Audio amplifier - Dc on output

Discussion in 'Analog & Mixed-Signal Design' started by pwnstars, Sep 23, 2016.

  1. pwnstars

    Thread Starter Member

    Feb 29, 2016
    44
    1
    Hello Guys..

    Im trying to fix an audio amplifier but im not sure about the trouble shooting process.

    I've measured 2.5V DC on one of the channels.

    Any suggestions on where to go from here would be appreciated.
     
    Last edited: Sep 24, 2016
  2. AlbertHall

    Well-Known Member

    Jun 4, 2014
    1,940
    383
    What is the problem with the amplifier? And on just one channel or both?
    What does the other channel measure?
     
  3. pwnstars

    Thread Starter Member

    Feb 29, 2016
    44
    1
    On the other channel it measure 20mV DC which is just fine.. What i want to figure out is why I measure DC on the output.
     
  4. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    U are getting the DC offset at the output.
    Good design will have a few mV at the output.
     
  5. AlbertHall

    Well-Known Member

    Jun 4, 2014
    1,940
    383
    Measure the supply voltages.
    Measure T14 and T16 collector voltages.
     
  6. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    Go back through R34 (47k) to the T10-T11 pair. Base of T11 must be nearly zero volts if it is properly matched with T10. I have seen this happen in real life. The differential pair doesn't match well enough and the DC offset is amplified, but this DC feedback loop doesn't have a proportioning resistor to ground at the base of T11.

    If the base voltages do match, look for DC leakage in C18

    Trust me, I'm good at this. If the differential pair isn't nulled, the output voltage is going to be stuck to the power supply voltage.
    This partial drift tells me C18 is leaky.
     
  7. pwnstars

    Thread Starter Member

    Feb 29, 2016
    44
    1
    Thanks for the help.. I will do the measurements on monday and return with the result :)
     
  8. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    783
    The TS says 2.5V - I'd say that's a bit much. On a 4R speaker - that's over 1/2 A.

    Could be a leaky output transistor, a possibility that should be eliminated before it goes critical and does a lot of damage.

    There should be a DC nfb resistor - the nfb should null out most of any DC offset, that resistor going high or open can cause problems.

    It should be possible to short the bases together on the input LTP - any differential between the collectors should be investigated.

    With closed loop DC amplifiers; the textbook fault finding procedures don't always work, sometimes you have to lift components and test them - the workload can be reduced by making informed choices.
     
  9. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    2.5V was not offset. It is a fault
    The 20mV channel was the offset
    I should have pointed that out.

    Sorry..My Bad.
     
  10. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    783
    I think most people would consider 1/2A through the speaker a fault - wouldn't you!
     
  11. pwnstars

    Thread Starter Member

    Feb 29, 2016
    44
    1
    It seems like i measure 2.5V on T15 basis which should be just about zero. If i could adjust that down the output should be adjusted too. Unfortunately nothing happens when i turn the potentiometer i've replaced it and still nothing happens.
    Not sure how to force it down.
     
  12. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    Yebbut...what's the voltage on the base of T11?
    If that isn't zero, the DC offset is going to transfer right across to T14 then T15.
    The feedback path through R24 is dominant in this circuit.
    An error on the collector of T11 can be caused by imbalance in the T10-T11 input pair.
     
  13. pwnstars

    Thread Starter Member

    Feb 29, 2016
    44
    1
    Base of T10 and T11 is -2.5V aswell.
    Whenever i've said 2.5V its actually -2.5V sorry for that.

    And i measure the same output on T10 and T11.. So there should be no imbalance
     
    Last edited: Sep 26, 2016
  14. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    Well, how did you get -2.5V on the base of T10?
    If you have -2.5V on the base of T10, the rest of the amplifier isn't broken. It's doing exactly what it's supposed to do..."deliver to the output the voltage which is on the input".
     
  15. pwnstars

    Thread Starter Member

    Feb 29, 2016
    44
    1
    The input (T10 base) isnt connected to anything.. Its being pulled down by T12. And i think that's due to an error in the in the T14, T15, T16 area that courses the base of T11 (and output) to be pulled down.
    hmm..
     
  16. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    T12 isn't designed to pull the voltage down, it's a constant current source. That means it's collector voltage is the dependent variable. It depends on where the emitter of T10 is, and that is 0.6 volts below the base.
    Put R11 back in and thus connect the base of T10 to zero volts.

    T12 is going to keep sucking down current until it gets the 1 ma it needs for the 5.6k resistor.
    Think about why T12 can't get 1 ma of current for its collector.
    The collector currents of T10 and T11, both added together, are supposed to provide that 1 ma.
     
    Last edited: Sep 26, 2016
  17. pwnstars

    Thread Starter Member

    Feb 29, 2016
    44
    1
    But the T12 collector voltage is also equally dependant on what's going on with T11 emitter.
    If i have -2.5V on T11 base, the emitter will be pulled down to -3.1V and that will cause the input (T10 base) to be pulled down to -2.5V aswell. Isnt that correct?
     
  18. pwnstars

    Thread Starter Member

    Feb 29, 2016
    44
    1
    True.. And that is approximately what is measured. 0.5A passes R12 and R13
    0.5A * 3.3k = 1.65V
    And i measured 1.61V on both.
    So that is part is correct.
     
  19. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    If the base of T11 is pulled down to -2.5V while the base of T10 is connected to zero through R11, T11 will shut off and zero current will flow through R13. That would make the base of T14 go high and there should be zero current through T14,15,16

    Is the T14,15,16 string shut off?

    ps, you meant 0.5 ma when you typed this:
     
  20. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    'Scuse me. I have to go see my drug dealer before the pharmacy closes, and buy food. This is gonna take about an hour.
     
Loading...