DCAudioDIY.com

Sometimes, when trying to complete a project in a short time. mistakes are made. I may be the king of them.

Last week I made a post looking for a transformer to replace one that had given up the ghost; that is if the ghost is gray smoke.

Just recently, my Aikido/DH500 project was completed. After maybe 10-15 hours of glorious listening, the 1963 Capehart console power amp transformer in use went "pffft!"

Rarely does a single mistake cause the real damage, and so it was in this case. I had a 1 amp fuse in series with both the heater and plate transformers. That should have even been massive overkill, considering that the B+ load is about 30mA @ ~300 volts DC. The heaters add up to 1.1 amps at 12.6 volts DC. Obviously, the AC load is greater, but not a significant amount.

So, as I was listening, suddenly the bass disappeared. A quick look at the Aikido section of the amp showed dark tubes. Ruh roh. So what did I do? Yup, bigger fuse. Now you might think that as an electrician with some experience, I'd have known that was a bad idea. Wrong. After turning the power back on that time, and within 30 seconds or so, the gray smoke poured out along with a lot of varnish on my nice new clean wood chassis (yup, soaked in). That actually was mistake #3.

To restore the unit to operation I found a much bigger and brand new transformer. Still, a console transformer providing B+ to a quad of EL84s and 2 12AX7s in a power amp along with the umbilical feed to the tuner section, should have lived forever in an application where only its high voltage winding was in use and that at 30mA. The new transformer can deliver 220mA at 460VCT.

It took a couple of evenings of pulling the amp apart, re-mounting things, covering now unused holes, etc, to get her back into fighting form. Powering it up for the first time, on the Variac, I saw an issue in the B+. The rectifier is a full wave connection using UF4007s. Through one, voltage was building nicely. Through the other one, very low voltage. Something is amiss. You see, I'd looked carefully at the power supply and amp circuit boards and couldn't see anything wrong. No signs of something shorted or blown up.

So now, once again the power supply board comes out and the testing starts. Rectifier #2 shows close to ZERO ohms in both directions! And then I saw it: mistake #1. I'd bypassed the diodes with Mylar caps. Mylar caps with a 100 volt rating! Yikes, how could I have been so careless? I was in a rush, the caps were in one of the "cap" boxes, and here we are. But in this case, the little green cap didn't blow up, it short circuited! That's what killed the transformer. The UF4007 was fine.

Mistake #2 was not testing EVERYTHING that could draw current before fitting a new fuse after the first one blew.

Now there are a couple of 630 volt ceramics there and all is well. But only after the needless death of that little Capehart transformer that powered my Dad's console and gave us so much listening pleasure for a couple of decades.

Don't be like me. THINK every move through!

Stuart

But you were able to find the problem. I think that's part of the DIY concept. So bravo to you!

Even after all these years I still occasionally make a dumb mistake. As with you, often when I am in a rush. It's a human condition.

Nobody's perfect, but you always take these little setbacks as opportunities to learn.

Mistake #1a: Not using an RC snubber in this application, which allows the use of 'fail open" resistor types in series with the cap.
Plus, if you want to absorb any transients from the diode switching, you have to have someplace for the energy to go. It's kind of hard to send it to ground in this area of the circuit, so just dissipate it as heat in the resistor. Two birds>> one stone.

I am absolutely my own worst enemy. It's been proven many times.

Me too
to err is human and I am very human

Pelliott321 wrote:Me too
to err is human and I am very human

There is a saying in Engineering, "If it works the first time, you obviously did something wrong."

Cogito wrote:

Pelliott321 wrote:Me too
to err is human and I am very human

There is a saying in Engineering, "If it works the first time, you obviously did something wrong."

With every failure experienced is at least one lesson learned, or re-learned if forgotten.

In this case, the re-learned lesson is that a capacitor is capable of not just failing, but failing shorted!

About 12 years ago a 4.0uF 250V Multicap used in an output coupling position of a crossover (150V at its input), going to the input of a Hafler DH500 failed shorted.

I'd replaced the Mylar input caps of the Hafler with bipolar Black Gates, rated 50 volts. The Black Gates passed the DC right along. Nearly fried one of my speakers.

David Berning told me something I didn't know then: when over-voltaged, an electrolytic cap can pass DC.

Those caps are still in my possession, and they test fine. Yikes!

Stuart