Sir aimen06 . . . . . . .
What I think has happened to YOUR unit is simply the mere failure of one or more of four electrolytic capacitors in the power supply circuitry of your JBL woofer amp.
That would result in a loud UNCONTROLLABLE buzz . . .or HUM as I probably would classify it.
Your unit is using amongst the first generation of discrete built " Class D " power audio amplifiers.
And getting to think about it, with impure DC power coming into it, its reproduction might take on quite a bit of raspiness.
Later generation D class amps have all of the unit built in a chip and being quite a small one at that.
The areas of principal interest are as being shown below:
The very top portion is showing their physical locations on the PCB.
The central position is showing them functionally, in both a block diagram and also schematically.
The people that I am usually assisting , tend to not even have " un pot pour faire pipi dans "
as far as any test equipment.
So we try to work it out with that which is being available.
Minimally, I am seeing a DVM, BUT, I'm not knowing HOW the particular brand and model that you have,
reacts to with a simultaneous AC and DC voltage presence.
In our situation here, we will typically be finding a higher DC level and a minimal to moderate AC
levels presence and we will be using the instrument in its AC range.
Knowing how "dodgy" some instruments used in the past have acted in this situation, lets just modify.
Use in the order of a 1 ufd paper or poly cap in series with the meters RED metering lead.
The series capacitor blocks the DC component and only the AC component passes through.
In this manner, the approximate 1300 ohms of series AC reactance of the cap at a 100~ ripple sampling should not diminish accuracy significantly when feeding into the multi decades of megs high input impedance of a DVM.
That particularly would not be a problem, when you are measuring either of those same capacitance pairs.
Because . . . the two readings would be very matched . . . .UNLESS . . . . one has appreciably declined in capacitance, and therefore WOULD be showing a high AC ripple reading.
Or, it would be showing an EXTREMELY high reading, if having opened up.
Sooooo get the set back to a state where you can power up and still be able to reach and take the AC readings across each of those 4 to 6 capacitors in the power supply.
THEN, see if you now can find your self a HIGH ripple /bad electrolytic capacitor.
( The last two are being a couple of small 100's which I don't really suspect.)
HINT:
In hard to get to / reach situations I just solder tack on wire connections and remote those then labeled wires out into the accessible world, where the readings can then be taken . . . .PLUS . . . no worry about slipping probes with their resultant
. . . .BAMMMMMMMM . . . .or SSSSSSSSHmoke . . . there went a resistor .
Now . . . . . the final "Cow Pee de Grass " (excusez mon français) . . . . is all of "youse guys" concern with the units side daughter board with those non polarized electrolytic caps that you were zeroing in on.
As previously mentioned, that section is being what really does the WORK . . . .being the Class D power audio amplifier for this unit.
I have put its schematic on, as the final reference at the bottom.
Basic rundown:
Audio comes in at the left side and is routing into IC1A and additionally has the presence of a ~350 Khz triangle generator waveform from IC1B at the bottom.
350Khz is minimally being 10X+ the audio high end range of 20 Khz, and there can be progressively Much more sampling points used on the lower audio frequencies.
The audio input and the triangle reference compare and create coincident variable pulse width square waves.
Whereas the analog audio was relating to frequency and amplitude.
Relating to the units digital aspects:
Refer to the very bottom comparison chart:
One then sees the transition of there being a FULL amplitude square wave, BUT, the audio has been converted to variable pulse widths that replicate the sampled audio.
The processed digital info is passed on thru the amp until the two inverted and non inverted outputs drive the final output power FET's to then sum the outputs and get a greater magnitude of output.
THEN the prior mentioned YELLOW boxed filtering network comes into play in stripping off all of the no longer needed digital square waves trash and harmonics and then filters /and/ transitions them back into POWER amplified analog audio .
ASIDE:
On the "yellow pookie" adhesive used on the board . . . . with my confirming . . . . indeed this is the very worst blend, being associated with hygroscopic tendencies.
Its worst, when its being in contact across any high impedance circuitry . . . .like the gate of a FET.
Or if the voltage under / across it, is up on the hundreds of volts . . . . or if that voltage could leak over to the prior mentioned type of transistor
OR to the base of a common junction transistor.
Trying to remove it . . . . INDESTRUCTIVELY . . . . I liken to trying to extract mortar from brickwork.
If I see that I have too great of an amount or something precludes its removal . . . .herewith . . . my work around.
As I mentioned, its humidity/ moisture absorption is being its problem.
What I do is use a hair dryer on the rework area and heat / warm enough and LONG enough to drive off ALL moisture. Then I brush on either Clear poly varnish or Clear airplane dope or Clear lacquer .
Then, as that sealant cools and dries, it gets pulled further back into the coated area.
No more moisture problems.
Thasssssit . . . . .
73's de Edd