Troubleshooting odd power supply problem

I'm working on a unit for a friend that displays absolutely the strangest problem I've ever seen. The secondary side of the transformer puts out the proper AC voltage (23vac), but the diodes are not rectifying. There is no or little DC voltage--should be +/-18 volts, but I'm lucky to find +/-2 volts past the diodes. All of the utility DC is there--the LED's and relays are switching. I've changed out the rectifiers to no avail. Filter caps look good, and the transformer isn't pulling down the fuse. I've seen some odd failures before, but this one has me stumped. Any ideas? Maybe a shorted opamp pulling down the rails? (Now that I think about it, I saw a Mesa amp with a problem like that once...)

The only really weird thing that can happen is that the laws of physics have changed while you were sleeping last night.
Having safely eliminated that, you can put away the "strange" concept and figure it out.

Think outside the box.

Like realty checking the meter by placing it on a 9 volt battery to make sure it is working.

Then, I would disconnect everything but the diodes and looking for pulsating dc on the dc side of the rectifiers.
Is this a bridge circuit or center tapped full wave or simple half wave?

It's a center tapped full wave rectifier. At first, I thought my meter had a bad battery, which makes it do funny things like this. I replaced the battery and still got the same reading. I checked it with a second meter, and got the same reading still. I checked my opamp theory, but that lead nowhere. I'll pull up the diodes and see where that leads.

tHEY COULD BE OPEN. mOST METERS HAVE THOSE DIODE CHECKERS.
but you already rrplaced so scrub thsat idera.

Let me see if I understand you. You say the "utility" stuff is working -- relays and LEDs. Are they on the same supply as the audio stuff?

Also: when you pulled the diodes and replaced them, did you measure the diodes with your meter to see if they were blown?

Peace,
Paul

Pstamler is headed the right direction, Check the logic stuff to see if it's on a separate winding. i bet it is. I've seen times where you read a certain voltage but then actually try to pull current and nothing, but this is usually with types of MOSFETs in my case..

Well, in terms of the utility power, there's about 2.5 vdc present. I don't know if that's what it's supposed to be, but it's enough to operate all of the relays and LED's properly. What puzzles me is that it's on the same winding of the transformer as everything else.

I went so far as to pull all of the filtration and regulation out of the PS, but there is still no DC on the output of the naked rectifier stage. I double checked the transformer output and grounding, and everything still looks as it should--23vac on each side heading into the full-wave stage. There's nothing after that. The diodes that I initially replaced test good.

Usually, in a situation like this, it ends up being operator error--I mean, it's like Occam's razor, right?

What happens if you resistively load the trafo outputs in question? Meters are high-Z devices, relatively; a little C is enough to give a reading into a 10M input, but there may be no way for many mA to flow.

The other thing is to put in some a.c. from another source to your bridge/caps et al. to absolve or implicate them.

You broke the problem already down to the smallest pieces! If it's not the transformer misbehaving under load it can even be a "near short" on the pcb...
I'm very interested to hear the solution for this mystery from you when you found it...

Looks like the transformer is faulty. I replaced the AC source in the circuit with one I had on hand, and everything looks fine now. I've never seen a failure like that before...I'm still puzzled.

[quote author="featherpillow"]Looks like the transformer is faulty. I replaced the AC source in the circuit with one I had on hand, and everything looks fine now. I've never seen a failure like that before...I'm still puzzled.[/quote]

I think as I suggested that you had an open winding but were seeing displacement current from stray C, plenty enough to give a false reading on your high Z meter.

I think as I suggested that you had an open winding but were seeing displacement current from stray C, plenty enough to give a false reading on your high Z meter.

Click to expand...

Oh, I get it! That makes complete sense once you phrase it that way. I was wondering how I could possibly see something like this.