On Fri, 21 May 2010 02:20:17 +0100, Arfa Daily wrote:

The first is the Warrior amp that I posted on here about, looking for
schematics. None were found, and as expected, the importer ignored my
pleas, so I decided I would spend a half hour on it 'blind'.

It turned out to not be too difficult to get the main PCB out, complete
with heatsinks and back panel. The wiring was long enough to allow the
board to be turned over, without having to disconnect everything. The
fault was that one of the two identical output stages was behaving as a
pretty good half wave rectifier, but only with a load connected. With no
load, an applied sine wave was perfectly symmetrical at the output
terminals, and of similar size to the good channel. With a load
connected, the negative excursions disappeared almost totally. Nothing
was burning, and the the output protect didn't even fire until the wick
was turned well up, which led me to believe that the problem may well be
back in the driver stages or earlier. As there are two identical amps, I
figured that I would start with a few comparitive resistance checks
between channels. Quickly, I found that at the base pin of one of the
driver transistors, I had a reading of 3k or so on the good channel, but
open circuit at the same point on the bad channel. I followed the print
back and took another reading and Lo! - 3k ...

So I went back to the transistor leg - open, but at the joint, 3k. I
tell you, I examined that joint with the strongest light and magnifier
that I have, but you could not see a problem with it. However, as soon
as it was resoldered, 3k on the leg as well, and the amp then worked
normally. This is the problem with lead free. You can no longer spot bad
joints by eye, and they don't behave like conventional bad joints any
more.

The second one was a Vox combo. This one was reported as "goes off after
a while - tap top to get it back". It actually ran for about 2 hours,
during which time I thrashed the output stage so hard you couldn't touch
the heatsink, and periodically knocked seven bells out of it with the
butt end of a large philips screwdy. At no time did it show any signs of
intermittency. I was actually on the phone to the store that it came to
me from, to check if they knew the owner, and whether he was savvy, or a
numpty, when it went off. Just like that. No provocation. You could then
lightly tap the top of the chassis just about anywhere, and it would
come and go at will. So easy was it to make it do it, you would have
thought that the joint causing it would have been really easily spotted.
I twisted and wiggled everything I could, but nothing made it do it, but
still the lightest tap, and there it went.

Eventually, after a frustrating session of blanket resoldering that did
no
good at all, I came to a power resistor standing up off the board. It
was a component that I had previously twisted. This time I pulled it,
and one leg just came right out of the board. The joint looked perfectly
normal - for lead-free that is - but it had not whetted the resistor leg
at all. How the hell could that take two hours to go bad, not be
responsive at all before that time, and then when it has gone bad, not
respond to twisting, but be so tap sensitive that you could make it come
and go with a feather? I HATE lead-free with a passion.

If it ever finds its way into avionics, be afraid, be VERY afraid ...

Arfa