Connect with us

Raytheon GA004 transistors, Digitronics 3500 Paper Tape reader

Discussion in 'General Electronics' started by Eric Smith, Apr 27, 2005.

Scroll to continue with content
  1. Eric Smith

    Eric Smith Guest

    At the Computer History Museum (, we're restoring
    a DEC PDP-1 computer ( In the process, we need to repair
    a flaky amplifier channel in a Digitronics 3500 paper tape reader. We
    isolated the fault to one channel. Each read amp board has three
    transistors, two of which are marked Raytheon GA004, with a 1962 date
    code. These appear to be PNP germanium transistors with a beta >100.
    One of the transistors on the amplifier board that is failing seems
    to have much higher leakage than the others.

    Does anyone have a data sheet on this transistor, a cross reference,
    or better yet some spares? We think we can probably get by with an
    NTE102A, but we'd be happier to get an exact replacement or at least
    enough technical data to prove to ourselves that the replacement we
    choose will be satisfactory.

    Also, if anyone has technical documentation or schematics for the
    Digitronics (or IOMEC?) 3500 paper tape readers, we would really like to
    obtain copies.

    Eric Smith
    PDP-1 Restoration Team

  2. If you find a part that is in the same case style that works you
    could try to re-mark it with the original part number.
  3. Eric Smith

    Eric Smith Guest

    Actually, we don't really care what it looks like. If we can't get
    the original Raytheon part (and I suspect we can't), we're perfectly
    happy to use a more recent replacement, even if it's in a different
    case style, as long as it's not too large to physically fit in the
    available space.

    What we're more interested in is finding out the specifications for
    the original transistor, so we can tell whether any particular
    candidate for replacement is actually suitable.

    Re-marking some other transistor to look like a Raytheon part is
    not required or even desirable. We do mark all the parts that
    we replace, so that it is obvious that they are not the originals,
    and we bag and label all of the removed parts.

  4. Robert Baer

    Robert Baer Guest

    Send me your address and i will send you some Germanium PNP
    transisors. Betas seem to increase with current on all of them.
    Have some new "bullet" JAN 2N128; beta seems to be about 40 at 4mA
    and max 6V operating. Can get lots of these.
    Have mixed old small cylindrical can, TO-1 or TO-44 (1 ea):
    *unmarked (?RCA) good to about 50V; beta near 100 at 2mA.
    *RCA 2N384 good to about 60V; beta near 75 at 2mA.
    *RCA 2N406 good to about 20V; beta near 100 at 2mA.
    *RCA 2N2613 good to about 20V; beta near 200 at 2mA.
    *? B175 good to about 50V; beta near 200 at 2mA.
    *NTE158 good to about 20V; beta near 120 at 2mA (obviously these can be
    readily obtained).

    *RCA TO-5 2N398 good to about 60V; beta near 100 at 2mA.
    *RCA TO-5 2N247 good to about 60V; beta near 100 at 2mA.

    Anyway i will send these to you on an "all or none" basis.
  5. Robert Baer

    Robert Baer Guest

    I forgot to say that i had refurbished a paper tape reader that used
    photo-diodes and Ge PNPs; changed it from a "step" read to a continuous
    or off/on read and added the sprocket hole as a sync channel using the
    same photo-diodes but used comparitors.
    Made a new PCB that was plugin compatible and changed the voltage
    from -48V to +5V.
    Speed increased from i step per second to feet per second.

    The PCB in your reader should be simple enough to chase it out and
    draw your own schematic if necessary That is what i had done).
  6. mc

    mc Guest

    Presumably you can use the address on the museum's web page.

    I'm not involved with this project, but nonetheless I want to thank you for
    doing your part to preserve history!
  7. Eric Smith

    Eric Smith Guest

    Cool! The Digitronics 3500 is rated at 400 lines per second. I'm not
    quite sure whether the interface is "step" or "continuous". The PDP-1 can
    definitely make it read one line at a time, but when the PDP-1 read-in mode
    is used, or software is used that doesn't have a significant delay between
    the tape read instructions, it appears to operate continuously at the
    rated 400 lines per second.
    We may do that at next week's team meeting. It's my understanding that
    we expect to desolder both the suspected bad GA004, and one that is
    believed to be working properly, and look at both on a curve tracer.

  8. The 2N1305 or similar one in that series were a common part in those
    years. As you said, a NTE germanium should work okay, since it's in an
    electromechanical device which probably doesn't put that great a demand
    on the transistors. Also, you might consider swapping two transistors
    so that the leaky one is in a later or less critical stage. But it may
    get worse as time goes on, so it should be replaced.
  9. This was used as a neon lamp driver, and should be capable of a lot
    higher voltage than 60V. Manual says 105V.

    Of course nowadays a high voltage germanium looks pitiful and pathetic
    compared to a silicon transistor.
  10. Robert Baer

    Robert Baer Guest

    Neon lamp driver?
    Darn; there was one GP germanium transistor that was made
    specifically for high voltage and was rather common in those daze - now
    i cannot remember the 2N number.
    In any case for that application, you could safely use a silicon PNP
    high voltage part as a replacement.
    Will get the package out in the mail later today (Friday).
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day