Connect with us

need help repairing old Heathkit weather computer

Discussion in 'Electronic Repair' started by tempus fugit, Oct 25, 2003.

Scroll to continue with content
  1. tempus fugit

    tempus fugit Guest

    Hey all;

    I'm working on repairing a Heath weather computer. Initially, none of the
    funcitons really worked (temp, pressure, clock, etc.) I got it to the point
    last night where everything was working except the clock, which kept time,
    but I couldn't actually set it. So, I (stupidly?) poked around a little more
    to try to get the clock going. I noticed a 4066 with some black stuff on the
    pins, and some weird voltages around it, so I pulled it out, fired the unit
    up, and checked the voltages again. They were still weird, but when I put it
    back in and turned it on again, the pressure and temperature stayed at 0. I
    was messing around with it again tonite, and the pressure and temperature
    readings were more normal, but I noticed a couple of opamps (in different
    locations) that were hot to the touch. There were no shorts in either of
    them, and the voltages around them pretty much matched the schematic. I now
    get bad readings in the temperature (-40) and pressure (0). I'm guessing the
    opamps are at least part of the problem. Any idea why they are getting so
    hot? Any other suggestions?

  2. Nigel

    Nigel Guest

    Personally, I always suspect 4066's, and as for op-amps I always suspect

    If things are getting hot, then are you sure that the supplies are present
    and correct and that nothing is shorted. If the opamps are getting hot, then
    it's a possibility they're drawing a lot of current, and that will affect
    the operation of the entire unit.

    You might also get some useful info from this page
  3. tempus fugit

    tempus fugit Guest

    Thanks Nigel.

    Guess what kind of opamps they are - yup, LM324. When you say "suspect" do
    you mean you suspect they have failed?

    I've been toying with the idea of replacing the opamps with a 074, but I
    don't want to end up just burning another opamp. I'm wondering whether other
    components have caused the opamp failure, although everything around them
    seems to be testing OK.
  4. Tweetldee

    Tweetldee Guest

    From your description, I'd guess that your problem is intermittent, and may
    be easily repairable.
    When you say that there is "black stuff" on the pins of the 4066, is it just
    tarnish or something else? If it's tarnish, then that could be the majority
    of your problems. I assume that the ICs are socketed. If so, you can use
    any of a few methods to clean the tarnish off the pins. The quickest is
    Tarn-X, but you have to rinse them very well afterwards with clean water.
    Don't leave *any* Tarn-X residue on the pins, since it is corrosive.
    Another cleaning method is with a rubber pencil eraser. Be careful with
    this, since it is easy to bend/break the pins.
    What condition are the IC sockets in? They may need to be cleaned or
    replaced if they are in bad condition.
    How hot are the opamps? Are they just warm or really hot? If they're just
    warm, it may be normal, depending on the circuit. Opamps usually are
    low-level components... that is, they don't handle any appreciable power,
    and usually run pretty cool.
    Is the schematic of the unit available on the web anywhere, or can you post
    it to the binaries NG? Might help with some suggestions to help you.
    Tweetldee at att dot net (Just subsitute the appropriate characters in the

    Never take a laxative and a sleeping pill at the same time!!
  5. tempus fugit

    tempus fugit Guest

    Thanks tweet.

    The black stuff is actually on quite a few of the components - diode leads,
    resistors, ICs, etc. I never really looked too close at it, but it almost
    looks like black paint.
    The opamps are HOT. They start out cold, and after about 2 minutes are too
    hot to touch. I have the original schematic, as well as the manual, but
    have noticed some errors on the schem (or the circuit board, I don't know
    which). One thing that I have noticed on a lot of components (mostly
    xsistors) is odd voltage readings. There might be 5v on one side of a bias
    resistor, and 0.3v on the other (xsistor) side, but the xsistor checks out
    OK with a diode test (i.e., not leaky or shorted).

  6. Tweetldee

    Tweetldee Guest

    Undoubtedly, the "black stuff" that you see on the leads is tarnish. It's
    not dangerous to the components, but if it's interfering with mating of
    component contacts, it will certainly prevent the circuit from operating
    correctly. This could be the problem with the opamps.
    There's something seriously out of whack if the opamps are too hot to
    touch.. Don't allow it to run for more than just a few seconds until you
    get it fixed.. the opamps will be destroyed, if they are not defective
    I can't tell a thing from your voltage measurements without seeing the
    schematic along with the voltages. It certainly would be a big help if you
    could scan the schematic and post it somewhere. Without it, it's pretty
    much a waste of time here.
    Tweetldee at att dot net (Just subsitute the appropriate characters in the

    Never take a laxative and a sleeping pill at the same time!!
  7. tempus fugit

    tempus fugit Guest

    Thanks again.
    Unforunately, I don't think I can scan the schem (it's huge for one thing).
    As for the voltages, they don't match up with what's indicated on the schem
    in a lot of places (even when I had most of it working), so I don't know
    which to trust. I'll try removing the tarnish and checking for continuity
  8. Volk

    Volk Guest

    Please get on this Heathkit list.
    You will find a great deal of help there.

    "Heathkit List"
    Send for info to :
  9. Tweetldee

    Tweetldee Guest

    You don't need to worry about the tarnish on the soldered component leads..
    just the socketed components, such as ICs and transistors. If there are any
    connectors that appear to be tarnished, then you should clean those as well.
    BTW, have you checked the power supply output(s) with your multimeter? If
    the supply voltages aren't correct, then you can't expect the rest of the
    circuit to operate properly. How abot ripple on the PS lines? If there's a
    lot of ripple, then there will be all sorts of chaos in the circuit.
    Tweetldee at att dot net (Just subsitute the appropriate characters in the

    Never take a laxative and a sleeping pill at the same time!!
  10. tempus fugit

    tempus fugit Guest

    The PS voltages were the 1st thing I checked. They were pretty saggy, so I
    replaced the main filter caps. That helped, but didn't fix the problem, so I
    moved on to the regulators, and found that the -15v regulator was out of
    commission, so I replaced it. This improved matters somewhat, but there were
    still problems, so I kept poking, and found a shorted PNP on the CPU board,
    which I replaced.
    I have a scope, but am kind of new to it. I basically attached the ground
    clip to the case and probed the regulator's output, which showed no ripple
    (which seemed a little unrealistic to me). Is this the proper method for
    checking for ripple?
    As for the tarnish issue, I'm just going to check for continuity from the
    IC's pins to the connections on the solder side of the board.

    Thanks again, and keep those suggestions comin'
  11. Tweetldee

    Tweetldee Guest

    Yeah, your scope connection will be with the ground connection of the probe
    going to power supply return (most likely connected to the case, but not
    always) and the probe tip to the circuit point you're monitoring. if you
    saw no ripple there, then the PS is probably OK.
    How long has this unit been out of commission? Is it possible that it a
    lightening strike or power surge may have caused its demise? If so, then
    it's possible that most, if not all, of the semiconductors in the unit are
    bad. If the sensors are outside, and close to a power line or phone line,
    then it's possible that may have been the cause of failure.
    It's going to be a bit difficult to diagnose it without a schematic to
    reference, but if you can adequately describe the circuits and their
    symptoms, then we may be able to get it going again. I searched all the
    usual places for freebie downloads, but no joy. +
    Tweetldee at att dot net (Just subsitute the appropriate characters in the

    Never take a laxative and a sleeping pill at the same time!!
  12. tempus fugit

    tempus fugit Guest

    Once again, thanks for your reply.

    I have no idea how long it has been dead. My brother in law got it for free
    (!) at a garage sale this summer, and was told that it was working.

    I was messing around with it again tonite, and I noticed a gross error.
    Remember I said I pulled a 4066 out because the voltage readings were way
    off from the schem? Well, it wasn't a 4066, it was a 4016, and (this is the
    embarrassing part) I had put it in backwards when I reinstalled it. Anyway,
    the good news is that (now that it is in the right way) the opamps are
    running at a normal temperature (slightly warm). I'm wondering though, if
    that part isn't screwed (the 4016) because when I say the voltages are off,
    I mean the wrong polarity as well as the wrong voltage. Maybe I'll pop a new
    one in and see what happens.

    Here's another thought: this thing was designed in the 70's, so maybe the
    CPU has a Y2K problem. I found a site that mentioned that the time was an
    issue because of this for this unit, but it didn't describe the symptoms
    I've got.
  13. James Sweet

    James Sweet Guest

    Does it keep track of the date? If not then obviously it doesn't have a Y2K
    problem, even if it did, the worst it would do is roll over to 1901, highly
    unlikely that it keeps track of the year beyond 2 digits.
  14. Tweetldee

    Tweetldee Guest

    If you plugged the 4016 in backwards, then it's probably damaged.. replace
    it for sure. The 4066 can be used in place of the 4016, with better
    performance in some applications.
    As I said in an earier post, opamps usually don't run hot, or even warm.
    If a new 4016 doesn't fix your problem, you might replace those opamps and
    see how it works.
    Does the unit keep track of time and date? If yes, then you may have a Y2K
    issue with the date display, but I doubt that could be causing your
    operational problem. Worst case, the date would be wrong, probably jumping
    back somewhere in the 1900's.
    Tweetldee at att dot net (Just subsitute the appropriate characters in the

    Never take a laxative and a sleeping pill at the same time!!
  15. tempus fugit

    tempus fugit Guest

    Yea, it does keep track of the time and date, but I can't actually set these
    parameters. Maybe I'll try replacing the ICs and go from there. The opamps
    are LM324s. I've got a spare TL074 around - do you think this would be an OK
    replacement, or should I replace them with the originals?

    Thanks again
  16. Tweetldee

    Tweetldee Guest

    Without seeing the schematic, it's a hard call as to whether the TL074 would
    be a good replacement. I haven't looked at the spec sheets of either unit,
    but I do know the the TL074 is a JFET input amps, whereas the LM324 is not.
    I assume the pinout is the same.. you will need to check that out.
    Probably wouldn't hurt anything to try the substitution, but couldn't
    guarantee they would work.
    Is the CPU in the unit a programmed chip or does it have a PROM/EPROM/EEPROM
    to store the code? If it's a programmed CPU, then you might surf around on
    Google to see if there are any replacements available. The switches that
    are used to set the time/date, etc, may be tarnished and not making contact.
    Check all of the switches for good, solid (not intermittent) low-resistance
    But, going back to the original problem, I would continue to look at the
    areas that may be affected by the tarnishing of the contacts on IC's,
    transistors, or any other components that may be socketed. Also, you
    should look for components that are loose. Sometimes, the plating on the
    leads of components comes off, and effectively breaks the connection. If
    there are any suspect soldered connections like that, then do the obvious
    and resolder them.
    Tweetldee at att dot net (Just subsitute the appropriate characters in the

    Never take a laxative and a sleeping pill at the same time!!
  17. They should be _first_! :)
    Not unrealistic at all, if the lowest point of voltage in the ripple at the
    input to the regulator is above what the regulator wants to see as minimum,
    you won't see any on the regulator output at all. This can be the case
    even when there's several volts of ripple on the input side.
  18. Pretty near the same part (different spec for "on" resistance is most of it
    as I recall).
    Ouch. I hate it when that happens... :)
    It still not be any good, though. Reverse power to a CMOS part like that
    can result in fairly high current, since the power pins are then looking
    at what amounts to a forward-biased diode. That thing may be cooked. But
    they're cheap.
    There ya go.
    What would that thing do with a date?
  19. tempus fugit

    tempus fugit Guest

    Hey Tweetldee;

    I don't know if I mentioned in my OP, but the 1st thing that was wrong was a
    defective -15v regulator. I'm wondering if all the ICs that needed +/-15v
    got damaged (these would be the 4016, 324, and a couple 1458s). The voltages
    around these seem OK though, and there are no direct shorts. Maybe I'll try
    replacing them and see what happens. They're cheap, so it will only end up
    costing a couple bucks.
  20. Those 40xx CMOS parts don't do real well with over-voltage on the supply

    See if any of them are at all warm to the touch -- if so, then _those_ are
    the ones that need to be replaced.
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