Harman Kardon HK6500, bias too high

elnaldo

Lunatic Member
Hello. Using the trimpot, the minimum is over the service manual specs. Should I simply "cheat" the circuit, and add more resistance in series with the trimpot? The manual calls for 40mV across 5.4 ohm (2 x 2.7 ohm emitter resistors), and I'm getting around 50-60 with the trimpot all the way to one side.

All the transistors are originals, all the soldering was reflowed (it was 600mV before the resoldering)


capture26417.png
 
Is the 60mV bias the same on both sides?

The schematic has some nice voltages marked throughout. I'd be voting for a basic
health check just to make sure all is in order.

Probably means Q429 or someone down stream is not that healthy.
Maybe measure Q429 bce and Q441 b and Q443 b should give you a quick idea.

A check of the power supply voltages is also suggested, but you know this.
 
I'd also suggest checking the fusibles - some may have gone high.

In any case, a problem like that with idle current calls for troubleshooting, not 'cheating', in my opinion.
 
Thanks.

The amp had a lot of cold solder points, one channel blew because of this (only one channel is installed now).

My 1st test was with a DBT, OK for the1st seconds, and after 6 o 8 seconds the bulb started to glow a bit more. So I suspected some transistor starting to conduct more. Checked DC offset, OK. Attached the multimeter to monitor bias, and the problem was there. After some seconds started to raise too much (where supposed to be 40mV, raised to 600mV in a while. I powered off, and after an inspection, I discovered a lot of bad soldering. Retouched all the board, and the next test gave around 100mV, and moving the trimpot it went down to approx. 60mV

The resistors I measured check OK, I didn't check every single one, just the fuse resistors around the bias and driver transistors.

I'll check voltages again and replace the bias transistor just in case it was damaged under some stress due to the broken solder.

One detail: Q429 is not mounted on the heat-sink, but on the output transistor body. It's factory installed that way.
 
I'd also suggest checking the fusibles - some may have gone high.

In any case, a problem like that with idle current calls for troubleshooting, not 'cheating', in my opinion.

I've seen some service manuals where you are instructed to cut a wire bridge if you can't adjust the bias current, that's why I was thinking in this route. I'll double check the voltages, even replace the bias transistor, and see how it goes.
 
OK, at Q801, the differerences are little with the schematic but the voltages are different:, being the largest difference at the Base, expoected 60mV, measured 185mV

Schematic / In Circuit
C: 0.9V / 0.95V
B: -0.06V / -0.185V
E: -0.5V / -0.85V

Resistors are check almost perfect. Bias network, at max, measures 688ohm instead of 720 ohm (220 + 500 trimpot). Could it be that small difference what I'm needing?

Edited: setting the unit to 240V instead of 220, the voltages match better the schematic, but still 135mV at Q801 base, and 70mV across the 0.54ohm emitter resistors instead of 40mV.
 
Last edited:
Another update: Replaced Q801, used a spare MJE340 I had around, and now I can adjust at 40mV, even 37 or 38mV.

It seems Q801 was not OK, or shifted too much from the original specs.
 
@elnaldo,

Sorry to bring this thread up again, but I'm curious also about the sound of this unit. I see it has good transistors, like 2SA872 / 2SC1775 and others. Do you still has this unit and how does/did it sound? The HK6550 appears to be a similar model (if anyone can confirm).

Thanks.
 
@elnaldo,

Sorry to bring this thread up again, but I'm curious also about the sound of this unit. I see it has good transistors, like 2SA872 / 2SC1775 and others. Do you still has this unit and how does/did it sound? The HK6550 appears to be a similar model (if anyone can confirm).

Thanks.
Hello. I can't tell you about the sound, I've fixed the amp and tested it with my test speakers but I didn't compare it to other amps.
 
Back
Top Bottom