Tube PS question: Acceptable ripple

[Guiseppe]

I've been working on a 6D22S-rectified PS (for a 6C33C PP) in PSUDII for a couple of weeks, and have managed to get the voltage ripple down to ~1.6mV, but not lower (at least not without voltage and current warnings). Is this generally considered "acceptable", or should I continue working on it? I've been beating my head against a wall trying to get it flat - then realized that it would be a good idea to ask here before putting any more time into this...

This is my first attempt at designing a PS from scratch; I'd like to get it right.

Thanks, folks.

[DualTriode]

Hello Guiseppe,
Tell us more!
What do you have going; capacitor, resistor, choke? Go too far past a Pi filter and start adding much resistance you get better ripple rejection at the expense of greater total impedance. Your power supply starts getting soft and acts like an unknown plate resistor.
DT
All just for Fun!

[SY]

Giuseppe, to determine what is the necessary ripply, you start with two variables: what is the required signal-to-noise at the output and what is the power supply noise rejection of the signal circuit. Then you can work backward to determine power supply noise.

[metalsculptor]

Ripple is better expressed relative to supply voltage then it can be related to the supply rejection of the amplifier that it is powering. Assuming a 400V supply the ripple is -108 db With an amplifier of 60db or better supply rejection I would be quite happy with a full load ripple of -40db as this would give 100db or better signal to ripple at full power. Your application may differ so it pays to define your goals everything is a compromise.

[Guiseppe]

DT:
Schem attached. Couple of notes: R1 is a NTC thermistor (R at 25ºC=200Ω; R at Current=5.001Ω; max current 500mA); I'd prefer to have it on the 0V leg (per advice received here), but PSUD won't allow it. Choke is a Hammond 193N (3H, 500mA). The power trafo is 350V/500mA. The PS will be supplying one 6C33C at 200-225V, 400mA max.

Not sure if this is gonna work or not. I'm concerned about the max current at the choke. Any thoughts?

Sy: Still working on the amplifier design. At this point, all I know is 6N1P-EV input/6N6P drivers/6C33C-B outputs in PP.
I started playing with PSUD, and couldn't let it go...

Thanks for the responses. Everything I've read says work backwards from the OPT - I need to listen, and get the amp designed.

Guess I'm putting the cart before the horse... but PSUD is fun to work with.

[quikie22]

Try switching C1 to around 40uF and let us know the results.

[SY]

Originally Posted by Guiseppe

At this point, all I know is 6N1P-EV input/6N6P drivers/6C33C-B outputs in PP.
I started playing with PSUD, and couldn't let it go...

Thanks for the responses. Everything I've read says work backwards from the OPT - I need to listen, and get the amp designed.

Well, this will be much easier than you think. It's a power amp, so voltage levels are pretty high. It's push-pull so there's ripple cancellation there, too. I think it will be difficult to screw this up.

The main things to be concerned about are keeping the ripple current reasonably low (don't use a monster cap at the input of the filter) and the way you actually do the grounding (don't induce ripple currents into the grounds). Past that, it's pretty simple. Don't get carried away!

[Guiseppe]

Originally Posted by quikie22

Try switching C1 to around 40uF and let us know the results.

IFRM hits 1.8A. Max for that bridge is 1.0A, so this is not good.

Ripple IS reduced to ~0.3mV, though...

[Guiseppe]

Sy:

Thank you, sir.

...don't use a monster cap at the input of the filter...

I take it C1 at 0.47uF does NOT qualify as a Monster Cap?

Quick question: Can I use the thermistor in the R1 position shown on my schem, or should I move it to the 0V leg? Does it really make any difference? I haven't tried to run the circuit without R1 - need to see what happens with that gone...

[SY]

Why not put the thermistor in the primary?