Bridge rectifier & arcing 5u4

[6267]

I am trying to prototype a bridge rectifier. It looks right to me, but when I plugged it in, I got arcing in the rectifier tubes.

The power transformer has two B+ leads, a 300-0-300 and a tap off that for 250-0-250. They share the center tap, in other words.

Previously, there was a full-wave tube rectifier for each of these windings, which worked great. The center tap was grounded to the first filter cap.

I am trying to make the 300-0-300 a hybrid bridge rectifier, so that I can unground the centertap and use it to provide a half-B+ screen supply. I wired two 1n4007 with their cathodes at each plate of the 5u4, and their anodes at ground. I ungrounded the center tap. I left the 250-0-250 winding disconnected until I put a bridge there too. I plugged it in, and got arcs in the tube and then it blew the reset breaker! Resistance from Cathode of 5u4 to ground is infinity, resistance from anode of 5u4 to ground is infinity, resistance from ground to anodes is the voltage drop of the SS diodes. I am fairly sure there are no B+ leaks to ground, resistances look good, so do filter caps. Anyone have any experience with hybrid bridges? Are my tubes coming up too slow?

[6267]

Here's a rough picture of what's going on, the 250-0-250 winding is not shown, neither is it's rectifier tube plugged in right now. I left the heaters off the drawing for convenience.

[Steve Conner]

You do know that changing to a bridge rectifier doubles the B+? If it was 400V before, it'll be trying to make 800V now. In other words, the center tap isn't half B+: it's the same voltage it was before, and the new B+ is twice that.

If your filter caps aren't 800V rated, blam. The PIV on the diodes isn't any worse than with a full-wave circuit, though.

Also, connecting several bridge rectifiers to different parts of a tapped winding never seemed like a good idea to me. There are two possibilities, either some of the diodes never conduct (you get different voltages to what you expected) or part of the winding gets shorted (you get fire)

In your case, I think the diodes from the 250V taps to ground would never conduct, because the ones on the 300V taps would conduct first.

[6267]

The 800v would explain the lightning!

This is the first time I've tried to design a bridge in place of a full wave, and I guess I somehow thought there was a difference between a bridge and a full wave doubler. I gave up and decided to use a standard full-wave center-tapped as it was before, no need for B+ quite as high as it was going. I just made my low voltage screen supply with a B+ divider and bypass/filter cap. I am fairly sure the best ways to learn things often involve surprises, like tripping the 15A breaker in the wall panel making kilovolts accidentally . No more late night experiments with reverse biased diodes!