From an online formum I got this:

I used a hot glue gun on the plastic enclosure speaker cable outlet – reused the original through-panel cable clamp and then just smothered the entire assembly with glue. This works well as it’s fairly easy to peel the glue off again if needed.
For the large components I used a few blobs of Araldite Precision as its fairly heat resistant once set – didn’t bother with the smaller stuff.

I’ve compiled a list of the replacement/substitute components I used, and where I got them:

Main Board
Power Amp Caps – Nichicon KG 4700uf 80v x2 (Parts Connexion) – these are 5mm wider than the original Su’scon items, but interestingly the screen printing on the main board has a placement outline exactly the same diameter as the Nichicons so they clip in without fouling any adjacent components.

Digital Amp Board (DS-150) (104-DS15000V-E)
Power Transistors – Vishay IRF9640PBF / IRF640PBF (Mouser)
Electrolytic Caps - Nichicon EP (Bipolar) 10uF 100v 105C (Mouser)
Opamp - Texas Instruments TL072BCD (Mouser) – has a lower input offset voltage (3mv) than the original ST Microelectronics TL072CDT (10mv)

Relays – Omron G2RL-1-E-24VDC 16A x2 (Rapid) - The original speaker relay is a Goodsky 5A DPDT with the two poles paralleled on the circuit board input/output tracks, the Omron 16A SPDT has the same terminal layout and works fine. The original power relay is a Goodsky 16A SPST, but the power board has the extra through holes for a SPDT layout blanked off with solder, so you can open up the holes on the board to take the Omron. The Omrons are approximately 2/3 the height of the Goodskys as well.

Power Board Caps (Farnell originally, I think) - Panasonic FC 1000uF 25v 105C x2, and 22uF 50v 105C x2 Large caps tested OK so  I left them as is.

So did all that.

Still nothing.

 All the ribbon cables were reconnected.  Most were corroded away. The original glue had become corrosive and conductive!
- Power relay not working. Re-flowed joints on  connecting cable. And all the joints on the power board.
- R54 is a bodge resistor on the reverse side and under glue. It was not connected at all. Cleaned and re-flowed.
- Removed all the shitty glue and re-soldered and components underneath. This often meant removal, leg cleaning and resoldering.
- Replace all the caps.

So obvious faults dealt with.

- Relay 101 NC contact was open. Replaced it.
- Now. To turn on Q103 and hence output relay, Q110 needs to be off.  It never was. Pulled tested and re-inserted Q110 and removed C110 just in case.  Now the (noisy) line is going low and turning on the base of Q103 but still no relay.  removed Q103 to find it's mostly missing!  Couldn't see that bottom side. Replaced with new and now RL101 turns on and we have output. Yay. Very noisy line on base of Q110 though. Replaced C110 on back side of PCB due to broken traces.
- Switch SW01 is iffy, sprayed with switch cleaner and hopefully this will fix.

I also did

There is a high audible Hz tone from the digital amp.
This can be heard even with no speaker attached. 
This may be due to some subtle interaction with the new op-amp IC1 and the different Q11 and Q10 and these were replace with slightly different types. IRF530PBF and IRF9530.
These are slighty lower voltage (100V) and are probably being over driven as the Vds can be between the +60V to -60V rails
when the complementary MOSFET is conducting.

The lower offset voltage of the IC1 op-amp may be a factor but I'm not sure what to change to counteract for this. 
I assume that the original oscillation was above hearing level and so not a problem.
The oscillation generation is the IC1B circuit and the output freq. is 44kHz.  
A component change may push this above audio. 20K perhaps? 
More knowledge is required to fix this. C6 perhaps? (Nope)

Cont'd:  Looking into offset voltages etc. for IC 1B, which produces the triangle wave, I noticed the the bottom of the wave form was a mess.
I figured this was an offset issue, so I paralled R10 with a 3K3 to drop the offset a bit. (Undone it made no difference).
This was after attaching a trim-pot and slowly reducing it's resistance.
The whine went away and the waveform looked clean.  
Attaching input produced output at the speaker. Yay! This does slightly change the gain though.
Not sure what the input levels are supposed to be though as increasing the volume causes the woofer, the relays trip out, to cut out.
Which I guess it's supposed to do in order to protect the driver.

Much Much Later

Power amp replaced with a unit from e-bay. £20
It works sometimes and I think the replacement of of Q141 with a BC547 is not a good enough match to disable (over temperature) mute. Correction they work fine it seems.
Shorting EC on Q103 trips the output relay on OK so the relay is OK.
The amp is still muted though so there's nothing to amplify and Q103 base is held low by the over temperature circuit.
This does not break the signal path though. But see below.
Awaiting new transistors. Now not needed. Oh well, more stock I guess.

OK. at some point the -15V went missing so no pre-amps were working, Another weaktrak had broken when cleaned up some mor gkue damage.

In addition another PCB track broke at the output connector to the power amp. This is because I'd had the amp module off so many times.

All the digital amp connectors were replaced with Molex KK254 headers to connect to the external amp which was mounted on the outside to the plastic sealing box.

All assembled now.  Happy woofer.

Final Note.

I found some schematics for later revisions of the Digital bord. These have an extra P-type MOSFET and some extra output filtering. So clearly ther's an issue with the one I had. I'll post all the schematics soon. Possibly on GitLab.

I learned a lot on the way too.

Last changed: 8. October, 2022 at 22:32