Harman Kardon SUB-TS15 Powered Subwoofer - Repair Help

[plat27265]

Greetings to everyone! 

I have undertaken the task of helping a friend with his Harman Kardon SUB-TS15 Powered Subwoofer that is in pieces.  He attempted to fix it or have someone fix it and it was left in pieces.   He also mentioned that it was not working and something was burned on the board.   

Before anything else, I am attaching a PDF copy of the sub's schematics so readers can follow me and offer some advice to a nagging problem that I have encountered.

Upon close inspection of the Power Amp board, I noticed that R146 (330-ohms, 2-watts) got burnt to a crisp along with a dislocated ground wire.   R146 is part of the +15V regulated power supply circuit.   I replaced it and repaired the burnt copper pad and reconnected the ground wire to the ground trace on the same board.   

I have identified several electrolytic caps that have either very high ESR and capacitance values that are out of tolerance.  I have replaced all the electrolytic caps (except for the large PSU filter caps) with new ones on all boards (power amp, pre-amp, and limiter board).

On two of the limiter board, I noticed dried up adhesive that has become conductive.  When I tested the adhesive on the limiter board, I measured 8 Megaohms.  I had to scrape everything off and I will apply hot glue once I confirm that everything is working properly.

I plugged in the fully assembled plate amp to the wall outlet with no input source.   I heard the speaker relay turn on.   This is a good sign.  I connected my multimeter to the speaker terminals and I measure no DC voltage.  Another good sign.  I turn the amp off.  Then, connected the subwoofer driver.   When I turned the amp on, I waited for a second for the speaker relay to engage.  When it did, I heard a loud thump then it was followed by another one and another one in succession.  It won't stop.  I turn the amp off and connected my multimeter across the speaker terminals and set it to read AC voltage on peak hold mode.   When I turned it on again, I heard the loud thumping sound.  The loud thumping happens around 0.5 to 1 second intervals.   The multimeter reads a peak initial turn on thump of 18-volts.    Succeeding thumps were recorded at 12 volts.

Note:  Everytime the thumping occurs, I can hear the speaker relay go on and off.  Everytime the speaker relay engages, the thump occurs.  The amp probably sensed an overload condition and turns off the relay.  It turns it on again, voltage spike goes to the speaker at 18V, turns off again. 

On the next test, I disconnected the negative terminal from the speaker and turned on the plate amp.  I heard the power relay turn on and then connected the negative terminal back to the speaker.  I hear nothing, which is a good sign.   I then proceeded to turn the amp off again.

On this next test, I connected my iPOD to the low-level inputs of the plate amp.   Set the power mode to Auto ON (or just ON, same symptoms occur).   No music playing yet.   Disconnect the negative speaker lead from the subwoofer driver.   Turned on the plate amp.  Hear the speaker relay click.  Then, connect the negative speaker wire to the subwoofer driver.   Silence, no thumping.   Hit play on my iPOD and the subwoofer works.  I left it playing for an hour, no issues.

During the test, I proceed to turn the power switch of the plate amp off then back on.  No loud turn on thump, just a soft one which is what it should be.  Repeat this action several times, all seem to work well.

All the tests mentioned above happened on one bench test session.

The next day, I started off where I left the test.  Everything connected including the music source.  Turn the plate amp on.  The loud thumping is back.   I disconnect one speaker wire from the subwoofer driver then connect it again.  No more thumping and I can hear the music source playing.  I leave the music playing for a while then suddenly, the loud thumping starts again.   

I am at a loss on where to look for the problem.  Based on the tests described above, the symptom manifests itself when the amp sees a large load connected to it immediately after the speaker relay turns on after the power switch of the plate amp turns on.  The symptom can always be observed to occur this way.   I said large load because when I connected a dummy 8-ohm, 5w resistor across the speaker terminals, I could not read any momentary voltage spikes across it.   It seems to react to an inductive load.   To stop the thumping, I have to disconnect the subwoofer and connect it again.  In most tests, the thumping does not re-occur and sometimes it does after a long while with or without a music source playing.

I am hoping that more experienced techs on this board can point me towards the right direction or give me some advice on where to start looking for the cause of this problem.

My initial suspect is the speaker protection circuit that drives the speaker relay.  I tested all the semiconductors in-circuit and all of them seemed fine.  But, what causes the power amp to output such a large 18V spike upon seeing a speaker load at its output and just continue doing so until the subwoofer is disconnected and reconnected.   

 

   

[plat27265]

Further observations I took today:

On Power ON with speaker connected:  DC Offset measured at speaker output (On Turn ON Thump):  79mV.
AC Reading:  anywhere between 4V to 18V during each thump.

DC Offset Measured if speaker is connected to amp output after power ON:  1.2mV

What could be the cause of the momentary spike in DC voltage  during the thump?  That's a very high value.

[floobydust]

The protect relay circuit trips on DC offset Q112/Q111 or over-current Q109 or high temperature Q115 sensor or Standby mode (diode by D108/Q102) or loss of AC mains Q114.

I'd guess something causes a DC offset/thump and based on the fact that the woofer causes troubles but not an 8 ohm load, I think next the over-current protection is triggering. It's a side-effect from another problem so I would not go after it. I would put a large resistor in series with the loudspeaker say 100R to protect it against the thumps yet hear what's going on.

Are the +/-15V rails OK? If R146 smoked then Q117 might be damaged.
Next I would work back and observe U101 pin 1 and it should always stay around 0V, then go to the op-amp outputs and see what they are doing for DC voltages (with quiet music play if you want).

At this point hard to know if the unit is thumping going in or out of Standby mode or there is a DC drift and offset from a bad part. Have to hunt that down. SW203 is ON, not AUTO? Look at the voltage on Standby STB.

The subs have very high vibration and with age some wires and connections fatigue. I would also look for fractured soldering joints or tap around with a plastic pen to find a bad connection.

[plat27265]

The protect relay circuit trips on DC offset Q112/Q111 or over-current Q109 or high temperature Q115 sensor or Standby mode (diode by D108/Q102) or loss of AC mains Q114.

I'd guess something causes a DC offset/thump and based on the fact that the woofer causes troubles but not an 8 ohm load, I think next the over-current protection is triggering. It's a side-effect from another problem so I would not go after it. I would put a large resistor in series with the loudspeaker say 100R to protect it against the thumps yet hear what's going on.

Are the +/-15V rails OK? If R146 smoked then Q117 might be damaged.
Next I would work back and observe U101 pin 1 and it should always stay around 0V, then go to the op-amp outputs and see what they are doing for DC voltages (with quiet music play if you want).

At this point hard to know if the unit is thumping going in or out of Standby mode or there is a DC drift and offset from a bad part. Have to hunt that down. SW203 is ON, not AUTO? Look at the voltage on Standby STB.

The subs have very high vibration and with age some wires and connections fatigue. I would also look for fractured soldering joints or tap around with a plastic pen to find a bad connection.

First of all, I would like to thank you for spending time to read what I wrote.  That was long.  I wanted to be detailed as possible so readers can follow and share my anguish (ha, ha, ha!)

The +/- 15V rails are fine (I think) but they're not exactly the same.  One is a fraction more than the other.  If I remember correctly, the (+) side is like 16.2 V and the (-) side is like 15.85V.   This comes across to me as something insignificant.  However, could this be a clue?

R146 got burnt but it just burned on the outside.  When I pulled it out and tested it, it was still OK but I replaced it.   C107 (Metallised Polyester - 0.1 uF/50V) is physically located beside R146.  When it burned, it blackened the section of board around it and C107.  It got hot enough to lift the copper trace that connects R146 to the collector of Q117.   I have tested C107 in-circuit using an ESR meter and I am getting a capacitance reading that is way off is rated value.  I am thinking that the heat has changed its electrical properties.  And, the ESR is very, very high.  I have read that the ESR on these plastic film caps should be very low.  I am going to pull it out and test in out of circuit.   Q117 seems fine.  Otherwise, the entire amp would not be working normally when I connect the speaker after turning the power ON. 

I will investigate for bad (fractured/dry) solder joints.  But, I am also going to look into C107 because of its proximity to R146 on the board (they're right beside each other).  My fear is that the heat that caused the PCB around these two including Q118 to blacken could be hot enough to alter the electrical characteristics of C107 which is a plastic film capacitor.   

C107 and R127 are tied across the speaker terminals.  Could C107 be leaky or out of spec to allow such a massive thump (voltage spike and DC offset)?   Testing in-circuit using an ohmeter tells me it seems OK.  I'll take it out of the board and test it.

Thank you for the tip about connecting a 100 ohm load across the speaker terminal to absorb that voltage spike (thump) while I go through the signal chain on the main amp board and find out the source of the thump.

I will report back with my findings.


 

[plat27265]

Just want to share with everyone a quick update concerning this thread.   

The good news is that the subwoofer is fixed and working as it should.  I tested it for hours and at varying volume levels.  It works fine now.

The culprit is some left over original adhesive that has become conductive over time (3 megaohms).   I thought I had thoroughly cleaned the adhesive found on the preamp board at the base of the limiter board.   I have missed a spot and it is straddling the -15V rail (jumper wires on the board).   Once removed and cleaned, the subwoofer no longer acted up as described above.

Cheers!