I have a PPI PC2400 that I had working at low levels. It originally came to me with blown PS FET's and a blown SG3525. It appears on this amp the PS FET's are driven directly from the SG3525, so it does not really surprise me the SG3525 went along with the PS FET's. I replaced both parts and all was good. I then went to power test it. I originally just used a MP3 player with WAV test tones. The drive level was not nearly high enough to get any useful power out even with the gain maxed. So I had to switch over to a function generator to get more voltage drive. When hooking this up my problems started. As soon as I hooked the ground of the function generator to the RCA inputs, the amp started oscillating (and hitting the 10A current limit I had set). The amp started making internal clicking noises, so I promptly turned it off.
I figured out the issue with the function generator, but after I got it running again it got to about 5W output, then the current just went straight down to 0. After this point I could not get the amp to turn back on at all. So I opened the amp up again to check the output and PS FET's. All of those are still fine. I powered it back on with just the +12V, remote, and ground hooked up trying to trace where power was and where power wasn't. I found that pin 15 of the SG3525 was only getting 1.8V. So I figured I had blown another 3525 during the oscillation. I removed the SG3525, looked at the pad without a part and it got the full 12V. I replaced the SG3525 again, but I still get 1.8V on pin 15. All the other pins sit at ~0V except pin 16 which sits at 0.5V. I figure Pin 16 is just trying to generate a 5V reference, but the input voltage is too small to get there.
So my next issue is trying to trace the power up to the SG3525. Does anyone have a schematic or diagram showing which transistors I need to check to get the correct power to the SG3525?
I figured out the issue with the function generator, but after I got it running again it got to about 5W output, then the current just went straight down to 0. After this point I could not get the amp to turn back on at all. So I opened the amp up again to check the output and PS FET's. All of those are still fine. I powered it back on with just the +12V, remote, and ground hooked up trying to trace where power was and where power wasn't. I found that pin 15 of the SG3525 was only getting 1.8V. So I figured I had blown another 3525 during the oscillation. I removed the SG3525, looked at the pad without a part and it got the full 12V. I replaced the SG3525 again, but I still get 1.8V on pin 15. All the other pins sit at ~0V except pin 16 which sits at 0.5V. I figure Pin 16 is just trying to generate a 5V reference, but the input voltage is too small to get there.
So my next issue is trying to trace the power up to the SG3525. Does anyone have a schematic or diagram showing which transistors I need to check to get the correct power to the SG3525?
That pin (on the 4400) is fed 12v via a 10 ohm resistor and a PNP transistor. In the 4400, the resistor is R104 and the transistor is Q33. I don't know if it will be the same in the 2400. If the resistor is out of tolerance or the transistor is defective, that would cause the symptoms you described.
It does not look like the numbering is the same as the 4400. I have included 3 fairly large pictures of the internals if that helps you out. I am about to go out to check some parts, but I thought I would post this up first. Do you think a troubleshooting step of powering pin 15 of the SG3525 with +12V directly could help?
Actually it looks like it is Q17 (next to R56 in pictures 1 and 3) I pulled that transistor out and applied a low current limit 12V signal to the pad and the amp powered on. With the transistor out of the circuit, it still seemed to test fine though. I tried to reinstall, but still no luck. I hope nothing else is pulling it down, but I could not find anything else tied to that line except pin 13. That is another difference between the 4400 schematic. It appears pins 15 and 13 are tied directly together on this board unless I missed something.
Now to order more parts... I swear these little transistors are going to be the death of me. I already have the surface mount SOT-23 version of this transistor, but not the TO-92 like I need.
Now to order more parts... I swear these little transistors are going to be the death of me. I already have the surface mount SOT-23 version of this transistor, but not the TO-92 like I need.
I will have to check the transistor next to Q17 later today. I did not find any continuity between any of the pins of Q17 and the one next to it. I did check Q17 in circuit yesterday, and here were the results. Certainly seems to point to Q17 being the issue.
Pin 1 (Emitter): 12.5V
Pin 2 (Base): 6.5V
Pin 3 (Collector): 1.8V
When Q17 was out, and no B+ cable hooked up, I measured R56. It measured 9.9 Ohm, so I assume that it is fine.
If you still think I should check the transistor next to Q17 I will. I did not get to play around with this amp a whole lot, I'm also trying to figure out a weird chassis vs earth ground issue on a kicker KX1200....
Pin 1 (Emitter): 12.5V
Pin 2 (Base): 6.5V
Pin 3 (Collector): 1.8V
When Q17 was out, and no B+ cable hooked up, I measured R56. It measured 9.9 Ohm, so I assume that it is fine.
If you still think I should check the transistor next to Q17 I will. I did not get to play around with this amp a whole lot, I'm also trying to figure out a weird chassis vs earth ground issue on a kicker KX1200....
I replaced Q17, and the amp turned on fine with an idle current of 1.1A at 13V. The power was very clean on the output, but very inefficient. I could only get to 10W out with about 110W in (amp was drawing about 8.5A at 13V). I left it idle there for awhile, then some magic smoke started coming out. Amp was still outputting clean power when I shut it down. When I opened it back up, R56 was badly burnt. R56 is the 10 Ohm resistor attached between B+ and the emitter of the replaced Q17 transistor. Obvious answer is the SG3525 is drawing too much current or something else downstream is the problem. Even after the burning, the resistor still measured 11.5Ohm. I will obviously replace the 10 Ohm resistor, but is there anything else I should be on the lookout for? Voltages and waveforms from the SG3525 seem fine still. I would think if the SG3525 was blown again (or anything in the power supply or output FETS), the idle current would be much higher.
Well after a month of having to work on some other amps, I'm back at this one. I got the amp to turn on fine. Its possible the SG3525 is damaged again, but for right now I'm gonna leave it. I replaced the 10 ohm resistor with a 1W version and MPSA56 that is connected to it. After reconnecting, the voltage drop across is was 0.35V, so I figured drawing 35mA is fine for the SG3525. I went looking at the gate and drain waveforms on the power supply transistors again. Below are some pictures of the waveforms.
One side gate and drain waveform:
Both PS drain waveforms simultaneously: Obviously looks dirty and not correct
At this point I figured pulling the rectifiers again was the next logical step. After I did this, I noticed that it looked like the waveforms started ugly like before, but eventually went normal. It initially made me think some sort of soft start issue, but I later figured that wasn't my biggest concern. I started taking some measurements at turn on delays.
200ms after turnon, the gate and drain waveforms looked almost identical to when the rectifiers were in normally (but with the rectifiers in, the waveforms always looked like this).
1000ms after turnon, the gate and drain waveforms looked good
After the waveforms looked good, I started going back to the simultaneous gate and drain waveforms.
Both gates simultaneously (Looks good to me, tell me if I'm wrong here)
Both drains simultaneously (One has more overshoot than the other, but I'm not too concerned with it)
Here is where I think I might have a problem. Correct me if I'm wrong, but isn't this classic shoot through because the drains are both on at the same time? So if that is my problem, is adjusting the deadtime my best solution? Between pins 5 and 7 on the SG3525 there is a 10 ohm resistor that looks original and hasn't been changed. I saw on some of your other PPI amps on your repair tutorial that it uses 100 ohm resistors there, but I can't say if that's a good value to try for mine or not. Is changing this resistor a good next step or do anyone have some other suggestions or tests I should try? With the rectifiers in (with no load) or out, the amp consistently draws about 1.1-1.2 amps at 13V.
Sorry for the long post!
One side gate and drain waveform:
Both PS drain waveforms simultaneously: Obviously looks dirty and not correct
At this point I figured pulling the rectifiers again was the next logical step. After I did this, I noticed that it looked like the waveforms started ugly like before, but eventually went normal. It initially made me think some sort of soft start issue, but I later figured that wasn't my biggest concern. I started taking some measurements at turn on delays.
200ms after turnon, the gate and drain waveforms looked almost identical to when the rectifiers were in normally (but with the rectifiers in, the waveforms always looked like this).
1000ms after turnon, the gate and drain waveforms looked good
After the waveforms looked good, I started going back to the simultaneous gate and drain waveforms.
Both gates simultaneously (Looks good to me, tell me if I'm wrong here)
Both drains simultaneously (One has more overshoot than the other, but I'm not too concerned with it)
Here is where I think I might have a problem. Correct me if I'm wrong, but isn't this classic shoot through because the drains are both on at the same time? So if that is my problem, is adjusting the deadtime my best solution? Between pins 5 and 7 on the SG3525 there is a 10 ohm resistor that looks original and hasn't been changed. I saw on some of your other PPI amps on your repair tutorial that it uses 100 ohm resistors there, but I can't say if that's a good value to try for mine or not. Is changing this resistor a good next step or do anyone have some other suggestions or tests I should try? With the rectifiers in (with no load) or out, the amp consistently draws about 1.1-1.2 amps at 13V.
Sorry for the long post!
The drain waveform isn't going to look like a 'normal' power supply in a cheap amp. The use of an inductor and having a regulated supply will make the drain waveform look strange. When the rectifiers are out and the supply goes to full duty cycle, the waveform looks normal because there's only a very short deadtime. I don't think there is a problem here.
Clamp all of the power semis to the heatsink and power up the amp. Start pushing on all of the various components to see if the resistor begins to overheat. I had a PPI amp that had a resistor end-cap shorted to a via and it would intermittently cause problems. Something similar may be happening here.
After looking through some other PPI documentation, I decided to take the amp out of the chassis and clean the EXCESSIVE thermal grease off and do it over. After I did this and reassembled the amp, I did not see any issues. I did cut away some random excessive leads, but I'm not sure that was my problem. I also reformed a couple leads on the outputs FET's so they would lay flatter. I tested each channel to 80W out. It was pulling 15A @ 13.4V to do 80W out, so it seems to be working OK. My MP3 player that I used for my test tones ran out of room past that.
For some reason I think the differential inputs on this PPI amp are REALLY unhappy when I hook up a single ended function generator. The amp oscillates and draws current like crazy. I haven't seen this issue on some other balanced input topologies (Rockford Fosgate, etc) I got it to "kinda" work with the function generator when I tied the power ground and signal ground together, but it has a 80-100kHz AM wave riding on the carrier. It wasn't clean by any means, so I just went back to the MP3 player with test tones. Do you have any experience in driving these inputs with a function generator? I want to be able to test it all the way to rated, but I would need my function generator to do it. What I am going to try next is to leave the shield ground on the amp floating, and connect the function generator ground to the power ground. Then just hook up the center pin of the function generator to the center pin of the RCA.
For some reason I think the differential inputs on this PPI amp are REALLY unhappy when I hook up a single ended function generator. The amp oscillates and draws current like crazy. I haven't seen this issue on some other balanced input topologies (Rockford Fosgate, etc) I got it to "kinda" work with the function generator when I tied the power ground and signal ground together, but it has a 80-100kHz AM wave riding on the carrier. It wasn't clean by any means, so I just went back to the MP3 player with test tones. Do you have any experience in driving these inputs with a function generator? I want to be able to test it all the way to rated, but I would need my function generator to do it. What I am going to try next is to leave the shield ground on the amp floating, and connect the function generator ground to the power ground. Then just hook up the center pin of the function generator to the center pin of the RCA.
They are 10 ohm on this one. They are within spec, I checked those out before. I'm pretty sure my problem has to do with single ended function generator vs differential input. I'm just about to go and experiment with it... will report back.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.