Image Map Image Map
Page 1 of 2 12 LastLast
Results 1 to 10 of 16

Thread: Tandy 1000 TL that eats power supplies

  1. #1
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    4,629
    Blog Entries
    3

    Default Tandy 1000 TL that eats power supplies

    I got a Tandy 1000 TL for $75. I checked out the power supply beforehand, everything seemed alright. I fired it up.. the machine made a clicking noise and the floppy drive light flashed on and off. I began to smell something burning, so I shut it off. I then disconnected all the peripheral stuff and tried firing up again. This time no clicking, but eventually one of the caps blew.

    I conferred with a few Tandy 1000 people and decided to borrow the working supply from my regular 1000. It doesn't have the same wattage, but all we were trying to do was power up the mainboard and see what happens. Once again, I got the clicking and a noticeable burnt plastic smell. I immediately shut it off and looked for damage, but did not see any.

    I'm guessing there must be a short of some kind on the mainboard - maybe a diode or something. I'm not sure how to safely proceed - should I be worried that the 1000 PSU will blow up on me once reconnected to the 'good' 1000? Any precautions I can take there?

    I'm trying to think how I could diagnose this without blowing up power supplies every two minutes.

  2. #2

    Default

    Use a bench supply with a current limit. Most often the short will get hot.
    If you are blowing up capacitors on the supply, it is likely that there is a short between opposite rails of the supply, such as +5V connecting to -12V. Usually the +5 will over power the -12v and reverse the voltage on that line, blowing one of the capacitors on that rail.
    Even with a bench supply you'll need to be careful bringing up the levels so as to not reverse the voltage on any of the supplies. It other words, don't just preset the levels and connect all up.
    There is another method but I'm sure it will be hard to understand. You need a sensitive volt meter. It needs to see micro volts. A 5 digit meter is usually needed.
    In this case, you use one current limited supply. You choose the rail with the most resistance. In the case of the +5 and -12, the -12 is likely just a wire trace and not a power plain. You locate two ends of the trace and connect them to the supply ( yes, this is essentially a short, that is why you use a supply that is designed to run in current limit ). You raise the current until you can get a noticeable voltage drop across the trace. 10 to 100 millivolts is usually enough.
    Now you connect up your meter. Place one lead on the supply rail that this rail is shorted to. With the other meter lead, you trace along the first trace looking for the lowest voltage, plus or minus. When you get to 0, you are at the short. Do note that if there is a branch in the trace, it may only bring you to the branch. You may have to move the supply connection and feed current through that branch.
    This method even works if both rails are power plains. It usually takes a large power supply to feed enough current through a plain to see a noticeable drop. You place the supply across two opposite corners of the board. You measure to the other plain and form a line across the board of 0 volts. I usually use a piece of string and some clear tape, since it is usually a curved line. Then move the supply to the other corners. You repeat the measures. Where the lines cross, is where the short is.
    Note, you are not feeding the current through the short, you are only feeding the current along a trace. Keep the open circuit voltage of the supply to less than 0.25V for safety of the components on the board. Remember to is the high current along a trace that we want.
    How this works?
    Imagine that you have a resistor that has a fixed tap some place along it and you have a movable wiper. You feed a current from one end to the other of the resistor ( in our case a trace of the board ). You put your meter with one lead on the fixed tap and the other on the wiper. You move the wiper and look for 0 volts. That must be the same point that the fixed tap is located at.
    Many have said, why not just feed the current through the shorted rails. In most cases the short will get hot and expose itself. Why go through all the troubles. In some cases, the shorted part will open up and not show itself. Now you have a part that has failed, you are not sure where it is. It may short out again while doing normal trouble shooting, taking out another power supply.
    Do note that you can not locate a short with an ohm meter. It only tells you there is a short, not where it is.
    Dwight

  3. #3
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    4,629
    Blog Entries
    3

    Default

    Thanks Dwight!!

    I tried a variation of this.. just running my DMM in ohms mode and seeing the resistance between power supply pins on the motherboard. But that unfortunately did not reveal anything. I'm guessing from your description that it is a bit more complicated than that.

    Would you be afraid to fire up the 'good' power supply on it's own computer after getting a slight burn like that?

  4. #4

    Default

    Quote Originally Posted by falter View Post
    Thanks Dwight!!

    I tried a variation of this.. just running my DMM in ohms mode and seeing the resistance between power supply pins on the motherboard. But that unfortunately did not reveal anything. I'm guessing from your description that it is a bit more complicated than that.

    Would you be afraid to fire up the 'good' power supply on it's own computer after getting a slight burn like that?
    Generally most supplies that we have don't have over current protection on the outputs, other than a input fuse. Also the only place I saw any kind of reverse voltage protection was on my lowly KIM-1. In general if a chip shorts a +5V rail to -12V, it will take out the -12V rail. This is why you want to first check the supply leads with an ohm meter to ensure there are no combinations of short before attempting to power up. You should also check that the supply doesn't over voltage as well with some load ( many older switchers needed some load or they would hic-up, often blowing them up ).
    To do the test I talked about, you so need to get good test equipment. I used a 5 Fluke meter as I recall. Some scopes have a low enough scale that you might be able to see the drops. You also need a supply that can provide amps. For ground plains, I used a supply that produced 20 amps under current limit. For traces leads one can get away with 2 to 5 amps, depending on the sensitivity of the meter you are using. Remember, you are only getting 10 to 100 mv across the wire or plain. Even for a trace of say 10 inches, 100 mv and 5 amps, that is .05 watts per inch. It is not going to damage anything.
    I suppose, one could use a galvanometer in place of a meter but I don't have one handy. You could take an old analog meter and just wire the movement as a galvanometer but use with care, starting at a lower current and increasing the current as you get closer to the short ( you don't want to be bending the needle of blowing the movement ).
    Look at my resister analogy and think of the one trace as being a resistor and the line that it is shorted to as the tap point on that resistor. Your meter probe is the wiper.
    You do need a good bench supply and some way to measure low amounts of voltage.

    A thought is to use a variable transformer to supply a low voltage transformer as a current source. You can then use an AC clip on current meter to control the current to something the transformer can handle. It has the advantage that you can now use an AC amplifier to look for smaller voltage drops. I know, I'm just thinking how else one might do this with things they might have in the scrap box. An opamp with AC coupling would be easy enough to have the gain to use a AC meter on.
    I'm just thinking a little outside the box on this.
    Anyway, look at my resistance analogy and think about how it works. I've used this method even to find a 10K high resistance short. One has to be aware of shunt paths when looking for high resistance shorts but most ICs draw little current at 10 mv. There are also tricks one can use to look for multiple shorts by moving the forcing supply connections around ( done that as well ).
    Dwight

  5. #5

    Default

    I should note why using an ohm meter to locate a short is usually useless.
    First the short is usually less than an ohm. Lets say it is 200 mlliohms. The traces supplying the power rail are usually in the order of 5-10 mlliohms. How good is you meter at seeing such small differences. I've used such a meter but its resolution was limited 2 mlliohms and used 4 leads.
    Dwight

  6. #6
    Join Date
    Sep 2008
    Location
    SE MI
    Posts
    4,208
    Blog Entries
    6

    Default

    Falter

    With the amount of work that you do it might be a good thing to invest in a Variac 'similar' to this one:

    https://www.ebay.com/i/303225484771?...4aAk1GEALw_wcB
    Surely not everyone was Kung-fu fighting

  7. #7
    Join Date
    Jun 2013
    Location
    Montevideo, Uruguay
    Posts
    391
    Blog Entries
    1

    Default

    if that board have tantalum caps remove all of them and check them for shorts, its common they short up in this old boards.

  8. #8
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    4,629
    Blog Entries
    3

    Default

    I found the source of the burnt plastic smell, I think. This 74LS244 is definitely showing signs of having burnt up.

    20191016_212328.jpg

    I think these are 5V devices.. so perhaps that hints at where to start looking.. if it's getting too much voltage. I don't know if ICs can short themselves out internally and cause something bizarre like blowing out a PSU.

    Interestingly, the legs on it are rusty, where as the ones next to it aren't. I'm guessing there is something going on underneath - I don't see any signs of a problem otherwise.

    Also I'm looking for diodes... I don't see anything labelled with a D.. but I do see what look like diodes labelled with an F?

  9. #9
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    4,629
    Blog Entries
    3

    Default

    I think I'll pull the board out tomorrow and do a check on the backside. Visually - that chip is the only one that seems to have any damage. Not sure why the one right beside it didn't also take damage.

  10. #10
    Join Date
    Jul 2009
    Location
    Central Florida
    Posts
    3,760

    Default

    Yes IC's can short out internally had a few of them do it. Good idea is to replace it. The LS244 is basically a line amplifier type of IC. A buffer chip. It takes as an example the 8 data lines from the data bus and amplifies the signal going to the next device along the bus. As to why it would blow is anyone's guess maybe a brief overload shorting the 5v line? I don't know. And you are right they are 5v devices. The F devices are ferrite cores with a wire running down the middle and act like an inductor to filter out any stray noises on the power lines. Specifically any stray AC

    Ok easiest way to replace an IC is to gut all the "legs" off close to the body. Once that is done remove the IC. Now take a soldering iron and remove the legs from the top side of the board using metal tweezers. Once that is done use a solder sucker to clean out the holes from the bottom of the board. Put in a 20 pin socket and solder it's leads. Then put the replacement LS244 into the socket. Hopefully this will fix the issue you have. Oh and this is guessing that you can solder and de-solder parts. Good luck.
    *FrankG*

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
  •