Connect with us

Stereo amplifier volume fades over time

Discussion in 'Troubleshooting and Repair' started by fredmeews, Mar 29, 2014.

  1. fredmeews

    fredmeews

    4
    0
    Mar 29, 2014
    Hello! A problem has developed with my solid state subwoofer amp: after a short (1-2 minutes) time, the volume fades.

    If I switch it back on immediately, the volume is still very low. If I wait 1-2 minutes, the volume is back at a normal level, then fades again.

    Any ideas? Thank you!
     
  2. davenn

    davenn Moderator

    13,264
    1,749
    Sep 5, 2009
    hi there
    welcome to the forums :)

    Maybe power supply failure due to heat
    show us some sharp and well lit pics of the circuit board(s)
    do you have any test gear ... a multimeter??

    cheers
    Dave
     
  3. fredmeews

    fredmeews

    4
    0
    Mar 29, 2014
    Dave - Thank you for the reply! I don't have any test gear, but could buy it if you think there's a chance of a total newbie figuring this out.

    Here are the pics of the circuit board. You can see where I desoldered / cleaned / resoldered the pot.

    Also here is what the outside of the unit looks like. The amp is inside a subwoofer box that powers 2 small satellite speakers.

    Thanks again!
     
    Last edited: Apr 19, 2014
  4. k7sparky

    k7sparky

    6
    0
    Mar 7, 2012
    On the back of the board looks like your power (output) inline chip or chips can't tell.
    I see some white heatsink goop and that I am guessing is your power chip.
    It needs a good physical contact to a heatsink. If it has an overtemp roll off that would do what you are describing.
    If that is the case it will be plenty hot. Might have had a clip of some sort holding it to a heat sink
     
  5. KrisBlueNZ

    KrisBlueNZ Sadly passed away in 2015

    8,393
    1,266
    Nov 28, 2011
    Thanks for the photos. Here are some disordered comments.

    It looks like there are two output ICs along the back edge of the board, with the white thermal goop on them. I assume you've unscrewed these from the heatsink to get the photo, right?

    It could be useful to know the markings on those devices.

    My guess is that one drives the left and right satellite speakers, and the other one drives the subwoofer in bridge-tied-load configuration.

    When the problem occurs, do all of the speakers lose volume? All equally?

    Is there distortion as well? If so, can you describe it?

    There has been resoldering done on one of the output ICs, and some of the pads running in a line next to it, as well as the potentiometer. You should clean up these areas and take some close-up photos.

    The best way to clean it is using a solvent (ispropyl alcohol, if you can get it) and an old toothbrush, then cotton swabs. You can press pretty hard, as long as the copper is firmly attached to the board.
     
  6. fredmeews

    fredmeews

    4
    0
    Mar 29, 2014
    Thank you both for the quick responses!

    I assume you've unscrewed these from the heatsink to get the photo, right?

    They're both screwed into this metal piece... each with a metal pad, the one on the right seems to have an extra pad behind that.

    My guess is that one drives the left and right satellite speakers, and the other one drives the subwoofer in bridge-tied-load configuration. It could be useful to know the markings on those devices.

    I think you're right! The left one has wires going to the satellite (TDA1554Q), the right to the sub (TDA1514A).

    When the problem occurs, do all of the speakers lose volume? All equally?

    No... just the sub loses volume.

    Is there distortion as well? If so, can you describe it?

    No distortion... it's just as if someone gradually turns down the volume

    There has been resoldering done on one of the output ICs, and some of the pads running in a line next to it, as well as the potentiometer. You should clean up these areas and take some close-up photos.

    Interesting... I only resolderd the pot. Ok, I'll get some solvent & clean it up. Thanks again guys!
     
  7. fredmeews

    fredmeews

    4
    0
    Mar 29, 2014
    When I look at the sub's output IC chip I can see that the little metal plate used to screw it to the metal piece is bent... this could be the loose physical contact that is causing the overheat you're mentioning. Is this type of heatsink made of a special material or is it just a piece of metal? Thanks again.
     
  8. KrisBlueNZ

    KrisBlueNZ Sadly passed away in 2015

    8,393
    1,266
    Nov 28, 2011
    Yes I see what you mean. The one on the right is pretty mangled, and that would be a likely reason for the problem.

    That bit of metal seems to be a spacing shim that conducts heat from the IC to the main heatsink. It looks like it's made from mild steel, I think. It needs to be very flat on both sides so it will make good contact all the way along. White thermal goop will help, and there should be some on each side of the shim, but it's very important to have close and even contact.
     
  9. k7sparky

    k7sparky

    6
    0
    Mar 7, 2012
    There are a few of the amp chips that have the tab bent back at 90 deg above the mounting hole / slot.

    As Kris says FLAT is necessary. In a perfect world all surfaces would be flat and non-pores. you would need no thermal grease.

    Metal to metal contact is the best. Thermal grease is made to fill in the microscopic pores for better heat transfer. Very little should be used.

    Installation torque on all semiconductors is also overlooked by almost everyone. The manufacturer will have a spec on his data sheet and usually around 4 to 6 inch pounds for consumer size stuff. That is snug not tight. The fact as most users "crank her down" and they still work is a credit to manufacturers quality of construction. To get an idea of 4 "#, hang a 1# weight 4" out on a ruler held horizontal.

    In general heatsinks are aluminum. Thermal conduction is what the idea is. Copper would be better, more expensive and heavier - looses in the tradeoff of slight thermal gain to $$ and # side. The spacers need to be metal. Have seen a vary few tinned copper in OLD stuff. Also spacers are generally avoided in the design phase, more parts more $.

    Make your bent one flat & smooth. Consider making a replacement if necessary. If it is magnetic it is ferrous.

    A can of freeze spray can help and use the tube if you can't feel the hot with your finger or you want a second check. Shoot the ICs one at a time and give it 30 seconds before moving to the next. Then on to other parts until you find the guilty party.

    edit: looking at your pictures more and noticing 2 mounting studs finally got my attention. Are the ICs. clamped between the plates and the heatsink? That would mean slowly tighten one end and then the other till you get to snug. If that bend in one plate is from over torquing so much to bend the plate the chip substrate may be cracked and even minor heating will cause it to quit.

    Looking like more reason to try freeze spray.
     
    Last edited: Apr 28, 2014
  10. KrisBlueNZ

    KrisBlueNZ Sadly passed away in 2015

    8,393
    1,266
    Nov 28, 2011
    I would definitely throw the old one out. You will never get it very flat, unless you have some kind of industrial press!
    Good point. If it's attracted by a magnet, and it's a dull grey colour and somewhat soft (easily gouged with pliers to a small depth), it's probably mild steel, I think.
     
  11. k7sparky

    k7sparky

    6
    0
    Mar 7, 2012
    Looking at the data sheet for the 50W output module TDA1514A it is interesting to note a mute function. As well as thermal and SOAR (safe operating area).

    The data sheet shows the SOAR input pin 2 tied to ground by a 470K resistor which could be, but not likely bad. that would be yellow violet yellow with a tolerance band probably gold for +/- 5% Who knows on your board compared to the data sheet. Looks like pin 2 tied to pin 3 for mute function.

    You also have some output feedback into pin 9 to balance against the input on pin 1 to trigger the mute after a period of time. Electrolytic cap from -Vp to pin 3.

    If freeze spray doesn't do it, try Ohming out the traces on the board for 2,3,4,9 for a cracked trace. Can be hard to see even with a magnifier.

    Also check the power supplies for +Vp and -Vp about the same absolute values on pins 4 & 6

    Your 50W power chip looks like a winner from the data sheet. I generally use http://www.datasheetarchive.com for data sheets. Lots of other good ones. I happened to grab the 7 page data sheet.
     
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

-