Rectifier snubber failure

[xtian]

Another new one to me. A ceramic cap, in parallel with one of the diodes in the FWBR, shorted in a Fender Deluxe Vintage Modified, blowing the mains fuse. This amp is a PITA because all the components are on the underside of the main PCB. But once I got the amp on my bulb limiter, I was able to use my thermal camera to see the hot capacitor easily. I clipped out all four snubbers, and the amp is back in action.

IMG_1735.jpeg

[solderhead]

A thermal camera would be nice to have.

Are you planning on replacing the failed snubbers, or is your plan to accept the noise and run naked?

[Stevem]

I would not bother with replacing all the.01 uf snubbers, I would up grade the diodes to a fast recovery type.
Then there's 4 less components to possibly fail.

If this is one of the Fender amps with 1/4 watt PI section plate load resistors then while your in there make them 1/2 watt.

[xtian]

I did not replace the snubbers. Amp is quiet at idle, no buzzing or humming.

[Snicksound]

I did not replace the snubbers. Amp is quiet at idle, no buzzing or humming.

Can't say I've ever heard a difference with and without.

I'm guessing it's more about EMI since commercially sold amps are subject to a lot more regulations nowadays.

[R.G.]

Caps across the diodes help sometimes, sometimes not. Standard diodes slam off when the stored charge in the junction gets depleted and it will again hold off voltage. This edge is so fast that it has lots of RF harmonics. The fault syndrome is when the wires and traces to the diodes have an RF resonance that rings with the transient and radiates the RF. This can be an EMI issue if it's in the RF bands that get tested, but it can also radiate to the amplifying parts of the amplifier.

Sometimes. Maybe you get it, maybe you don't. On PCBs it's fairly predictable where the trace resonances will be, but not for hand-wired. Just a cap across the diodes are not great, either, because sometimes a cap can merely tune the wiring resonances to a lower RF band. It needs an RC snubber to damp the ringing along with sucking the energy out of the transient.

It's not surprising that clipping the caps out didn't necessarily cause hum or hiss. Nor is it particularly surprising that a high voltage cap fails once in a while. Kind of "hey, look at that" instead of "holy cow! a unicorn!"

Me being me, I'd have used this as an excuse to replace the diodes with FREDs. This would have had the same result, but I'd feel better knowing that I'd over-fixed it.

[xtian]

I'd have used this as an excuse to replace the diodes with FREDs. This would have had the same result, but I'd feel better knowing that I'd over-fixed it.

Except for the fact that this is a Fender production PCB amp, meaning the traces are paper thin, and every component swap is a gamble.

[solderhead]

I was reluctant to write a post contradicting the terminology used on the forum, but since RG opened that door:

When you see caps like that placed across a FWB the role of the caps is to create a very short loop for the high di/dt current at diode turn off. This lowers emissions which may be picked up by nearby circuitry.

Believe it or not, I have a textbook entitled, "Snubbers". Written by a guy who spent his whole life designing snubber circuits. I picked it up for $5 in a used bookstore for reference use. It's a boring read.

There are experts who claim that placing true RC snubbers across each diode in a rectifier circuit is a waste of both time and components. they insist that just placing cheapest possible caps on the diodes, and designing a snubber to go across the transformer secondary is all that's needed. Of course FRED are a better solution than 1N4007 if you're worried about noise.

[maxkracht]

Just found different failure in the same area of another fender. Caps were too oxidized and didn't take solder, one pulled right out, the others didn't look great. Not sure if this would have caused problems down the road.

Is there any benefit to using FRED vs UF diodes? I got a large stock of uf4007 a while ago that I have been using by default.

[xtian]

Is there any benefit to using FRED vs UF diodes?

Not in my experience.

[R.G.]

I agree - probably not.

FREDs were all the buzz for a while when the 1.2kV 8A? 10A? devices came out in TO220s at a reasonable price. The UF4007s were a bit later, and obviously rated for lower currents. I don't know how the recovery times and softness of the recovery compare between the two, although it would be interesting to see. There are probably many more Fast Recovery Epitaxial Diodes now at a range of voltages and current ratings.

I suspect that the fast recovery prevents a lot of reverse voltage from building up while the junction is being swept out before slamming shut, which would mean that fast is good. Soft recovery would improve that even more.

[maxkracht]

Thanks guys, I appreciate the input. I think I looked into buying some FREDs a while ago and couldn't find anything with the specs and price range I was looking for. Nice to know i'm probably not missing anything.