Tube PSU for PP Amp

[prelius]

Hi folks,
I've been using PSUDII to design PSU's for SE amps and it never failed me. Recently, I built a PP amp, and used PSUDII to model its power supply. The screenshot is attached... The power supply is supposed to provide three voltages: 395V, 310V and 250V. From schematics I am pulling 395V from C3; 310 from C4, and 250 from C5. When I powered the amp and measured the B+ all voltages were 30% less than what they should have been...
I can rule out the hardware: the tranny gives out 350-0-350 as it should, and both chokes test good. Any ideas why a built PSU gives voltages 30% lower than its model in PSUD?
The amp is a PP 6L6 amp, largely based on HK250 schematics...
Thanks, Paul.

 

[DF96]

Is C1 OK? If open circuit, or very low in value, you have a choke input supply which could easily be 30% low.

 

[gorgon53]

You state 3 different voltages you want to pull but no currents. How are we supposed to know what the individual load current-dependend voltages schould be?

 

[gorgon53]

Put appropiate current loads into your sim. A 5k load at the 250V output is hardly representative for your 6L6 pp amp.

 

[prelius]

Folks,
Apologies for the delay having a look at the forum, was busy at work..., and thank you for the replies!

Anyway, I figured out what the problem is... Gordon, you are absolutely right... My sim is incorrect, because the amp is pulling significantly more current 5K Ohm load represents. When using PSUD for SE amps I've always used CCS to load the power supply, but in a PP amp power consumption varies, so I used the resistive load with the default R value provided with PSUD, without really thinking what I was doing...

I will re-run the sim tomorrow with properly calculated load.. One question regarding the load calculations for PSUD...

Suppose the amp draws 100mA at 395V, 75mA at 310v and 50mA at 250V, representing 3.95K, 4.1K and 5K loads respectively. The three loads are in parallel, so the total load on the power supply will be 1/(1/3950+1/4100+1/5000)~=1.5K. Is this a correct way to calculate total load for the PSUD?

EDIT: I noticed one can insert "Current Taps" in PSUD. Is this what I need to use to specify current draw at each filter?

Thanks, Paul.

 

[gorgon53]

use current taps

 

[SY]

It's also VERY important to step the current to determine if there's an unintended resonance! This happens more often than you'd think in multiple LC filters.

 

[gorgon53]

in this example Q is well above 0.5, the thing rings like a bell in a pp amp