JBL PowerBass PB12 repair tips (aka how to get a nice sub for $42!)
The following experience applies ONLY to Revision 2 of the JBL Class D amp! Revision 1 uses an AMI module (rare, expensive and junk!) while Revision 2 uses a discrete output stage.
These tips may be useful with JBL E250P and JBL P12SW amplifiers as well. They are almost identical. Some component numbers may have changed but the topology is identical.
I bought this sub from a local electronic junk shop for $25 with a tag "lights up, no sound". My kind of project! It even had my Electronics I professor interested.
Upon disassembly, I found a badly damaged output filter. The caps in the first stage of the LC filter were destroyed. The NP electrolytic (C72) was vented and the SMD ceramic (C71) had been set alight with visible flame damage to the board. Ouch.
If it was just the electrolytic, I'd have chalked it up to cheap caps. This was obviously not that simple....I checked the output stage. To my surprise, the MOSFETs (IRF640, Q18, Q22) were good!
My attention was then drawn to a darkened section of the board in the power supply. It was obvious that the two large power resistors, R6 and R9 feeding the aux supplies had gotten extremely hot. They had burned the solder mask from the traces near the connection points but the traces were intact. The solder joints were in very poor condition. Suspecting a short, I tested the transistors and Zener diodes that were part of the simple pass regulators. (D6, D7, Q2, Q3) They were good. I replaced them anyway as it was obvious that they had become extremely hot.
I then changed out the caps in the aux supplies as they were also exposed to the excessive heat. (C9, C4, C2)
The resistors were also swapped out. They had become so hot that the color bands were discolored, almost like being sun bleached. I ordered them as prescribed in the service manual and I was surprised to see that the ones mentioned in the service manual were far larger than the ones originally installed. The schematic calls for a 1.2K 3W for R9 and a 1.2K 5W for R6. The ones installed from the factory looked to be 1 and 3W respectively.
The main filters were over 10 years old and only 85C units mounted near the heatsink, I figured these too should be changed for long term reliability. I used 105C units of slightly higher capacity. The board is silk screened for larger ones. A common 4700uF 80V fits just fine and inrush is manageable.
The flame damage was scraped away and the hole in the board filled with silicone. A new SMD cap was added 2mm away. A film cap was put in place of the 6.8uF NP electrolytic as that exact size wasn't obtainable locally in an NPE cap. The other NPE cap, a 4.7uF unit was swapped out for a new one. (if it fails, the output stage could fail with it)
Testing time....click...lights dim...bright flash. Dead. Uh oh. I made it worse, the outputs have failed shorted.
I realized that during the cap replacement,the via feeding -Vs to the aux supplies was removed with the old cap. They are very fragile on this unit; be careful! Fixing this allowed me to pop in a spare set of MOSFETs and try again.
Click...green lights! It's working...wait what's that noise? Before I could verify my input source, I was again greeted with the bright flash of a fuse blowing and I decided to shelve the project for the night.
The next day, I got smart and decided to test the waveforms. Clearly something was causing both outputs to simultaneously turn on. So with MOSFETs removed, I started testing ALL the little supplies. I found the mute voltage grossly out of spec and a badly deformed wave coming out of the modulator op amp (pin 7 of U6) I now had a lead on the MOSFET killer.
The culprit ended up being a small SMD transistor on the other side of the board that failed partially shorted. If I recall, it was Q26 and it was in the region that was overheated from R6 and R9.
I put in a fresh set of IRF640 MOSFETs, put on some silicone thermal pads (lower junction resistance than a mica+grease), bypassed L5 as prescribed in the service manual, swapped out C65 as a preventative measure and fired it up. Neighbor relations are at an all time low...luckily I've figured out when they go to class. :D
So here's the condensed version:
When faced with a PB12 with the R2 board, replace R6 and R9 with the proper wattage resistors. This was the initial failure. Their heat caused Q26 to fail and also discolored the board. The failure of Q26 led to the modulator not working well, which destabilized the OPS. If changing the main filter caps, verify via continuity. Use a jumper if you break one.
For reliability's sake, change ALL the semiconductors and electrolytic caps in the overheated region. If the aux supplies go down, this amp will likely self destruct and the parts are inexpensive. C65 is critical to the half bridge driver (IR2111) and should be swapped for something newer to improve long term reliability as well. If you have a turn on chirp, C65 can be downsized to 47uF. (see E250P service bulletin)
Verify that a 100kHz 24Vpp square wave exists on pin 7 of U6 with respect to the negative speaker terminal. Bypass L5 if present via a thick jumper wire or just desolder L5 and use the single winding as the jumper after removing the ferrite. (This is far more elegant than a jumper, IMO). Install MOSFETs ,a new 2A time delay fuse and enjoy.
I have a pb 12 it cracks when turned on , once on its fine.
I opened up the amp its revision2 c
the main caps look like they leaked at the base of them(rust colored stain)
where did you buy all your parts?
Im also dealin with the same amp board having issues. I know c72 needs to be replaced. Can you help point out on the board what else should be replaced.
for someone else dealing with same unit. I found the JBL service manual as well.
|All times are GMT. The time now is 09:05 PM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio