(right side)
vee- = -46.2v
vcc+ resistor462 = +47.3v
q446 readings
base -14v
collector -28.6v
emitter -14.6v
(left side)
vee- = -46.2v
vcc+ resistor462 = +47.3v
q446 readings
base -14v
collector -28.6v
emitter -14.6v
Like above, check with VR402. Strange that it first works a bit and then goes bad but that could be thermal related as that VR402 warms a bit.
Here's what I would do:
Warm up the amp and measure for the voltages on the different test pins in the alignment procedure on page 3 of the service manual. Big chance they are out.
If so, I'd follow the alignment procedure and do a long test with a multimeter hooked up to the speaker output (esp. on the channel that showed the jump).
The reason for this is to rule out other components than the potentiometers or cracked solder joints as a possible cause.
If the DC offset were to return after the alignment procedure, you have some more searching to do, but no need to worry about that now.
Warm up the amp and measure for the voltages on the different test pins in the alignment procedure on page 3 of the service manual. Big chance they are out.
If so, I'd follow the alignment procedure and do a long test with a multimeter hooked up to the speaker output (esp. on the channel that showed the jump).
The reason for this is to rule out other components than the potentiometers or cracked solder joints as a possible cause.
If the DC offset were to return after the alignment procedure, you have some more searching to do, but no need to worry about that now.
Guys ,In your opinion, what is the possibility that the two DC offset potentiometers (left and right) fail in the same way and at the same time?
maby q404 406 is the problem ? aren't they the ones who spans up the circuit and are called differential pair transistors ?
Rather large I'd say. Both components are the same batch of parts bought and used. They both have the same time and cycles of power on/off. They both got the same temperatures, they both are subject to the same vibrations. The chance is they are both fail at a similar time is thus real. And since they are obviously of the same construction it will also be the same mode of failure.
Probably that 4V and a bit offset is what the amp has if that potmeter is defective. So if they both fail then that one goes up and the other down in polarity is probably due to tolerances of other components.
From the Microchip AN219 application notes:
It is my experience that potentiometers can exhibit not only temporary drift from setpoint, but also permanent. By how much depends on the type and construction. A sealed multiturn point will be way better than the typical exposed single turn pot found in consumer devices. These can have a specification of a max of only 20 (yes, twenty) full turns before end of life, and while adjusting jumps of up to 5% are to be expected (the so-called "CRV" Contact Resistance Variation).
So I'd say offset drift could be caused by the pots, but may not be the actual cause in this particular case. That still needs to be determined.
It is my experience that potentiometers can exhibit not only temporary drift from setpoint, but also permanent. By how much depends on the type and construction. A sealed multiturn point will be way better than the typical exposed single turn pot found in consumer devices. These can have a specification of a max of only 20 (yes, twenty) full turns before end of life, and while adjusting jumps of up to 5% are to be expected (the so-called "CRV" Contact Resistance Variation).
So I'd say offset drift could be caused by the pots, but may not be the actual cause in this particular case. That still needs to be determined.
But on on both channels... there has to be something else going on here.
Thanks for the voltage readings earlier, nothing obvious there though 🙁
What about this:
It is a valid and important question.
In post #1 you said:
Not sure what you mean by that but if you have been replacing parts then perhaps the offset voltage is actually instability and oscillation that shows as a DC shift in level.
Mathematically this kind of error in equal value but in opposite directions could be caused by a failing transistor (or resistor) in the parallel pairs of Q422/426 and Q424/428.
I would measure (in both channels, when DC is present on speaker output) the voltage over their emitter resistors R444/446/452/454 - they should be conducting in equal manner (at least in a pair and also in both channels).
With power OFF, mark the position of the pots and then move them a few times. return them to the marked position and measure the output voltage, again.
There is no danger in moving the pots a few degrees with the amp on.
These amps are quite easy to repair by measuring the cirquit transistor by transistor, starting at the amps input, but, I'm sorry to say this, have a limited life time IMO. HK did not use the best capacitors, they are often a cause for multiple problems. It is not unusaual for them repairing and carefully adjusting them, to fail again, just a few operating hours later.
Even as I like the HK design, I wouldn't buy one.
Usually, with a scope and some sine wave at the input, you should find the fault in a few minutes. If it has been repaired or modified before, things look different. Heavy parts often have broken solder joints, a good idee to refresh them before doing other analysis.
There is no danger in moving the pots a few degrees with the amp on.
These amps are quite easy to repair by measuring the cirquit transistor by transistor, starting at the amps input, but, I'm sorry to say this, have a limited life time IMO. HK did not use the best capacitors, they are often a cause for multiple problems. It is not unusaual for them repairing and carefully adjusting them, to fail again, just a few operating hours later.
Even as I like the HK design, I wouldn't buy one.
Usually, with a scope and some sine wave at the input, you should find the fault in a few minutes. If it has been repaired or modified before, things look different. Heavy parts often have broken solder joints, a good idee to refresh them before doing other analysis.
There's something wrong in the schematic in post 5--- it must be a "simplified" drawing.
Look at Q420 and its two resistors in series to the base. This circuit can't work, as Q420 would never bias on. So I went to OP's original post and downloaded PDF and found some excised muting circuitry on pages 27 and 29. The full circuit disables the entire bias to the PA stages. I'm not saying this is the culprit behind the strange symptoms, but it might be and certainly deserves scrutiny.
I still advocate looking at the bias pot's adjustment range and any erratic sensitivity to rotation. Measuring the PA output and base voltages at Q404 and Q406 can give good insight to bias behavior.
Look at Q420 and its two resistors in series to the base. This circuit can't work, as Q420 would never bias on. So I went to OP's original post and downloaded PDF and found some excised muting circuitry on pages 27 and 29. The full circuit disables the entire bias to the PA stages. I'm not saying this is the culprit behind the strange symptoms, but it might be and certainly deserves scrutiny.
I still advocate looking at the bias pot's adjustment range and any erratic sensitivity to rotation. Measuring the PA output and base voltages at Q404 and Q406 can give good insight to bias behavior.
Last edited:
I don't understand the trimming circuit, there are resistors pulling the inverting input down, nothing pulling it up, does it always assume Vbe difference only goes one way?
Assume for a moment that the bias pot and the resistors to -14V are not present.
Bias current into Q404 will pull its base to a negative voltage via R414, a relatively large resistance (22k). in contrast, the bias resistance to Q406 is much lower, less than 500 ohms, so little voltage will be dropped compared to R414. Feedback will drive Q406 base voltage to the same negative base voltage developed at Q404. The PA output has to be negative as a result. Then correcting current from the bias trim path compensates.
Bias current into Q404 will pull its base to a negative voltage via R414, a relatively large resistance (22k). in contrast, the bias resistance to Q406 is much lower, less than 500 ohms, so little voltage will be dropped compared to R414. Feedback will drive Q406 base voltage to the same negative base voltage developed at Q404. The PA output has to be negative as a result. Then correcting current from the bias trim path compensates.
I have measured around q420 and q419 which is on the opposite side
q420
base -42v
collector -43v
emitter 42.8v
q419
base -43.5v
collector -43.5v
emitter -43v
"opposite side" means "the other channel"?
Emitter voltage for Q420 is positive - typo?
To make it easier to follow here is the schematic of that area:
What are the voltages on Q433 and Q435 bases and what is the voltage difference between them (in both channels)?
yes i mean the other channel.
o sorry my bad q420 emitter is -42.8 typo not +
these are the measurements... i think something might be wrong around q436
(Right channel)
[q435]
base -8mv
collector -45.3v
emitter -5.16v
[q433]
base -3.15v
collector +46.1v
emitter -3.66v
(Left side)
[q436]
base 2.9v
collector -46v
emitter 3.5v
[q434]
base 5.42v
collector 45.3v
emitter 4.8v
o sorry my bad q420 emitter is -42.8 typo not +
these are the measurements... i think something might be wrong around q436
(Right channel)
[q435]
base -8mv
collector -45.3v
emitter -5.16v
[q433]
base -3.15v
collector +46.1v
emitter -3.66v
(Left side)
[q436]
base 2.9v
collector -46v
emitter 3.5v
[q434]
base 5.42v
collector 45.3v
emitter 4.8v
Correct me (anybody) if I am wrong but the base voltages of these transistors are usually symmetrical (to ground) i.e. have roughly same values but with opposite polarity?
If so then there are problems in both channels.
If so then there are problems in both channels.
If there would be no DC at the output then very likely so. But we know there is something wrong because there is DC offset.
Has those offset potmeters been checked yet? If so, what was the result?