power supply PLEASE HELP

[soundbadger]

hello there
i have designed a power supply using psud2 (see attached jpg) , it is to power 2 kt88 tubes and 2 pre-amp tubes (undecided on exact type), i just wondered if someone could take a look and tell me what they think. is there anything i should do differently. i need a power supply to give 2 different voltages for the power and preamp tubes this seemed the easiest way.

the transformer resistance of 150r is just a guess but the chokes 82r is a hammond 193j.

according to psud this circuit provides 410v at 160mA and 345v at 12mA

any help appreciated,,,,
soundbadger

[zigzagflux]

So this is a single ended amp?

If it were me, I'd remove R1 and C3. If you are aiming for a specific ripple at the output stage, increase C2 as needed. Look at the current through L1; make sure it doesn't go near zero for all operating conditions.

Do you have any requirements for ripple on the preamp stage? If it doesn't have to be extremely low, you could also replace the RCRC with a single RC.

[soundbadger]

i had thought about replacing the rcrc with a single filter.
but with regards to removing r1 and c3 the ripple going to the power tubes (the 160mA current tap) is higher than i would like when i just use the c-l-c which is why i added the extra r-c.
is this acceptable, / would you have done it differantly ,,

as always any advice greatly appreciated

cheers
soundbadger

oh and yes it is an SE amp design

[zigzagflux]

Well, I see very very little difference in ripple between C2-C3 and C4-C5.

C2 measures 395V with 47mV ripple.
C3 measures 360V with 25mV ripple. Not extremely different, and you ended up adding 200 ohms of resistance in series with the bulk storage of your output stage to reduce ripple by half. The actual output stage essentially has only 10uF to work with; the rest comes from a high impedance source; not good practice for an output stage.

If 360V is your target, I would place something like a 4uF cap for C1 and 200uF cap for C2. Any resistance to tweak your output I would add immediately after the tube rectifiers, and it doesn't need to be much.

Similar with C4 and C5, one stage might be better with a 20uF cap. Extremely low ripple would require another choke. Up to you.

Most importantly, you need to model your transformer correctly in order for PSUD to predict correct output voltages. Do you actually have the xfmr?

[soundbadger]

yes i know the transformer but i dont know its resistance to tell psud2
its a hammond 300bx

[Jeb-D.]

The actual output stage essentially has only 10uF to work with; the rest comes from a high impedance source; not good practice for an output stage.

This is true. It will cause your frequency response to become non-linear as that 200 ohm resistor will become part of the output stage. When frequency drops below where the 10uF(C3) cap becomes effective.

If it were me, I'd remove R1 and C3. If you are aiming for a specific ripple at the output stage, increase C2 as needed. Look at the current through L1; make sure it doesn't go near zero for all operating conditions.

I'd have to agree.

[soundbadger]

"and you ended up adding 200 ohms of resistance in series with the bulk storage of your output stage to reduce ripple by half. The actual output stage essentially has only 10uF to work with; the rest comes from a high impedance source; not good practice for an output stage."

could someone please explain this
cheers

[zigzagflux]

Well, the idea is to have a good deal of capacitance right at the output stage, in your case the B+ is the top side of the SE transformer, with circuit common at the bottom of the cathode resistor.

With large capacitance available, the current requirements of the amp, at all audio frequencies, are supplied by the stored energy in the capacitor. The capacitor is only allowed to be charged up by the wall voltage 120 times every second (60Hz * 2), so the more capacitance you have right at the B+, the better off you are.

The larger the capacitance, the more energy storage you have. As far as the output stage knows, it "sees" a voltage source with a low source impedance, this impedance being primarily the ESR/ESL of the final stage capacitor. Typically, this source impedance will also decrease as you increase the capacitance, ignoring the electrolytic vs. film cap debate. Ideally a zero source impedance at all audio frequencies is pursued, so large capacitance values are selected. There is a practical limit to this goal, of course; no one puts in 1 Farad of capacitance for a 200 mA load.

You are essentially providing a 10 uF cap with let's say, 10 milliohm ESR. That's your voltage source. The larger 100 uF caps you have in your earlier power supply stages are isolated from the B+ portion by the 200 ohm resistor. 10 milliohm in parallel with 200 ohm is 10 milliohm, so the 100uF capacitor essentially does not contribute to the performance of the audio circuit. All it does is serve to reduce ripple.

We are recommending eliminating the resistor to allow your B+ to effectively use the energy in the 100 uF capacitor.

Make sense?