Connect with us

newbie solder question

Discussion in 'Electronic Repair' started by Dave, Jan 19, 2006.

Scroll to continue with content
  1. Dave

    Dave Guest

    I've been following the electronics newsgroups for some time, and decided to
    try a small repair job. I have a 20Wx2 small single-board amplifier which
    has only worked on one channel since it was given to me. Due to its' small
    size, I'd like to repair it. Anyway, I used a DMM to test each of the 20 or
    so electrolytic capacitors on the board. I know, I know, you can't really
    test them in the circuit and expect a 100% accurate result. But as it's a
    two-channel amp, I figured that at least the equivilant capacitors in
    Channel A and Channel B should exhibit identical behavior.

    I located two 470uF caps which differed in that one of them had zero
    resistance which did not rise, even on the highest resistance scape of my
    DMM, 20M ohm, as did all of the other caps. So, I figured I'd swap the two
    identical caps (channel A and channel B). I tried using desoldering braid
    (which I've used many times with great results) and found that the solder on
    this board just wouldn't melt with my 30W iron. Well, says I , I'll just
    use a bigger stick. I grabbed the 45W iron and still couldn't seem to melt
    the solder through the braid, although the 45W iron WOULD melt the solder
    directly. I heated each lead and wiggled and jiggled it loose. When I had
    both caps removed I put the desoldering braid directly over the hole in the
    PCB and with a pointed tip leaned on the board until the braid cleared up
    the solder. This took maybe a minute or more of continuous heat.

    When I removed the brad, AAAAGGGHH, it seems to have removed all of the
    metal around the hole, not just the solder. Now the trace itself to which
    the cap is connected is on the other side of the board and I could put in a
    cap with slightly longer leads and solder it on the top. But if there's a
    way to fix what I've done that'd be even better. When I put in the original
    cap the solder just won't seem to go into the hole any more, it just sticks
    to the iron.

    Any advice greatly appreciated, I suppose this is how folks learn.

    BTW, when I got both caps out they tested identically out of the circuit so
    I'm pretty sure my problem is elsewhere. The circuit consists of a simple
    power supply, unregulated, a whack of tiny (mainly 10uF 50V) electrolytic
    caps, some mylar caps, resistors, a bass/treble IC, two 20W TO-220 amplifier
    chips. I can very very faintly hear sound in the bad channel. When I turn
    the volume up and down on the bad channel it makes a sort of "thumping"
    sound from the bad speaker but doesn't amplify the sound. I've replaced the
    amplfier chip on the bad channel as it was cheap and simple but this did not
    help. Anybody have any suggestions what to check next? I do not own an
    oscilloscope. I have checked and cleaned all of the knobs (bass, treble,
    balance, volume), they all work as they're supposed to.


  2. Mike Berger

    Mike Berger Guest

    You often have to isolate components or parts of the circuit to
    diagnose things anyway. Personally, I prefer a very hot temperature
    controlled soldering iron and vacuum type solder sucker. Work fast
    and don't let the heat travel beyond where you want it. I know
    others prefer your method.

    If you do lift a pad or trace, cut it with an exacto knife to keep
    it from lifting any more and replace it with hookup wire wrapped
    around the component leads and soldered to the edge of the broken
    trace. If the traces are covered with a solder mask, you'll have
    to scrape some of that away before you can solder.
  3. GregS

    GregS Guest

    You didn't mention measuring at the same points with caps removed.
    I usually recompose circuit traces with solid conductor wire.
    Probably at least a 50% chance the chip is bad.
    Sometimes I have to double up on irons. I've also lit the Bic
    next to the iron for more heat. My Radio Shack 250 watt gun for big jobs!

  4. Mr. Land

    Mr. Land Guest

    Give up using the braid...that stuff sucks (or, rather, it doesn't).
    Get yourself a solder sucker if you're going to do stuff like this.

    BTW, if I had an audio power amp that didn't work, I think the caps
    would be near the end of my list of things to check. Start with the
    power (aka output) transistors and work your way backwards. Unless
    this is a really esoteric unit, both channels probably share the power
    supply, so if one channel is working, you can probably assume that's OK.
  5. Dave

    Dave Guest

    As I noted before I've used the braid quite a bit in the past and found it
    to work well... but will pick up a sucker anyways as obviously I need it.
    I did start at the output transisters and replaced them all on the bad
    channel (they're all on an integrated 20W amp chip, NS LM1875). I swapped
    the caps in an effort to avoid trying to source a 10-year-old base/treble IC
    which is likely the problem. Other than the output IC's, mixer IC, and
    rectifier, there ain't a whole lot on the board except caps and resistors.
    Resistors all checked out OK, move on to electrolytics... No?
  6. Dave D

    Dave D Guest

    My general rule is a solder sucker for standard work and braid for smaller
    or awkward tasks, especially certain surface mount work. A solder sucker is
    far more appropriate for removing components like electrolytic caps from an
    amp IMO.
    Do you still have a short? If so that makes things very easy- a short is
    probably the easiest fault there is to diagnose. Simply trace the two PCB
    tracks which are shorted and find all other components across these tracks.
    One of them must be shorted out. Is the capacitor in question on the supply
    rail? Check that the new IC you fitted doesn't have any solder bridges, and
    if it has a mica heatsink insulator, make sure it's not damaged. I doubt
    it's a shorted cap, but it is possible nonetheless.

    It's possible the 470uF capacitor is part of the decoupled supply to the
    preamp section, if so it may have a zener diode in parallel to give a
    regulated supply. Therefore, check for shorted zeners.

    The output ICs no doubt are coupled to the speakers via large electrolytic
    capacitors, what is the voltage level on the IC side of this cap with
    respect to ground? It should be close to half supply potential. Compare it
    with the good side, but take care not to short anything out.

    Are both output ICs getting a supply? Are they both getting a ground? Have
    you measured voltage levels on each pin and compared them with the good
    side? Have you downloaded a datasheet for the ICs, found the input pin and
    injected a signal there to listen for output? Even touching a finger to the
    input pin may give a buzz from the speaker, or a screwdriver may give a
    click. There's ways and means, even without test equipment!

  7. default

    default Guest

    Well back in the "old days" we used "eyelets" - didn't have plated
    through holes for circuit boards. The eyelets were just miniature one
    part hollow rivets - or similar to grommets used in sails or
    tarpaulins. That would be an ideal fix if you burnt the plating off.

    Solder Wick for single sided boards - solder sucker for double sided

    Regarding the amp. Have you checked the DC voltages between the
    channels? Is this a single supply or plus and minus supplies? TO220 -
    five lead power op amp? built in thermal, short circuit, and safe
    operating area protection? I assume you only have one chip per
    channel (or "bridged" with two chips per channel?)

    If you can identify the chip and its input(S), you might try using a
    capacitor to bridge signal from the working channel's input into the
    input of the dead channel. (observing polarity if using an
    electrolytic). Unplug the input to the dead channel while testing if
    possible. The cap size will be 10 uf or less for a high impedance
    audio stage input. That should localize the problem between the input
    or output of the power amp chip.

    If it is like the TO220's I'm using, there's a differential input
    (inverting and non-inverting) inputs - make sure you drive the correct

    One assumes you switched speakers at some point.
  8. Dave

    Dave Guest

    I think I'll solder to the trace on the other side of the board, just means
    that the cap won't sit flush anymore but that's okay, lots of room in the
    Yes, must buy solder sucker. It's on my list now.
    No. Do you mean between the channel A and channel B signal outputs with no
    signal applied to the inputs? If there WAS a DC voltage differential what
    might this tell me? Failed coupling cap? There are no output caps, the amp
    output pin (see below) goes straight to the center pin of the RCA jack on
    the back.
    Both the amp and bass/treble/volume/balance IC's run off +Vcc.

    TO220 -
    One chip per channel. They're LM1875T's, yes they have built-in thermal,
    short circuit, etc. As I noted I have already replaced the one on the bad
    channel (they are a TO-220 package).
    Does it matter if I use an electrolytic or not? See pinout below, but I'd
    connect pin 1 of channel A (working) to pin 1 of channel B (dead). If I do
    use electrolytics, which end is positive, A or B? Why do I need to use a
    capacitor to bridge as opposed to a wire?
    Here's the pinout:

    1 +IN
    2 -IN
    3 -V(EE)
    4 OUTPUT
    5 V(CC)

    From what I can see, pin 1 is signal in, pin 2 is used as the negative
    feedback loop, being connected to pin 4 via a 20K resistor. There are no
    coupling caps on the outputs.
    Yes, first thing I tried.
  9. Dave

    Dave Guest

    I checked for solder bridges on the new IC VERY carefully when I put it in,
    as I did also inspect the mica insulator.
    The balance/tone/volume IC (which is basically my entiere 'pre-amp section')
    has a zener internally, BUT I see that it also uses an external decoupling
    cap on the internally regulated voltage supply... must check which cap
    performs this function.
    No, no output caps at all. Output pin of amp IC goes directly to RCA plug
    via large trace.
  10. Arfa Daily

    Arfa Daily Guest

    All of the advice given so far regarding the fault, is good stuff. As far as
    solder wick / solder sucker / desoldering stations go, it's horses for
    courses. Contrary to what one of the posters said about solder wick "
    sucking " ( or not ... ), in my experience, this is only the case when you
    either buy cheap, or use the wrong size for the job. Buying cheap at radio
    rallies, will just get you old stock, where the flux has gone off, and then,
    it doesn't suck. Using a solder wick with too big a size, with an iron
    that's too small tip wise, or power wise, results in insufficient heat
    transfer to the joint, and then it doesn't suck.

    A solder sucker is very good for medium sized joints, but there is a real
    technique to using one well, and if you use it on a ' poor ' quality board,
    it will readily suck the print off as well as the solder. There is a real
    balance between getting enough heat into the joint to melt the solder to a
    point where it will stay molten enough to be sucked cleanly off the board,
    and not destroying the bonding between the copper and the substrate.

    If you do get yourself a solder sucker, get a good one for which all parts
    are available - nozzle, neoprene suction washer, circlip ( you'll lose it
    ! ) etc. Practice a lot on a scrap board to get the heat / time thing right.
    Clean it out regularly, and don't buy cheap !!

  11. Dave

    Dave Guest

    Well, my desolder braid is cheap and old, and PERHAPS a bit too thick for
    the job, but it's worked well recently on similar-sized joints using the
    same iron.
    Thanks for the advice, have not used one before. Unfortunately where I live
    there are very few stores which sell electronic parts and buying one off of
    the internet generally costs twice as much when you get done with shipping
    and taxes and the like.
  12. Arfa Daily

    Arfa Daily Guest

    This is true. I used to use solder wick years ago, and it wasn't very good,
    so I stuck to a solder sucker. Then, a few years ago, huge amounts started
    appearing at rallies for like 1 UKP per reel, so I bought loads, as did
    friends. It was crap. I then read an article in my national ham radio
    journal about this, and it said that the reason that the stuff didn't work
    was because even though it was housed in a basically airtight container,
    over several years of storage, the oxygen in the atmosphere reacted with the
    powder flux that it's impregnated with, and destroyed its fluxing
    characteristics. Hence the reason that it had been dumped cheap on the
    amateur market, as it was of no commercial value at all.

    Soon after this disappointing experience, I had occasion to be in a fellow
    professional's workshop, and saw him using some solder wick made by "
    Multicore ", and was amazed at its efficacy, so I immediately ordered a
    reel, and now always have some at the ready. It's not cheap at 30 UKP for a
    30m reel ( about $48 ) but it lasts a long time. I also use a Weller vacuum
    desoldering station, but that's too expensive for amateur use, unless one
    can be picked up cheap on e-Bay.

  13. Smitty Two

    Smitty Two Guest

    Liquid flux. If you solder, you need it. Liquid flux. If you desolder,
    you need it. Liquid flux. Good solder braid is better than any sucker
    I've ever used, with the exception of the heated, power suckers, which
    are a tad pricy for the casual hobbyist. Liquid flux.

    Did I mention liquid flux?

    A minute of continuous heat? I'd recommend a better quality soldering
    iron. You shouldn't be taking any longer than a second or two to solder,
    and two or three to desolder, small components like that.

    Also, get some liquid flux.
  14. My experience with solder sucking bulbs and plungers has been a lot
    worse than with braid, except when the joint had a huge amount of
    solder, and even then I sometimes had to remove the final traces of
    solder by using braid

    I once used braid that was over 1/8" wide, and it was terrible because
    it absorbed too much heat. 0.06" - 0.08" works best for me. I also
    avoid braid that's tarnished (won't stick to solder) or coarse (not
    enough surface area, probably low quality as well), and I keep the
    soldering iron tip really clean and cut off used braid right away
    because dangling braid just draws heat away from the joint. I cut it
    even if it isn't full of solder because its flux has been ruined by the
    heat, keeping it from stick to solder.

    When solder won't melt into the braid, I add more solder to the joint
    because it seems that factory solder has a higher melting point or some
    kind of film over it.

    I don't understand why 45W isn't enough heat, but some irons seem to
    deliver more heat to their tips than others of the same watt rating do,
    or they get weak with age. 35W has been enough for me with
    single-layer boards, except on large copper areas, and 45W should be
    able to do 2-layer boards or even the smaller components on multilayer
    boards, even when the lead connects to the internal ground or voltage
    plane (I normally use 50W).
  15. Ray L. Volts

    Ray L. Volts Guest

    That's far too long to hold an iron to a joint.
    Aside from using functional (i.e. quality) wicks, something you need to keep
    in mind is tip cleaning and tinning.

    As for wicks, I've tried lots of different brands and from my experience
    Tech-Spray's Pro Wicks and Easy Braid's Quick Braids give the best bang for
    the buck. I've never had a spool of Pro Wick fail me, even after many years
    in storage. Quick Braid is a little less expensive, though both are priced
    very reasonably. Pro Wick performs better, with a little quicker heating
    and a faster solder draw. Avoid Pro's Tool wicks like the plague.

    A good compromise between an expensive professional desoldering station and
    a cheap solder sucker is one of the powered solder suckers. These have a
    heated barrel like the desoldering station guns and a simple vacuum chamber
    like the inexpensive solder suckers. The wattage of these irons is
    typically 30-45W. They can be had for around $20-$30US.
  16. GregS

    GregS Guest

    I would not be without my flux pen.

    I never tried to make some. Just alcohol and regular

  17. jakdedert

    jakdedert Guest

    BTW, one use for the flux--I use the paste stuff, must try the
    liquid--is to 'activate' old desoldering braid. When doing a lot of
    work on old boards, I keep the tin of flux open, heat the braid and dip
    the end into the flux between uses. It activates the braid and makes
    for much easier desoldering.

    Is that also the way to use the liquid stuff? It would probably be a
    less messy than my procedure.....

  18. GregS

    GregS Guest

  19. Mr. Land

    Mr. Land Guest

    So, Dave...

    This topic seems to have drifted off of the subject of helping you
    troubleshoot your amp and onto desoldering techniques and tools...(not
    that that isn't great content).

    So have you fixed it?

    If not, what's this 10-year-old mixer IC you mention? Do you have a
    pinout of it? How do the DC voltages on all the pins of the output IC
    compare? If you have a meter-based meter (was that redundant?), can
    you see any DC fluctuation on the input or output pins of the output
    IC's when you spin the volume control? Check both channels and compare

    I've seen speaker "thumping" occur when DC was getting to an input it
    shouldn't be.
  20. Dave

    Dave Guest

    Um, not yet. As you noted below, I also thought about DC getting to the
    speakers being the cause of the thumping. That's what started me on my
    capacitor-testing adventure in the first place. Which led to my frying the
    PCB traces trying to remove the caps and associated solder with solder
    braid. So I've got to procure a new pair of 470uF caps (with nice long
    leads) that I can solder onto the other side of the board. After (hmm,
    maybe at the same time) I procure a solder sucker and some new
    decent-quality solder braid and some liquid flux (from other posters) to
    clean up the mess I've made. I see that there is quite a bit of technique
    involved with this de-soldering stuff... I'll get there.

    The IC in question is an LM1036N, combo volume/tone/balance control. It's
    available from Digikey for <$5, maybe I'll just order one for the price of a
    Big Mac and fries, get my caps and assorted tools too. When I get the caps
    back in I'll bridge one amp IC's input to the other one (via a 10uF cap
    according to another poster perhaps to limit any DC voltage from getting
    from the "bad" channel to the "good" and wreaking havoc there). This should
    at least tell me if the output stage is working correctly. I guess I should
    check input voltages at both op-amp IC's to make sure the bad channel is
    being powered first. As I noted I've already replaced the bad channel's
    output stage (LM1875T 20W amp TO-220). I've already tested all of the
    diodes and resistors and (to the extent that they can be tested in-circuit)
    caps with a DMM (not the best test but will show if a cap is open circuit or

    I didn't think of using an analog meter, that's a good way to roughly
    compare outputs... I'll make up a CD with continuous tones to feed a signal
    via the unit's RCA inputs and test the signal at various points moving
    backwards from the output stage. I have a feeling it's the combo
    volume/tone/balance IC but then I had a feeling it was the amp too and it
    ain't.... I think I suffer from "jump to unsubstantiated conclusion"
    syndrome and I just have to slow down and think things through before
    running off to the store to replace another part.

    I'll keep folks updated (a retain-my-pride way of saying "I'll likely have
    more questions") via this thread.

    Thanks for your help so far...

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