Greg Mayman wrote in a message to Roy J. Tellason:

 -=> Roy J. Tellason said to Greg Mayman
 -=> about "POWER SUPPLY" on 10-31-03  20:15.....

 RJT> Nope.  Unless I'm getting a spurious reading from them still being
 RJT> in-circuit. I didn't get anything suspiciously low,  though,  just
 RJT> normal transistor type readings...

 GM> Yeah, that sounds about right. Of course they could be breaking
 GM> down under voltage but that's not so easy to test.

 GM> Connect power, measure voltage across the big caps. Should be about
 GM> 300v total IIRC. Check for high frequency signal from the power
 GM> transistors to the transformer - caution: HIGH VOLTAGE.

 RJT> Somewhere I have an isolation transformer,  but having this board
 RJT> flopping around loose while powered up doesn't have much appeal to it.
 RJT> I've put off that step until I can come up with a comfortable way of
 RJT> dealing with this.  :-)

 GM> On thinking it over I realized there's not a lot to be gained by
 GM> checking the primary of the transformer for signal. If everything
 GM> is alright up to and including the transformer, you will see signal
 GM> on the secondary. OTOH if the transformer is faulty, then you
 GM> probably wouldn't see any signal on the primary anyway.

I'm thinking that this is a startup problem of some sort.  I keep reading
posts about higher-value resistors that go bad,  stuff like that...

 GM> Yes, for powered up testing it really needs to be stabilized, but
 GM> in such a mounting that you can get to either side of the board
 GM> when you need to. Not the easiest thing to do.

What came to mind right off is the thought of clamping one of the heatsinks
in a vise,  but then I remembered the other board -- this unit has a small
board that screws on to the top of both heatsinks,  and appears to almost
contain a small separate power supply.  At least there's a power transistor
on it, a little transformer,  and assorted other bits.  There are wires
that appear to be soldered in to the other board near where primary DC
would be,  and a couple of plugs for the output side of it.  Darned if I
know what that's all about without extensive tracing.

 RJT> In the meantime,  if I can get some better lighting over here and
 RJT> maybe better magnification,  then I'll have another look at it,  and
 RJT> test some of the smaller passive components over on the primary side.

 GM> "Maggylamps" are wonderful things, a 6 inch magnifying glass with a
 GM> circular flourescent lamp mounted around the glass. I wish I had
 GM> one here.

I wish the one I had were more stable.  It was clamped on this other bit of
furniture here and kept falling off...

 GM> Meanwhile I have to make do with a pantograph desk lamp and some 2x
 GM> magnifying spectacles.

I have one of those desk lamps here too,  but it isn't enough.

 GM> I have a pocket sized 10x magnifier somewhere, but it's never handy
 GM> when I need it...

I have a couple of magnifying glasses here on the desk,  of somewhat
different strengths.

 RJT>I can't read the color-codes on the bloody things any more... :-(

 GM> And the surface mount devices make it even worse.

Oh yeah!  And this doesn't even have any of that in there!  But when you
start getting that small,  what's the difference between a 1/8W resistor
and a 1/10W unit?

 GM> I had to troubleshoot a board with those little b*ggers a while 
 GM> ago. That's when that 10x magnifier was VE-E-E-E-RY handy!!!!

I had to actually deal with working with those parts once,  and it was a
real bugger.  This one musical mfr. (Roland) sent me a rom upgrade for this
keyboard I was working on that was supposed to fix the problem. 
Unfortunately when I put it in *nothing* worked.  A phone call revealed
that the problem was due to changing from mask rom to eprom or maybe it was
the other way around.  Instead of jumpers,  there were two small surface
mount resistors that had to be unsoldered and moved to another set of pads,
 nearby,  in order for things to work...

Not my idea of fun!

---