Hi, I'm currently repairing a Power 500a2. The left channel had Q125 & Q115 shorted, and R182 & R163 source resistors visibly overheated/burned. D7 was also shorted. The right channel appears intact, but not fully verified.
I pulled all the left channel source resistors to replace with all matching, and clipped the legs of Q125, Q115, & Q127 (all the IRF540), replaced D7, and set both bias pots fully CCW.
I powered the amp up with a 10a fuse inline and it drew about 0.74A at idle. The right channel produced low, possibly distorted sound. I later realized the RCA input had both shield legs broken between the pcb & RCA riser block, but before I realized this I attempted to bring up the right channel bias. I slowly rotated the bias pot looking for a .05A rise, but never got any increase before going full CW. I turned the bias back full CCW at this point. I then realized the broken RCA when I pushed on the RCA block and got brief loud playback, but I couldn't maintain the connection well enough to judge if it was without distortion.
Nothing got hot at any point. I've now replaced all the left channel source resistors and replaced Q125, Q115, & Q127. I still need to repair the RCAs with a replacement coming tomorrow.
I checked every resistor, diode, and transistor comparing left & right channels and everything was within tolerances, or matched right to left, as far as I could tell. Except...when I check the IRF 9540's Q126, Q119, & Q124 and compare them to Q226, Q219, & Q224, I get a different reading. In diode mode when I connect my positive lead to the drain leg, and negative lead to the source leg, I get 667 for all left channel IRF9540, but I get 753 for all right channel IRF9540. Measuring resistance gives me 550ohm and 618ohm. Source to gate in diode mode I get 832 vs 825 between channels.
The IRF9540 are matching date codes on a given channel, but are different date codes left vs right channel. Is this normal to be that different? The IRF540 seem to give much closer readings between left & right.
Why did my bias adjustment have no affect on current draw?
I have not powered it back up since replacing source resistors & IRF540's. I wanted to see if you guys had any input on anything I might have overlooked. Thank you
I pulled all the left channel source resistors to replace with all matching, and clipped the legs of Q125, Q115, & Q127 (all the IRF540), replaced D7, and set both bias pots fully CCW.
I powered the amp up with a 10a fuse inline and it drew about 0.74A at idle. The right channel produced low, possibly distorted sound. I later realized the RCA input had both shield legs broken between the pcb & RCA riser block, but before I realized this I attempted to bring up the right channel bias. I slowly rotated the bias pot looking for a .05A rise, but never got any increase before going full CW. I turned the bias back full CCW at this point. I then realized the broken RCA when I pushed on the RCA block and got brief loud playback, but I couldn't maintain the connection well enough to judge if it was without distortion.
Nothing got hot at any point. I've now replaced all the left channel source resistors and replaced Q125, Q115, & Q127. I still need to repair the RCAs with a replacement coming tomorrow.
I checked every resistor, diode, and transistor comparing left & right channels and everything was within tolerances, or matched right to left, as far as I could tell. Except...when I check the IRF 9540's Q126, Q119, & Q124 and compare them to Q226, Q219, & Q224, I get a different reading. In diode mode when I connect my positive lead to the drain leg, and negative lead to the source leg, I get 667 for all left channel IRF9540, but I get 753 for all right channel IRF9540. Measuring resistance gives me 550ohm and 618ohm. Source to gate in diode mode I get 832 vs 825 between channels.
The IRF9540 are matching date codes on a given channel, but are different date codes left vs right channel. Is this normal to be that different? The IRF540 seem to give much closer readings between left & right.
Why did my bias adjustment have no affect on current draw?
I have not powered it back up since replacing source resistors & IRF540's. I wanted to see if you guys had any input on anything I might have overlooked. Thank you
The different date codes could account for the different readings, D-S. G-S is the board.
The bias won't make a difference if the biasing isn't enabled or if there is damage to the circuit.
The bias won't make a difference if the biasing isn't enabled or if there is damage to the circuit.
How does biasing become disabled? Any particular failure modes you're familiar with. The other amps I've repaired have had a couple of the commonly failed resistors or transistors, but I hadn't seen the bias not do anything til this one.
With it fully powered up or only remote power? I will check tomorrow evening after kids are in bed. Thanks, and good night
15.20V
PC-2335-F
If it matters, I get -3.6mV out of the right channel and 20.7mV out of the left
Bias is acting normal now with everything in place. Both pots give the same 0.05A rise at about the same position. 540's got slightly warm, not hot, after a few minutes of being on (not clamped to heatsink, just on mehsa). I will have to wait til tomorrow to check the outputs. Hopefully all is well.
Hooked it up and seems to be good to go. I'm still curious if the voltage on the LM339 is correct, though?
When you stated that the amp wouldn't bias, I wanted to know if the biasing was being enabled. With pin 14 at 15v, it is.
Perhaps it didn't like operating correctly with the 540's and source resistors off.
Thanks Perry
Thanks Perry
Never pull the source resistors and try to pass any current through the outputs, It forces all of the current through the protection circuit resistors which generally causes them to fail.
Sorry, I'm confused then. I had all three 540 out, all three 9540 still in, and all six source resistors out. No speakers hooked up. Still a concern?
Probably not without a load and only one bank of the FETs in the circuit. In the future, I'd recommend that you never apply power with the source resistors out of the circuit.
Okay, I'll remember that. What's a good method for testing one channel when the other is surely blown?
Generally, only one FET fails in each bank of outputs. I generally cut those out to see if the channel has other problems (in the drive circuit...). If it does, it's generally not a problem to test the good channel with the shorted outputs out of the amp in the bad channel as long as the bad channel isn't drawing excessive current.
Okay, so as long as the source resistors are still in what the rest of the fets, it's fine. This one had two shorted 540's, tested out of circuit and still shorted. The third seemed questionable and went ahead and replaced all three with new ones from the same date code.