The First Watt J2 dates from ~2009, and when on, which is a lot, averages 65-75C on the heatsinks, depending on ambient temps.
With that in mind (and little else, TBH, it sounds fine), I decided I’d replace the 15,000uF smoothing caps which were rated at a stingy 25VDC and 3000hr/85C lifetime rating -
Popped a couple of caps off the board and, well, some naughty microfarads are AWOL -
Once all the caps were out, it became obvious the flush-mounted resistors were also cooking the PCB -
So I whipped 'em off…
Nice new caps (R) vs. slightly off-colour old ones (L) -
The new ones are 10,000hr/105C rated, so prolly see me out…
I’ve kept the old ones for charging-up and then chucking towards unsuspecting people to catch…
Here’s the board part re-populated with uprated Mills resistors stood well off the board -
I also pulled the fucking stupid T4L microfuses - sure they’ll probably never pop (and if they do, the amp is possibly scrap, 'cos dead silicon), but what a PITA to change if not! So, boggo 20mm cartridge holders and fuses into place instead.
Also binned was the X2/Y2 cap across L&E - completely pointless considering the filtering at the other end of the power cord - and more-pertinently, in the way.
That’s as far as I got today. I’ll pop the caps back in next time I get a chance to fanny-about with a soldering iron. No rush, waiting on some other parts anyway.
Meanwhile, I’m suspiciously eyeing the NTC thermistor (green disc) that couples system-ground to mains Earth. I get why he’s used it, it’s clever - in overcurrent situations its already token 10R resistance drops to nothing - but it won’t do so quickly (1 - 12s), it won’t do so repeatedly if its hit hard enough, and it’s kind of hard to derate sensibly across the amp’s operating temps - it’s probably sitting at closer to 2R most of the time, which makes it a mite pointless, so I may pull it and put something more conventional in there.
Grounding overall is also a bit random - e.g. the standoffs could ground to the chassis - except no-one removed any anodisation, same was true for the star-ground point on the chassis until I sorted it, and only one XLR socket was grounded, so I’ll get everything back to one point via as few routes as possible.
Not that it’ll make the slightest difference to the sound…