Harmon Kardon TS7 sub amp issue

[Journey11]

Hello All,
I was hoping to get some advice on an issue with a Harmon Kardon TS7 sub amp. Initially I had done a minor repair job on it for some broken solder joints ... one on the LFE input jack so that fixed that non functioning input. And then the whole amp out put went bad ( worked at first minus the LFE input ) so found some bad solder joints at Q105 and Q103. Fixed it all up and it worked for a little over a week then the two main power transistors blew. So replaced them and powered up ... it worked for all of 5 seconds and they blew again. I checked every other transistor on the board along with all the polarized caps for shorts ... almost all the resistors surrounding the main driver and power transistors. Couldn't find anything wrong with any other components. So I'm left scratching my head as to what caused the power transistors to blow. The only thing I could come up with was this.

When I repaired the bad solder on Q104, I noticed the base connection for Q106 had a long solder bridge between the pads so I got rid of most of that. So during the search for more bad components I noticed when I had Q106 out of circuit for testing, I noticed a break in the pcb trace going to R114. So then I thought that was the reason for the big solder bridge.
So with all that said ... my question to anyone who is familiar with type of amp design is this ... If the base connection (red arrow on pic ) of Q104 was open ... would it have caused an over current situation down the main output channel ? ( blue line in the pic ) The text in the pic describes how each transistor blew out so take that into consideration when/if Q104's base were to go open.

 

[Harald Kapp]

If the base connection (red arrow on pic ) of Q104 was open

You mean the connecztion of the collector of Q104 to R114, do you?
That would place a positive voltage on the base of Q106 which in turn would place a negative voltage on the base of Q108 and yes, if the Q107 is controlled to conduct you have that short circuit (blue line) as indicated.

When I repaired the bad solder on Q104, I noticed the base connection for Q106 had a long solder bridge between the pads

Depending on which pads were short circuited this bridge may have prevented the short circuit (e.g. if between emitter and base. The cost for this would have been distorted sound.

 

[Journey11]

You mean the connecztion of the collector of Q104 to R114, do you?
That would place a positive voltage on the base of Q106 which in turn would place a negative voltage on the base of Q108 and yes, if the Q107 is controlled to conduct you have that short circuit (blue line) as indicated.

Depending on which pads were short circuited this bridge may have prevented the short circuit (e.g. if between emitter and base. The cost for this would have been distorted sound.

Hi Harald ... thanks for the quick reply, much appreciated.
The connection from Q104 collector to Q106 base was good ... it is a big wide trace straight across. It was a thin trace going from Q106 base to R114 that was suspected (almost open in the 1st week of operation til it blew) to go open when I removed the large solder bridge between the two. I suspect that when I removed most of the bridge, a very small bit of solder was still making connection so the amp worked for about a week ... that is until I'm thinking the amount of current was large enough to break that connection ( acting like a fuse so to speak ) So when that happened it caused the blue channel to run wide open current wise.
So in your comments above you were assuming that the connection from Q106 base was open to the Q104 collector which was not the case that trace was good ... nice big wide one. Not sure if this will affect your analysis.

As for the large solder bridge I spoke about ... I believe this was done to make the connection between Q106 base to R114 good and solid. The bridge was not causing any shorts to anything near it. When I removed most of the solder I suspect I made that connection to R114 very very thin and it eventually melted what little solder making the connection so was pretty much like a fuse and somehow caused the blue channel to open up full blast.

 

[Harald Kapp]

So in your comments above you were assuming that the connection from Q106 base was open to the Q104 collector

Sorry if I wasn't clear enough: I assumed from your initial description that the connection to R114 was broken. Your arrow doesn't make this very clear:

 

[Journey11]

Ya sorry bout that ... you are correct ... that is where the connection would have been lost. So I installed a new set of power transistors yesterday afternoon and posted this to get some advice on it. I didn't want to power it up just to blow a fresh new set of transistors again. So your original assessment was correct then ?

 

[Harald Kapp]

So your original assessment was correct then ?

Sounds likely. Only a test can show - after you repaired the broken track

 

[Journey11]

KInda of exactly what I suspected but sometimes I get dislexic when it comes to transistor operation ... comes with age I guess LOL !

 

[Journey11]

I was thinking of using my variac to run a little bit of juice in the circuit to see if those power trannys start heating up before they pop ? Waddya think ?

 

[Journey11]

And thot maybe open up one of the 0R1 resistors and put an ammeter in there to see how the current is doing as I ramp up the voltage with the variac ...

 

[Harald Kapp]

I was thinking of using my variac to run a little bit of juice in the circuit to see if those power trannys start heating up before they pop ? Waddya think ?

Or use the light bulb trick. Better for limiting the current.

And thot maybe open up one of the 0R1 resistors and put an ammeter in there to see how the current is doing as I ramp up the voltage with the variac ..

Why not connect a millivoltmeter across the resistor instead of de-soldering it? 1 mΩ -> 1mV / 1A

 

[Journey11]

Ah yes the ol light bulb trick ... I tested those 0R1 resistors and they seemed ok but I bought new ones anyway. I built a mOhm setup to test those 0R1 resistors and they looked close to the rating. Problem was/is my meters are just not accurate enough down to that level.

 

[Journey11]

Just opened up an old QSC dual mono block for repair ... what an old boat anchor ... it's a wreck. Might buy it for 50 buck an old QSC series 3 3800 but before I buy it gotta see if can repair it.

 

[Journey11]

So Harald ... just to be sure ... where would I put that light bulb in the circuit to do this test. I've heard of the light bulb test but have never actually done it/used it ?

 

[Harald Kapp]

The light bulb comes into the mains power supply circuit, between mains outlet and power supply mains input.
It is meant to limit the power input to the amplifier.

 

[WHONOES]

Another horrible power amplifier design with no discernible thermal feed back. What you are probably experiencing is thermal runaway.
After repairing any broken tracks you may find and if you feel up to it, short the input connections together, place an ammeter in series with the + supply. Disconnect any load you may have on the output of the amp so that it is totally open circuit. If it is possible, place a large switch between the amplifier power supply and the amp' and also put a small value fuse say 250mA in series with each arm of the supplies. Whilst watching the ammeter, switch on. Make a note of the current draw if it is not excessive i.e. less that 50mA then leave on. Keep a watch on the meter. If the current increases over the space of a few minutes then you probably have the dreaded thermal runaway. Switch off before it becomes unmanageable.
In a well designed amplifier with the appropriate thermal feed back, I would expect the quiescent current to stabilise after a couple of seconds and perhaps diminish slightly as the unit warms up to normal.

 

[Journey11]

Another horrible power amplifier design with no discernible thermal feed back. What you are probably experiencing is thermal runaway.
After repairing any broken tracks you may find and if you feel up to it, short the input connections together, place an ammeter in series with the + supply. Disconnect any load you may have on the output of the amp so that it is totally open circuit. If it is possible, place a large switch between the amplifier power supply and the amp' and also put a small value fuse say 250mA in series with each arm of the supplies. Whilst watching the ammeter, switch on. Make a note of the current draw if it is not excessive i.e. less that 50mA then leave on. Keep a watch on the meter. If the current increases over the space of a few minutes then you probably have the dreaded thermal runaway. Switch off before it becomes unmanageable.
In a well designed amplifier with the appropriate thermal feed back, I would expect the quiescent current to stabilise after a couple of seconds and perhaps diminish slightly as the unit warms up to normal.

Hi WhoNoes,
Thanks for chiming in here. You say there is no thermal protection on this HK design ? Then I have to wonder what these two transistors Q101 & Q115 that are clamped down on the heat sink are for if not some sort of thermal detection ? Possible thermal biasing of some sort ?
The amp worked fine for many many years and the client brought it to me cuz LFE input had broken solder/connection. While fixing that after I reassembled the output went very low and distorted. This was when I found the bad solder around Q104 & Q106. After dealing with that it ran fine again for about a week and a bit then the main trannys blew.

With your thermal runaway angle ... I would have to think that there would have to be a trigger/reason for it to just do this out of the blue after years of operation w/o issue. For example a resistor in the drive section that has gone way out of tolerance ?
I measured almost all the support resistors in the drive section and all were good.
I did see some discoloration (on pcb) from heating in and around the other set of drive trannys Q103 & Q105. I checked both of these out of circuit and they seemed fine. However, it is possible these two might be the reason for the runaway if they are out of spec even though they test good.

Not sure if you read Harald's assessment ... if you did what do you think of his analysis and my suspicion that the base of Q106 had gone open from R114 ? There was a large solder connection there before I cleaned up the bad soldering around these two trannys. After re-inspection when the mains blew I found that thin trace was broken at one of the pads, hence I'm thinking why someone made a large bridge on that trace to make sure that connection was solid.

I like your ideas for testing the rig before I actually power up and see if the mains blow again. Already lost the 1st new set. Not sure how easy it will be and/or whether I have some fuses @ 250mA to run with. I'll see what I can do ... I have a variac and mentioned using that to slowy ramp up the voltage and watch the mV across the 0R1's ... what's your take on that approach ?

Thanks again for your input
Journey

 

[Journey11]

Problem solved ... looks like Harald was correct. I took advice from you both who responded with the testing phase and after a careful setup I was able to power up the amp without issue. I'm quite convinced the Q104 to R114 disconnect was the problem. Thanks much guys ! My client will happier than a pig in poop ... he's going crazy during this ridiculous lock down without his home entertainment MINUS the SUB ... I can relate !

Now on to the Bose L1 Model II PA system which is switching on and off ... anyone got a schematic ? Ha Ha LOL good luck with that eh ? Got the service manual minus schematics ... typical eh ...

 

[Harald Kapp]

Now on to the Bose L1 Model II PA system

Better open a new thread for this new issue, thank you.

 

[WHONOES]

Glad you got it fixed without too much issue.
Re the Bose L1. Just a thought, with out any schematic to see. It could have developed larger than allowed output offset voltage. If the amp has an offset monitor then it may be why it is switching on and off.

 

[Journey11]

Glad you got it fixed without too much issue.
Re the Bose L1. Just a thought, with out any schematic to see. It could have developed larger than allowed output offset voltage. If the amp has an offset monitor then it may be why it is switching on and off.

Interesting thought .... are you familiar with this system ? Worked one on before ... If I don't have any luck ... I'll open a new thread on this one.