Connect with us

power amp advice sought

Discussion in 'Electronic Repair' started by Dan K, Jul 26, 2006.

Scroll to continue with content
  1. Dan K

    Dan K Guest

    Hi group

    I have a 240w single channel solid state transformer coupled power amp that
    isn't working. There is not a whole lot to the power amp board, but I do
    not have a matching schematic. I did find a schematic to a more recent
    version of this amp and its close enough so I was able to figure out the
    main power path, but not the front end, feedback, and protection circuitry.
    It looks like the company re-designed the front end, feedback, and
    protection circuitry on this later model. The power amp board had 2 open
    transistors, 3 shorted transistors, and showed signs of burns. The amp blew
    fuses when turned on.

    I replaced all the bad transistors and an open resistor (burned). I
    reinstalled the large heat sink that connects to the two power transistors,
    the two pre-driver transistors, and the bridge rectifier. I verified that
    all transistors connected to the heat sink were electrically isolated from
    the heat sink and powered the amp up. It blew a fuse.

    I removed the heat sink and the power transistor pair and powered things up.
    I found very high voltages throughout the output stages (like around 60 vdc)
    with a large amplitude distorted signal riding on it that was the input
    frequency I was feeding into the amp. The amplitude of this distorted
    signal followed the volume control. I assumed that this particular amp
    cannot be run w/o the power transistors, so I put them back in. I powered
    the amp back on and it worked. The only difference being I was running
    without a heat sink now.

    There is a fan in this amp. One fan wire is connected to a ground wire that
    is connected to the power supply. These two wires are connected via a screw
    that screws into the heat sink. I found that as soon as this ground from
    the power supply is connected to chassis ground (via the heat sink
    originally) I lost all audio (but did not blow a fuse or apparently damage
    anything from the momentary contact). Anyway, I was running this
    heat sink, fan wire not grounded, just audible output...thinking about what
    I should do next when snap, crackle, pop I had flames shooting out of the
    board. Both pre-driver transistors are physically cracked, I presume the
    power transistors are shorted, a couple of resistors have burned up, and the
    pc board really shows burn signs now. So, here's my question:

    Could this failure be due to just running for 3-4 minutes at extreme low
    volume without the heat sink? I would guess not, but I don't have a lot of
    experience with class B power amps. If it could be, it might be worth
    replacing the transistors and trying again, otherwise I think the garbage is
    the best place for this amp.


  2. Arfa Daily

    Arfa Daily Guest

    Refailure of this kind is common. You really need to be able to run the amp
    up on a variac to see what's going on without risk of re-frying all the
    parts that you have replaced. It is possible that the transistors could have
    failed as a result of having no heatsink on them, but probably only as a
    result of something else still being wrong. With the amp running at just
    above idle, the output stage dissipation should be minimal, but not if there
    is a problem with the bias setting, if this is variable, or bias supply if
    fixed. Once the transistors start to get hot, they will quickly go into
    thermal runaway without adequate heatsinking, and rapidly fail.

    If you can lay hands on a variac, it might be worth having one more go, but
    after that, without proper schematics, you will probably struggle to get to
    the bottom of it, and finish up throwing good money after bad

  3. jakdedert

    jakdedert Guest

    It might also be useful to post the brand and model # of the amp.
    Someone might have experience with--or be able to supply a schematic

  4. Dan K

    Dan K Guest

    Its a CSI P-240A, which is from Speco.
  5. GregS

    GregS Guest

    Its possible that you heated the parts. Try again?

    I ALWAYS use a light bulb in series with power amp testing.
    60 - 200 watt or so. I sometimes also use a variac in
    conjunction with the light bulb. That would not have cured a heat
    problem if that fried the amp, but it may have prevented sparks and

    I laughed, sort of, after fixing a brute of complexity of a Crown
    amp. After "fixing" it, I sat it on the floor playing some soft music,
    and started walking upstairs, heard a pop, looked around at the bright
    light comming from inside the power amp, then some smoke, and
    that was that.


  6. default

    default Guest

    Like someone already suggested - a variac is necessary when fixing
    solid state amps. Monitor current and keep an eye on the output stage
    to see that it comes up with the voltage you expect. I lack
    experience with transformer coupled amps (except a few old Sansui amps
    from the 60's)-

    If the bias is adjustable set it to minimum until the full voltage is
    present then set it to some reasonable value. (like 20-30 watts for a
    250 watt amp)

    Class B or Class AB? I have a pair of 500 watt amps - class AB and
    they work class A up to about 100 watts and above that start working
    class B. (24 output transistors and massive fan cooled heatsink)

    It definitely needs a heatsink if it is class AB. Without the sink -
    the output stage gain increases dramatically when the transistors warm
    up - the bias current tracks the gain of the output stage - unless
    there's some thermal feedback to control it. (old designs usually
    had a pair of diodes, thermister, or bias transistor that would mount
    to the same heatsink as the output transistors)

    I've seen amps with two pairs of Darlington transistors in the output
    - and a small thing like too little heat sink compound on one will
    cause it to overheat.
  7. default

    default Guest

    PS - old transformer coupled Sansui amps had a pair of series
    connected bias diodes in a molded plastic case with a bolt hole for
    mounting. That was the thermal feedback system for that amp. The
    diodes would develop an intermittent open connection and an otherwise
    perfectly working amp would crackle the speakers and start blowing
    fuses and transistors.
  8. I powered the amp back on and it worked. The only difference >being I was

    Perhaps the output transistors weren't insulated from the heat sink?

    Mark Z.
  9. Arfa Daily

    Arfa Daily Guest

    That was my first thought also, Mark, but if you go up 4 posts, you will see
    that he makes a point of saying that he checked that the devices were
    isolated from the heatsink. It's a mystery ... |:-{

  10. Could have a damaged insulator...

  11. Arfa Daily

    Arfa Daily Guest

    Could be. Might just be the way we are interpreting " checked ". When I
    check that devices are isolated from a heatsink, I prove it with a meter
    reading. Perhaps he was just visually inspecting ...

  12. Dan K

    Dan K Guest

    Nope, meter reading. Isolated. It wouldn't have worked for a while if they
    wasn't isolated, Vee would be shorted to Vcc. Could always be a crack in
    the insulator I suppose that only shorts at certain times...They look ok

  13. Arfa Daily

    Arfa Daily Guest

    Yeah, it kinda sounded like you had a done a proper test when I first read
    it, and you are of course right that it wouldn't have worked ok for a
    while - I'd forgotten that bit ...

  14. jakdedert

    jakdedert Guest

    Sometimes you'll get a little 'burr' of sharp metal around the holes in
    the heatsink. It reads fine when cold, but when it heats up the burr
    expands enough to short through the insulator.

    Clean off the sink, and feel for any sharp points.

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