I am trying to model a PS for a PP EL34
I have a 375-0-375 22R Trafo, a GZ34 for rectifier and want close to 400Vp
I can choose any Caps I want, but I would like to use one or both of the 2H/52R chokes i happen to have. I can but new if I have to...
When I model in PSUD II, if I use a constant current load, I can't do it without a warning that I am exceeding the maximum current of the GZ34 without dropping my voltage too much.
The current load I estimated by the bias current of the 4 EL34s which depending on the plate voltage is between 40ma and 60ma or so. That gives about 200ma roughly with marginal current for the driver/splitter. Is this right?
The current draw as presented by PSUD is pulsing, so it is not continuously over, but I still understand that is not good for the rectifier.
If I model a resistive load, I can easily manage.
I have seen several PS schematics that have values that I am trying that are used in PP amps, they also don't model well as current loads.
I strongly suspect "operator error" here. Any help? Am I worrying needlessly? Am i off base and need a different plan?
Thanks for any help!
I have a 375-0-375 22R Trafo, a GZ34 for rectifier and want close to 400Vp
I can choose any Caps I want, but I would like to use one or both of the 2H/52R chokes i happen to have. I can but new if I have to...
When I model in PSUD II, if I use a constant current load, I can't do it without a warning that I am exceeding the maximum current of the GZ34 without dropping my voltage too much.
The current load I estimated by the bias current of the 4 EL34s which depending on the plate voltage is between 40ma and 60ma or so. That gives about 200ma roughly with marginal current for the driver/splitter. Is this right?
The current draw as presented by PSUD is pulsing, so it is not continuously over, but I still understand that is not good for the rectifier.
If I model a resistive load, I can easily manage.
I have seen several PS schematics that have values that I am trying that are used in PP amps, they also don't model well as current loads.
I strongly suspect "operator error" here. Any help? Am I worrying needlessly? Am i off base and need a different plan?
Thanks for any help!
What are you using for transformer R? Look at the GZ34/5AR4 data sheet for min plate R @ 375V..................it appears to be about 117R for Rs at 375V
The datasheet is here: http://www.mif.pg.gda.pl/homepages/frank/sheets/093/5/5AR4.pdf
where Rs=Rsec +Nsquared X Rprimary + R(additional)
Rsec=secondary R (=22R)
N=turns ratio
Rprimary =primary R of your transformer
and R(additional)=additional R in series with each plate of the rectifier (between the transformer and the plates)=whatever you need in the equation above to get Rs above 117R
Use Rs=117R for your transformer R in PSUD, that will lower the peak current and reduce the risk of flashover. You can increase this series R to lower peak current but it will lower your B+
FWIW, I modeled your PS using CLC with 22u-2H/52R-220U-200ma load and got 393V B+ with the max forward current at 750ma (right at the limit of the GZ34) I used 122R/375V for the transformer parameters.
Also, your driver tubes will eat a little current as well, and may or may not be included above depending on your chosen quiescent current for the EL34's.
If you need more B+, you can always use SS rectification..........😎
The datasheet is here: http://www.mif.pg.gda.pl/homepages/frank/sheets/093/5/5AR4.pdf
where Rs=Rsec +Nsquared X Rprimary + R(additional)
Rsec=secondary R (=22R)
N=turns ratio
Rprimary =primary R of your transformer
and R(additional)=additional R in series with each plate of the rectifier (between the transformer and the plates)=whatever you need in the equation above to get Rs above 117R
Use Rs=117R for your transformer R in PSUD, that will lower the peak current and reduce the risk of flashover. You can increase this series R to lower peak current but it will lower your B+
FWIW, I modeled your PS using CLC with 22u-2H/52R-220U-200ma load and got 393V B+ with the max forward current at 750ma (right at the limit of the GZ34) I used 122R/375V for the transformer parameters.
Also, your driver tubes will eat a little current as well, and may or may not be included above depending on your chosen quiescent current for the EL34's.
If you need more B+, you can always use SS rectification..........😎
Last edited:
Excellent
Thanks so much! That makes sense to me. I did try different values for the transformer R, but guess I didn't put it together in a way I understood.
So, when I model it now, it still can handle a higher load current, I put in 250m, so that is reassuring.
Additional question though. The simulation does show a large current surge, way over rectifier capacity for the first one second, before settling down to just under 750ma. I know that this is only a simulation, so do I need to worry about that or is it an anomaly of simulation.
Side question: How can I estimate the current capacity of my chokes? I salvaged them and managed to figure out the resistance and inductance, but i am a little worried as they were split, one per channel, in the original amp. That makes me concerned that they are only rated in the 150ma range rather than the 250ma range that I need to run just one. I guess I could do the same, and split the PS at the inductor, but I was thinking of running them CLCLC just for kicks, if I can afford the voltage.
Thanks so much! That makes sense to me. I did try different values for the transformer R, but guess I didn't put it together in a way I understood.
So, when I model it now, it still can handle a higher load current, I put in 250m, so that is reassuring.
Additional question though. The simulation does show a large current surge, way over rectifier capacity for the first one second, before settling down to just under 750ma. I know that this is only a simulation, so do I need to worry about that or is it an anomaly of simulation.
Side question: How can I estimate the current capacity of my chokes? I salvaged them and managed to figure out the resistance and inductance, but i am a little worried as they were split, one per channel, in the original amp. That makes me concerned that they are only rated in the 150ma range rather than the 250ma range that I need to run just one. I guess I could do the same, and split the PS at the inductor, but I was thinking of running them CLCLC just for kicks, if I can afford the voltage.
Try the "soft start" option in PSUD. When using tube rectification, currents are not instantly high since things need to warm up a bit to conduct. I'm no expert here but that's what I'd try. IIRC from reading other PSUD posts here, it is a anomaly of simulation.
A similar Hammond 157R choke (2H/57R) is rated for 200ma so your choke is probably a little undersized. Keep in mind that it also needs to handle the additional charging current based on the conduction angle which Morgan Jones 3rd ed. simplifies to Vin(rms)/1386L (for 60 hz)
375/(1386*2)=135ma
So peak current through the choke = 200ma+135ma=335ma
Note that more henries give a lower ripple current, so a 5H choke max current=254ma and 10H=227ma.
More henries in a low R package means a physically much larger (and more $$) choke. You need to keep the R low to stay close to your 400V B+ target.
Check out the Hammond 159T (2.5H/300ma/43R) open frame choke or the 193P (5H/500ma/26R) enclosed choke. They'll both meet the ripple current criteria above and still have a little lower R than you are presently using.
Since both of these have lower R than you existing choke, you should be able to get closer to your 400V B+ target.
A 33u-(5H/26R)-220u-200ma PSUD gets the B+ right at 400V with a 375/122R transformer, and has less than 60mv ripple.
Also keep in mind your other constraint which is the max value of the first cap, which is somewhere in the neighborhood of 40u-60u depending on who's datasheet you look at.
If you are using a new (non-NOS) GZ34, you may want to be a little conservative with the cap and current ratings.
I've also had this problem when using the sim. So I just tried the "soft start" mode and it looks logical. However I think real world startup may be in between the two modes since the warmup time of the output tubes, compared to the charge time of the 1st cap(s) is a determinator of the "softness" of the startup. Some of my tube rectified amps smoothly come up to full voltage on the outputs. Others start up with higher voltage on the outputs, then settle down as the outputs fully warm up and conduct.
Maybe the resistive load is needed when it's for a high-current tube, and CC's are fine for low current tubes when using the sim?
Is a single GZ34 traditionally acceptable in a 4xel34 amp? Would 2 rectifiers be more standard for that (other than SS diodes of course.)
Maybe the resistive load is needed when it's for a high-current tube, and CC's are fine for low current tubes when using the sim?
Is a single GZ34 traditionally acceptable in a 4xel34 amp? Would 2 rectifiers be more standard for that (other than SS diodes of course.)
Last edited:
Graph the diode current. If all you see is a really short spike at the very beginning, you don't need to worry. A few PSUs I have built using 5AR4/GZ34 simulate this way and they work fine. The soft start option will suppress that warning.
The long current surge is more of a concern and that can lead to tube arcing. What size cap are you using after the rectifier?
The long current surge is more of a concern and that can lead to tube arcing. What size cap are you using after the rectifier?
- Status
- Not open for further replies.