Current draw 0.38A they are all 0.68ohm resistors.
And this draw only with one channel at 15mV Vdrop, another 5 .3mV.
And this draw only with one channel at 15mV Vdrop, another 5 .3mV.
This does not make much sense. How can the current draw change by 18% when the bias current changes by nearly 300%?
Is there any chance something is wrong with the measurement?
Judging by the mains current draw and the heatsink temperature the amp is indeed overbiased.
Is there any chance something is wrong with the measurement?
Judging by the mains current draw and the heatsink temperature the amp is indeed overbiased.
The amp should not exceed 80W of power draw when idling. Taking into account the heatsinks are not external, 50 deg C also seems like a sensible limit or even 45 deg.
None of this explains why your amp pulls more than 70W from the wall when the idling current is just 8mA. 8mA x 20 x 75 = 12W. There is of course additional current draw from the drivers and pre-drivers, so let's say another 5W. Together with transformer losses i cannot see more than 25W of power drawn from the mains, yet it appears yours is several times higher.
I can offer no explanation other than wrong measurements or wrong resistor values. It is perhaps worthwhile measuring the actual resistance and taking measurements across all the emitter resistors, not just one.
None of this explains why your amp pulls more than 70W from the wall when the idling current is just 8mA. 8mA x 20 x 75 = 12W. There is of course additional current draw from the drivers and pre-drivers, so let's say another 5W. Together with transformer losses i cannot see more than 25W of power drawn from the mains, yet it appears yours is several times higher.
I can offer no explanation other than wrong measurements or wrong resistor values. It is perhaps worthwhile measuring the actual resistance and taking measurements across all the emitter resistors, not just one.
As Analog says , measure each ER. Also try and measure the temp on each OP device separately.
Also, remove fuses of one channel, then bias the active channel and check current draw at mains. The do the reverse for other channel ( ie check current draw fire each channel independantly). It might indicate an anomaly....
Also was this amp repaired? Maybe some OP’s are fake, they might draw more or less current.
Also, remove fuses of one channel, then bias the active channel and check current draw at mains. The do the reverse for other channel ( ie check current draw fire each channel independantly). It might indicate an anomaly....
Also was this amp repaired? Maybe some OP’s are fake, they might draw more or less current.
I think amp after repair, i find only that some transistors was changed.
But only in one channel.
But i try measure both channels separately and measurments are identical.
Only do not try to check current draw separately - also i noticed thats secondaries from power supply was connected wrong there are 4 sec. Wire, two green and two yellow, i found they connected vise versa but i fix this around one month ago.
But only in one channel.
But i try measure both channels separately and measurments are identical.
Only do not try to check current draw separately - also i noticed thats secondaries from power supply was connected wrong there are 4 sec. Wire, two green and two yellow, i found they connected vise versa but i fix this around one month ago.
So channels separately, draw identical 0.18A
One channel resistors from 3.4 to 6.7 other channel from 4.2 to 5.8.
One channel resistors from 3.4 to 6.7 other channel from 4.2 to 5.8.
Last edited:
That's a 38% variation in current draw for the devices on one channel (4.2mV to 5.8mV... is this the side that was never fixed?? ) and close to 100% variation on the other channel !
On the bad channel, (assuming this was fixed) 3.4mV to 6.7mV suggests a problem. Some devices will be pulling 15.5 milli-Amps and others 30.5 milli-Amps... that's almost a 100% variation! That seems to be your problem. In my experience, anything over a 15% variation suggests an hfe or vbe mismatch. The ones pulling 30mA will heat up more than the others and hog current and heat up the heatsink a lot faster.
I wonder what others think about this... a 100% spread across the OP devices seems excessive. Maybe the repair was done with fake OP devices?
On the bad channel, (assuming this was fixed) 3.4mV to 6.7mV suggests a problem. Some devices will be pulling 15.5 milli-Amps and others 30.5 milli-Amps... that's almost a 100% variation! That seems to be your problem. In my experience, anything over a 15% variation suggests an hfe or vbe mismatch. The ones pulling 30mA will heat up more than the others and hog current and heat up the heatsink a lot faster.
I wonder what others think about this... a 100% spread across the OP devices seems excessive. Maybe the repair was done with fake OP devices?
If you wanna go easy way... leave biasing to temperature in the 6-9mV range, if you want to do it the right way, you will need to gain and vbe match all OP transistors... this will be a lotta work.
So i found one channel that not repaired, hotter than other, and no matter if i lower bias or not.
This makes sense. Start with the OP that is highest vdrop across the ER. Monitor it’s temperature vs the others.
You could have mismatched outputs or even a few fakes in there. You will need to gain match them.
The other thing to look out for is oscillations on that channel. Hopefully someone can walk you on how to troubleshoot oscillations. I am no good there.
You could have mismatched outputs or even a few fakes in there. You will need to gain match them.
The other thing to look out for is oscillations on that channel. Hopefully someone can walk you on how to troubleshoot oscillations. I am no good there.
Bias at highest OP match with two other, lasts are 2mV lower, nothing intresting, - maybe oscillation.
I see no way forward until it is established whether the transistors are genuine and until the fakes have been replaced. And not just the outputs. Yes, oscillation is about the only reason for low dc current yet high dissipation.
Replacing all fakes with originals should sort the oscillation worry all by itself, provided no other changes have been made while repairing the amp. Ime Krell don't go nuts matching transistors, at least in the amps i have had matching was no better than 20%.
Replacing all fakes with originals should sort the oscillation worry all by itself, provided no other changes have been made while repairing the amp. Ime Krell don't go nuts matching transistors, at least in the amps i have had matching was no better than 20%.