diytube.com

With a tube rectifier there are no switching glitches so not sure what the comments above are about.

Even when you use the solid state "relief" diodes in the anode feeds to the tube rectifier you still do not get switching glitches (relief for the tube rectifier from handling the full Peak Inverse Voltage (PIV)).

Switching glitches arise with SS diodes when the voltage reverses polarity. The forward biased junction of the SS diode has junction capacitance which has a charge. As the voltage reverses you get a "splat" of current in the reverse direction to "sweep out" that junction capacitance charge before the SS diode turns OFF. The size of that junction charge is what determines the severity of that noise spike and it is what is different between normal SS diodes, ultrafast soft recovery diodes (approx 1/5th the junction charge) and Schottky Diodes (about 1/10th or less the junction charge) - Slightly simplified explaination but covers the "essence" wihout getting into solids state physics..

You need a current path for that reverse direction current "splat" and the tube rectifier simply does not allow it. That is, the reverse current "splat" is blocked by the tube rectifier.
That also means that your SS relief diodes don't need to be anything special, bog standard "garden variety" 1N4007 or similar diodes will do. All you need to worry about is their volatge (PIV) rating.

Cheers,
Ian

Geek wrote:I've been using PIV relief diodes in series with the AC side of the tube rectifiers as of late. This seems to work and you still get that tube rect. sound while the rect. survial rate being much, much higher than without them.

Cheers!

I was thinking about that. Could you use any full wave rectifier with UF4007s to increase the PIV? Like a 3DG4 with UF4007s in series and rectify 600 volts?

I use 2 x UF4007 in series per anode for safety. But Gingertube is correct in that any garden variety including 1N4007 will work.

Matter of fact, if it starts to "buzz" with 1N4007's as relief diodes, it's an indicator your tube has packed it in.

Never used the 3DG4 before, but a quick look at the specs shows it to be a bit of a "tank", so I don't see why not.

Cheers!

Geek wrote:I've been using PIV relief diodes in series with the AC side of the tube rectifiers as of late. This seems to work and you still get that tube rect. sound while the rect. survival rate being much, much higher than without them.

Gregg, are you pretty much using a pair of UF4007s in series (four total) exclusively? Cheap, simple and good for sure.

Shannon

My PSU Designer simulations look pretty close to what I want for B+ level and ripple.
Obviously simulating the load is tricky but I'm going Class A so that simplifies it some.
I have all the parts in hand now, including a decent range of values for the filter components if I need to adjust or just want to experiment.
I'll report my results along with a schematic once I run through some tests.
Thanks for all the input.

Hi Shannon,

Shannon Parks wrote:Gregg, are you pretty much using a pair of UF4007s in series (four total) exclusively? Cheap, simple and good for sure.

Yep! I demonstrate this in my posted Procyon schematic, too if anyone needs a reference

I am looking at a new Edcor transformer to use with the above 5U4GB circuit discussed.
This particular one has 2, 6.3V center-tapped secondaries instead of one 6.3 and one 5V.
Would using a 6.3V secondary instead of a 5V secondary for the rectifier heater be an issue?

Yes, don't do it.

Even with a dropping resistor, not all 5U4 types draw the same current. And what if you want to drop in a 5R4 or 5AR4 later?

It's no thing to add a little 5V, 3A Triad transformer. They're what, $10?

Cheers!

Thanks Geek.
I should have looked at the datasheets with a better eye.
All manufacturers I can find state "5V filament" and show absolutely no +/- range, so I suspect there is a reason for 5V only based on resistance/current.
I do not want to overstress the tube and limit the type tube I can roll just to use this transformer.
Space in my chassis may limit the addition of a separate filament transformer but I suppose that might allow me to reduce the size of the power transformer a bit.
So if I went with a separate filament transformer like this: https://www.edcorusa.com/lvp5-3-120 would the rectifier filament/high voltage output connections still look the same as second pic in OP?

Take the HV output from the centre tap and you're ready to go!