diyAudio (
-   Tubes / Valves (
-   -   Hrrrrmphhhff. PSU design error and how to cope with it (

Rundmaus 28th September 2012 01:19 PM

Hrrrrmphhhff. PSU design error and how to cope with it
Hey there,

due to the idea of building a one-channel test setup for my amp, I made a design error I now have to cope with. Situation is as follows:

The PSU uses one custom-wound power transformer with separate windings for both channels, each channel gets its own GZ34 rectifier and filter chain.

Now, guess what does *not* exist twice on my carefully planned (and ordered/shipped) power transformer... Yes, right, the 5V winding for the rectifier heater. F**k! :eek: :eek: :eek:

As the GZ34 has its heater internally connected to the cathode, I have the following options:

a) Live with the fact that the rectifiers are paralleled. Stretching SOA to its absolute maximum, one GZ34 should be able to supply the whole B+ current of 250mA (at 480-0-480 into L-C-L-C) in case one of the rectifiers fails. So failure of one tube would lead to a detectable (protection circuit) sag in B+, not to the immediate destruction of the second rectifier.

b) Rewire the rectifiers with both anodes in parallel, so each of the GZ34 acts as a single diode in the full-wave rectifier. Now we're safely within current limits, and the failure of one rectifier leads to half-wave rectification with B+ sag and audible hum. BUT: Startup of the amp might be a little hum-my if both GZ34 don't come up exactly the same instant.

(c)) Not really an option: Use only GZ34 with the internal connection burned away intentionally :rolleyes: :rolleyes:

What do you guys think?


SY 28th September 2012 01:24 PM

(d) Buy a small toroid with two 5V windings, tuck it in a convenient spot in the chassis.

Rundmaus 28th September 2012 01:36 PM


due to a lot of ideas that came up quite late, I am slowly but definitely running out of convenient spots in the chassis ;)

Your idea was the first that came to my mind, but I hoped I'd get around putting another lump of iron under the hood...


andRookie 28th September 2012 01:51 PM

Use ss diodes and the gz34 socket for some kind of delay tube for B+?

pointy 28th September 2012 01:56 PM

can you put windings over the transformer windings that you already have.

(not safe but could work!)

yagoolar 28th September 2012 01:56 PM

Add another 5V winding to the trafo?
(Pointy was faster)

12E1 28th September 2012 02:14 PM

In my experience there are acoustic benefits in keeping the rectification and subsequent filters separate for the left and right channels. (I have that set up on my current amp.) It greatly helps channel separation.

The transformer that I have only had one 5V winding but it had a surplus 6.3V one, so I used that for the second rectifier with a resistor to drop the 1.3V (ideally 0.72ohms - I used two 1.5 ohm 5W wire-wound in parallel). The actual resistor dissipation is just over 2W in total.

If you don't have a spare 6.3V winding, don't be tempted to use one that feeds the other valves since it will be directly connected to +HT of course.

Other options are as already suggested.

Vinylsavor 28th September 2012 03:56 PM


If your heater winding is specced for enough current for two GZ34, you might be able to use two 6BY5 instead. They only need 1,6A. If your heater winding delivers a bit over voltage due to the lower load, this might be an option. They have no connection beween cathodes and heater

Best regards


DF96 28th September 2012 04:08 PM

If your two HT secondaries are identical then parallel them and feed two GZ34 in parallel. This like your option (b) except the anodes are arranged differently. Avoids hum problem during warm-up, but at the expense of GZ34 overloading during warm-up during the few seconds when only one may be hot enough to function. No worse than (a) though, and only for a few seconds.

wa2ise 28th September 2012 06:17 PM

maybe you can make use of a 5V switching power supply to feed the 2nd rectifier. Or you could use TV set damper diodes, those have separate heaters and cathodes, and H-K voltage ratings that are pretty high. Problem with that is that damper diodes are just half wave.

Is the power transformer a toroid? If so You could add a winding easier than if it's a regular transformer.

All times are GMT. The time now is 08:58 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2