When this type 7 amp came to me, it undoubtedly had had a reverse polarity issue. One of the reverse protection diodes (D2) was blown apart and it was showing a short from battery(+) to ground. Replacing the diode fixed the short. On power-up the amp went into protection. In checking, I removed the outputs (IRF640Ns) and the output driver card. I traced the protection issue down to the over-current transistor (Q283) testing bad. While waiting for my order of transistors to come in, I went ahead and replaced the 3 each PD2 transistors(U1, U2 and U3), 2 each 1D transistors (Q5 and Q7) and the 2 TL072 op amps (U4 and U13) for good measure.
After replacing Q283 and checking the PS, I reassembled the output section and everything looked good on the oscilloscope. I mounted it back into the heatsink and connected a woofer and music input. The amp maxed out and the supply voltage dropped. The woofer sounded "muffled" and had a lot of static. I connected my larger power supply and removed the current limiter. It did the same thing (the amps only got up to about 8A. I wasn't really pushing it).
I removed the amp from the heatsink to do more tests. I used 13.8vdc supply limited to 4.5A. My signal input (when used) was 50Hz at 2vpp. These voltage readings are referenced to the (-) speaker terminal. I also tried referencing the battery ground and got the same results.
With no signal input, I found that I had -3.42vdc across the speaker terminals when the relay clicked on. I traced that voltage back to the low side drains and the high side source legs of the outputs. The voltage appears when the remote is turned on.
When I input my signal, there was -2.5vdc on the the low side drain legs and high side source legs. When the relay clicked on, I had the -2.5vdc across the speaker terminals. When I removed the remote, the voltage across the speaker terminals went to -58.9vdc. I checked this seven more times. The voltage readings I got were 12.31, 33.2, -3.32, -58, 58.2, -59.2 and -32.4. These spikes only show up on the speaker side of the relay.
when I connect a speaker ( I tried non-inductive loads, too), I get 6vdc across the speaker terminals, the 4.5A power supply maxes out and the supply voltage drops to 9.3vdc The voltage does not spike when the remote is removed. The signal at the speaker terminals on the scope looks like a half rectified sine wave with the crest chopped off.
I am running out of ideas other than to put all new outputs in just trying to make it do something different,
After replacing Q283 and checking the PS, I reassembled the output section and everything looked good on the oscilloscope. I mounted it back into the heatsink and connected a woofer and music input. The amp maxed out and the supply voltage dropped. The woofer sounded "muffled" and had a lot of static. I connected my larger power supply and removed the current limiter. It did the same thing (the amps only got up to about 8A. I wasn't really pushing it).
I removed the amp from the heatsink to do more tests. I used 13.8vdc supply limited to 4.5A. My signal input (when used) was 50Hz at 2vpp. These voltage readings are referenced to the (-) speaker terminal. I also tried referencing the battery ground and got the same results.
With no signal input, I found that I had -3.42vdc across the speaker terminals when the relay clicked on. I traced that voltage back to the low side drains and the high side source legs of the outputs. The voltage appears when the remote is turned on.
When I input my signal, there was -2.5vdc on the the low side drain legs and high side source legs. When the relay clicked on, I had the -2.5vdc across the speaker terminals. When I removed the remote, the voltage across the speaker terminals went to -58.9vdc. I checked this seven more times. The voltage readings I got were 12.31, 33.2, -3.32, -58, 58.2, -59.2 and -32.4. These spikes only show up on the speaker side of the relay.
when I connect a speaker ( I tried non-inductive loads, too), I get 6vdc across the speaker terminals, the 4.5A power supply maxes out and the supply voltage drops to 9.3vdc The voltage does not spike when the remote is removed. The signal at the speaker terminals on the scope looks like a half rectified sine wave with the crest chopped off.
I am running out of ideas other than to put all new outputs in just trying to make it do something different,
Pin 1) .040
Pin 2) .013
Pin 3) .011
Pin 4) -15.10
Pin 5) -.019
Pin 6) .010
Pin 7) .010
Pin 8) 14.84
Strange, but I noticed that the amperage went up on my power supply when I was probing pins 2 & 3.
I was using my DMM referencing battery ground. The amp is pulling 1.5A with remote engaged and while probing all the other pins.
The amp pulls 3.1A while probing pins 2 & 3.
Pin 2) .013
Pin 3) .011
Pin 4) -15.10
Pin 5) -.019
Pin 6) .010
Pin 7) .010
Pin 8) 14.84
Strange, but I noticed that the amperage went up on my power supply when I was probing pins 2 & 3.
I was using my DMM referencing battery ground. The amp is pulling 1.5A with remote engaged and while probing all the other pins.
The amp pulls 3.1A while probing pins 2 & 3.
Not shorted to ground. I had figured it was either the output card (maybe some of the components that I didn't change?) or something with the output FETs. Didn't really think about it being a bad new component. I will pull the output card tomorrow and replace that TL072. I should have used my jumpers with the card instead of soldering it in. I will this time. Can you think of any other components that may need checking?
I will also change out D5 and D6. When I was replacing components on the driver board, using my hot air gun to remove the TL072s I forgot that I had turned the air volume up and I blew 3 or 4 components off the board. I lost the 2 diodes. I was told that they were 1N4148 diodes. I ordered some but when they got here, they were the wrong package type and were too long for the pads. I was able to bend the leads down and around and used solder paste to get them in. I have 1N4148 SOD-323 now instead of the 1N4148 SOD-123s. They do have a different designation printed on them (T4) as opposed to the originals that I believe were A2s.
I will also change out D5 and D6. When I was replacing components on the driver board, using my hot air gun to remove the TL072s I forgot that I had turned the air volume up and I blew 3 or 4 components off the board. I lost the 2 diodes. I was told that they were 1N4148 diodes. I ordered some but when they got here, they were the wrong package type and were too long for the pads. I was able to bend the leads down and around and used solder paste to get them in. I have 1N4148 SOD-323 now instead of the 1N4148 SOD-123s. They do have a different designation printed on them (T4) as opposed to the originals that I believe were A2s.
Last edited:
Not at this point.
In the future, I wouldn't replace parts unless you have a good reason. In ALL of the amps of this type that I replaced, I don't remember ever having one of the large transistors be defective. In almost all cases, the two 1D transistors in the center of the board were all that needed to be replaced. In a few amps, the top-op-amp was bad but that was rare.
AFTER you get the amp working perfectly, if you want to replace working parts for some reason, you can do so but replace only one at a time, especially on class D amps. After replacing the single part, confirm that the amp is working properly.
In the future, I wouldn't replace parts unless you have a good reason. In ALL of the amps of this type that I replaced, I don't remember ever having one of the large transistors be defective. In almost all cases, the two 1D transistors in the center of the board were all that needed to be replaced. In a few amps, the top-op-amp was bad but that was rare.
AFTER you get the amp working perfectly, if you want to replace working parts for some reason, you can do so but replace only one at a time, especially on class D amps. After replacing the single part, confirm that the amp is working properly.
I appreciate that. I was told by another tech that it would be "good practice" to replace those if you ever had to pull the board.