a friend of mine built a LM3886 amp.
wiring the pot the wrong way round he started the amp with full volume.
his speakers just went completely dead and stunk like hell.

he just wanted to try it and took an underrated trannie (120VA,48V ctp.) for both channels.
with all the protection of the IC´s that shouldn´t have been the problem!?
he came along and i tried to find the mistake.
before he took the inputs off the board.
we measured -1.5V DC at the input and 30V! at the output.
after getting the pot back in everything was back to normal.
the chips were placed in a professionally designed PCB.

has anybody encountered problems like this?
is that DC-offset (without the input connected) normal?
pots can always lose contact, go defect... and then you get dead speakers??
was it the trannie?

thanks in anticipation
Firstly, never perform preliminary tests on a power amp with speakers connected unless you a willing to sacrafice them. Use a dummy load.

I don't have a data sheet handy but with 1.5 volts in, 30 volts out sounds about right, given the power supply you mentioned.

1.5 volts at the input with the input floating (no connection) is definitely a problem that needs looking into. There should be a resistor to ground at the + input of the chip to provide a DC path to ground. There may be one internal to the chip and maybe not.

In any event, I believe it is unwise to turn on an amp with an input that would cause it to deliver full power. I've never tried it though and it may be dependent on the amps design.

Properly designed amps never die if the input level control fails - there just ceases to be an output.


2001-06-01 4:53 pm
Try it again with no load (no need for even a dummy) and the input shorted. If you're still getting those big offsets, something is seriously wrong. These parts can fail catastrophically and it smells like it took the speakers with them.
Agree with Bill. As a wise man once said, Good judgment comes from experience, and a lot of experience comes from bad judgment.
yeah my friend was kinda overexcited to test it
we took dummys after that.
but still strange that there is no resistor from the input to ground to eliminate the big offset of -1.5V.
maybe the chips got a little ill with the full volume.
will put a 47k-100k in and measure again.
after the accident i took a look in the datasheet
and even there it says not to take a too big Rin (which was the pot in that case) to cause DC offsets.
guess my friend is getting into speaker-diy now as well :)

actually i like smoke tests when it´s not my stuff.
exploding caps, touching cathode ray tubes with restload and stuff can really losen up the atmosphere at work.
but this time it was a good friend and really nice yamaha speakers.
but **** happens
and humans make errors (which technicians shouldn´t)

the amp works by the way, it´s "just" the two speakers.
i´m sure my buddy will soon direct a thread to speakers forum to ask which drivers make a nice substitute :)

keep on burning (your ears and not the coils)