valve rectifier in PSUD

[jrenkin]

I am confused about rectifier current and how to interpret PSUD.
I am using a GZ34 rectifier to a CLC to a PP amp drawing max 250mA.
My Transformer is 417V,2 x 150R to the rectifier, the first cap is 10u, inductor is 8H 50R, the second cap is 100u and the load is a current source at 250mA modeled in PSUD.
Simple, right? The actual cap values may change, but that is not the point.
The rectifier requires a transformer resistance of 2 x 125R, and the first cap cannot exceed 60u, so I have both of those parameters covered, yet PSUD tells me I have exceeded the current capability of the rectifier and the initial peak current is huge (under 1/2 second). Even after turn on, the ongoing chart per PSUD is a series of current peaks that exceed the rating of the rectifier. I couldn't begin to approach a 60u cap with out PSUD telling me I will blow up.
I do not know how to interpret this. The total current draw of 250 mA pushes the limits of the rectifier, but shouldn't the input resistance and first cap values protect against this current inrush killing my rectifier?
What am I doing wrong?

[Rod Coleman]

This happens to me quite a lot. Sometimes it is combinations of current sources and voltages that seem to not get along together. In this case, I usually try substituting a resistor instead of current source.

THe other way I get this error comes from not getting the cap prefix right. Check that the schematic really shows uF and not F!

If it still wont work, rename the file supply.txt and load it here & we'll try it!

[jrenkin]

Thanks. I guess the real question is, if I know my transformer resistance and first capacitor are in spec per the tube data sheet, do I even need to worry about what PSUD tells me in regard to these current peaks, or should I ignore it and just use PSUD to model the other data?
In terms of my build, I may end up paralleling two GZ34s so i am not pushing the limits of the tubes anyway. For a PP like this, where I want to bias the EL34s around 50-60mA (x4 = 240mA) and need about 20-30mA more for the front end I probably should be using solid state, but i don't want to....

[Rod Coleman]

I think it is well worth making it work, because you are near the top of the dc current range for GZ34. A lower 1st C will give you more margin on rectifier lifetime than 60uF. You can make up for the increased ripple by using a larger 2nd C.

The new production GZs have a patchy reliability record.

[jrenkin]

Originally Posted by Rod Coleman

I think it is well worth making it work, because you are near the top of the dc current range for GZ34. A lower 1st C will give you more margin on rectifier lifetime than 60uF. You can make up for the increased ripple by using a larger 2nd C.

The new production GZs have a patchy reliability record.

Hmmm, I am not too worried about ripple, the input stage will be regulated and the output should take care of normal stuff. But reliability and all, I think I will just parallel the tubes for the heck of it. It will bring my parameters way down per tube, decrease sag, look cool and allow easier rectifier tube rolling when my GZ34s crap out
Besides, I have got two....

I still don't really understand how to interpret PSUDs rectifier current information. How am I getting such huge peaks when I am within all the parameters of the rectifier? Either I don't understand and I should be looking at RMS values not peak or PSUD emulates incorrectly in this regard or the rectifier recommended values are not really that much protection.

[TheGimp]

one caviat to lowering first C and increasing second C will also lower the final voltage.

[boywonder]

I just modeled it based on your info above and did not get any warnings in PSUD, although 250ma seems to be pushing the envelope on the rectifier. I used 310R for the transformer R (150R X 2 plus 10R for windings-probably more severe than real life)

Seems like the RMS current value is not too bad.....

[Rod Coleman]

Are you aiming for class AB? If so keep in mind the current is more like 4 x 140mA at full power (425V B+) if the bias is 4x 50mA

[jrenkin]

Originally Posted by boywonder

I just modeled it based on your info above and did not get any warnings in PSUD, although 250ma seems to be pushing the envelope on the rectifier. I used 310R for the transformer R (150R X 2 plus 10R for windings-probably more severe than real life)

Seems like the RMS current value is not too bad.....

Yes, but the model for the transformer is 50R (this is a calculated impedance from PSUD's calculator) for the transformer and adding 100R before each plate of the rectifier. This actually models as 150R, not 310R. I can't recall where I read to do it that way, but I did. Is it wrong? I will try to look it up again.
I can add more resistance to reduce the current load, but then I reduce my B+ too much. I want around 430V.
I am getting the impression that the transient peaks in PSUD are something to worry about and if I want my B+ where I want it and my current too, I need to parallel rectifiers.

[jrenkin]

Originally Posted by Rod Coleman

Are you aiming for class AB? If so keep in mind the current is more like 4 x 140mA at full power (425V B+) if the bias is 4x 50mA

I am actually aiming for class A.
But well taken, I may want to parallel rectifiers anyway to make sure I have plenty of headroom.