I think that I have made a correction to the component ID comparison above.
I had >35VDC across the output terminals before lifting the transistors. I shut it down as that 475 Ohm resistor was getting too hot. The amp is drawing about 6A with my 3 Ohm resistor in line.
After lifting what I think is Q21 & Q22, it still draws about 6A and there is ~44V across the output terminals. Something else started smoking on the driver board, I shut the amp down but I didn't really get a good look at exactly where other than it was a different spot than the resistor I know has been getting hot.
I have not measured it lately, but I am sure that I don't have 14V across the amp's GRD & B+ terminals as the fans are not turning while powered.
I have removed both drivers that were driving the problem side and components stopped smoking on the driver board. This is good in that it allowed me a little more time to take measurements.
I still had ~41VDC on the output terminals, however I measured much less DC across the gate and source of the outputs, down to ~1VDC from ~25VDC.
I measured resistance across the each of the three output legs in circuit and I didn't find any low readings. When I removed all of the outputs from the side that had ~25VDC (now ~1VDC) across the gate and source, I found another two shorted again, across all three legs. I have not kept track of which spot they are coming out of since I do not realize any are shorted until after they are removed. How is that possible, if they are shorted they should be found in circuit right?
This leaves me with a few more questions.
Why does it keep damaging two outputs each time? Is this just a coincidence? Is it that none of the original outputs can be trusted to be good even if they have measured ok? Could it be a certain pair of board locations that are damaging them? I'm certain the previous two times I have found shorted outputs, they came one from each side not both from one side like this time.
As the amp sits now, both drivers are still out as well as the entire side of outputs they drove. The other side of outputs is only partially populated, otherwise everything else is in place.
I still had ~41VDC on the output terminals, however I measured much less DC across the gate and source of the outputs, down to ~1VDC from ~25VDC.
I measured resistance across the each of the three output legs in circuit and I didn't find any low readings. When I removed all of the outputs from the side that had ~25VDC (now ~1VDC) across the gate and source, I found another two shorted again, across all three legs. I have not kept track of which spot they are coming out of since I do not realize any are shorted until after they are removed. How is that possible, if they are shorted they should be found in circuit right?
This leaves me with a few more questions.
Why does it keep damaging two outputs each time? Is this just a coincidence? Is it that none of the original outputs can be trusted to be good even if they have measured ok? Could it be a certain pair of board locations that are damaging them? I'm certain the previous two times I have found shorted outputs, they came one from each side not both from one side like this time.
As the amp sits now, both drivers are still out as well as the entire side of outputs they drove. The other side of outputs is only partially populated, otherwise everything else is in place.
If they read 0 ohms out of the board between all legs, you should have been able to see that they were shorted in the board.
It damages only 2 outputs because after those short, the rest are protected (by the shorted ones).
It could be that they're failing in one location if there is a defective gate resistor, an open trace or a damaged insulator. One of the FETs in the other bank would fail due to the first one failing.
It damages only 2 outputs because after those short, the rest are protected (by the shorted ones).
It could be that they're failing in one location if there is a defective gate resistor, an open trace or a damaged insulator. One of the FETs in the other bank would fail due to the first one failing.
I removed the rest of the 640Ns from the other side and one of those was shorted also, well really low resistance, now down to 7 of the original 14 outputs.
I find that all gate resistors are between 10 and 10.1 Ohms.
None of the traces to the outputs appear to be opened. I checked each gate pad of every output back to the corresponding pin on the driver board. I checked from one bank of drain pads back to the rectifier. The other drain pads connect to the opposite bank of source pads. I checked the remaining source pads back to the other rectifer. All good there.
I will put in new insulators, but I don't think that was the problem either. You had mentioned that earlier and I checked for continuity between the drain and heatsink and found nothing.
I have yet to find the defect on the driver board.
If none of the above conditions are the cause, the outputs have to be getting damaged by a defect on the driver board or is there something else I am missing?
If I reinstall the drivers, is there anything to learn with none of the outputs installed or do I have to have at least one 640 per side?
I find that all gate resistors are between 10 and 10.1 Ohms.
None of the traces to the outputs appear to be opened. I checked each gate pad of every output back to the corresponding pin on the driver board. I checked from one bank of drain pads back to the rectifier. The other drain pads connect to the opposite bank of source pads. I checked the remaining source pads back to the other rectifer. All good there.
I will put in new insulators, but I don't think that was the problem either. You had mentioned that earlier and I checked for continuity between the drain and heatsink and found nothing.
I have yet to find the defect on the driver board.
If none of the above conditions are the cause, the outputs have to be getting damaged by a defect on the driver board or is there something else I am missing?
If I reinstall the drivers, is there anything to learn with none of the outputs installed or do I have to have at least one 640 per side?
With no output transistors installed but with the driver board installed (with all components in place on the driver board), drive a strong 100HZ signal into the amp. Do you see the drive signal on both the high and low side FET gates?
You may have to ground the output filter inductor to see the signal on the high side gates.
You may have to ground the output filter inductor to see the signal on the high side gates.
With the scope grounded at the main power ground I see nothing at either gate, operating the scope settings through it's ranges. I have not grounded any output inductor yet.
Some new things to possibly consider are the yellow voltage light is lit as soon as I apply 14VDC at the power terminals and the amp draws ~.025A before remote is triggered. After remote is triggered it only draws ~.150A. I am NOT using any resistor in line with B+ at this time.
It appears that the amp's power supply is no longer working now.
Some new things to possibly consider are the yellow voltage light is lit as soon as I apply 14VDC at the power terminals and the amp draws ~.025A before remote is triggered. After remote is triggered it only draws ~.150A. I am NOT using any resistor in line with B+ at this time.
It appears that the amp's power supply is no longer working now.
At post #57 just above, that would be the first time I attempted to power up the amp without any outputs installed. I have now installed two 640Ns per side and it remains in protection.
Previously, I have powered it up successfully without rectifiers and without the driver board installed (two seperate times). It would appear that I have created another problem to repair somehow. There is an LM339 on this board labelled U2.
Previously, I have powered it up successfully without rectifiers and without the driver board installed (two seperate times). It would appear that I have created another problem to repair somehow. There is an LM339 on this board labelled U2.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.