Input filter cores are a good source for such toroidal cores. The wires you use can be small as the current needed for EPROM programming is generally quite small.
    Dwight

  4. #24
    Join Date
    Jul 2014
    Location
    Rhode Island
    Posts
    1,375

    Default

    I bought a bag of 25 MPS6560 transistors on eBay. The bad Q10 from the Bytesaver had reasonable diode drops base-collector and base-emitter, but had a Hfe of 5. The first new transistor that I tested was bad, the rest had betas that clustered around 50, 100, 150, and 200. I installed a transistor with a beta of 150, and the 30V supply works. I just programmed a 2708 using the CP/M program Bytesave, and it worked! Now we will program the other 7 2708 EPROMs with the rest of 8k BASIC.

    The duty cycle of Q10 is a lot longer than I expected to see.

    Q10 Base: Q10-Base.png
    Q10 Collector: Q10-Collector.png
    D1: Cathode: D1-Cathode.png

    We noticed that the 30V power supply frequency changed from 231kHz at idle to 313kHz when programming the EPROM.
    Q10 Base and Collector when programming.png
    Last edited by m_thompson; October 31st, 2020 at 11:53 AM.
    Member of the Rhode Island Computer Museum
    http://www.ricomputermuseum.org

  5. #25

    Default

    Ok, don't leave us dangling, what did you do for the coil? Your waveforms look good.

    Dwight

  6. #26

    Default

    Quote Originally Posted by Dwight Elvey View Post
    Ok, don't leave us dangling, what did you do for the coil? Your waveforms look good.

    Dwight
    I think he already had the transformer, it was 8008guy who was missing one, but I'm not sure.

    If m-thompson could measure the voltages on the transformer pins it would help work out the winding ratios and help 8008guy make a transformer, or if he made it the details would help.

    The base voltage doesn't look like its zenering the transistor, when it does the negative peak gets sliced off flat, and the charge rate taking about 1uS to get high enough to switch on the transistor is about right for the time constant of the 680pF and the 1k resistor of about 0.68uS, so that makes sense at least. And know we know the operating frequency and duration that the transistor should conduct when regulating. Also it looks like the peak collector voltage is nearly 40V, which makes one wonder why the secondary was required.

  7. #27
    Join Date
    Jul 2014
    Location
    Rhode Island
    Posts
    1,375

    Default

    Quote Originally Posted by Hugo Holden View Post
    If m-thompson could measure the voltages on the transformer pins it would help work out the winding ratios and help 8008guy make a transformer, or if he made it the details would help.
    Next Saturday when I am at the RICM I will take 'scope images for the transformer pins. I will also measure the resistance for the coils, if that will help.

    During the week I will continue working on an LT SPICE model of the power supply. As Hugo Holden said, I may have to add something to the SPICE model to get the oscillation started.

    I have an RLC meter. I will try to measure the transformer coil inductance in circuit.

    I also need to finish undoing a modification that the previous owner did to support a RAM chip in the last socket. I tried to get 8k BASIC to load from 7x 2708 EPROMs. It looks like it loaded, but the serial console is unresponsive. Mike Douglas said that there is initialization code in the last EPROM, so I have to fix that socket.
    Member of the Rhode Island Computer Museum
    http://www.ricomputermuseum.org

  8. #28
    Join Date
    Jun 2014
    Location
    Salt Lake City, Utah
    Posts
    561
    Blog Entries
    1

    Default

    Yep, me no coil.
    Spread the joy of Vintage Addiction

    -->www.chronworks.com/<--->www.i8008.net/<--

  9. #29

    Default

    Ok, wind a coil. It looks like the primary and secondary are a 1:1 and the feedback coil is 4:1, or 1/4 of the primary.
    Something like 24 turns for the primary and secondary with a 6 turn feedback.
    Dwight

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
  •