Help me build my Choke I/P preamp PSU

[andyjevans]

OK guys - cooking time. I'm planning on a stereo preamp with on each channel a 6SN7s in diff pair over a CCS, probably with 47k anode resistors - though open to ideas. I could use 76 or 6P5 or 2C22 tubes as well. I guess a maximum draw of 20mA.

For the power supply, the ingrediants I have on the bench are:

C-core Mains trannie with 240-0-240, 290-0-290, 335-0-235, 375-0-375 secondaries.
Chokes as follows:
One 9H, 275 R rated 100mA (very solid construction)
One 7H, 98 R rated 150mA
Four 10H 150 R rated 100mA
Four 10H 90 R rated 200mA

Ideally I'd keep the 150mA and 200mA for power amps, but I don't have to if there's a real benefit.

Caps - various polypropylenes:
Several 10uf at 600vDC
Two 80uF at 600v
Six 56uf at 400v
Eight 40uF at 340v

also Russian Teflon caps .1 or .06 at 600v

Now - I've never built a choke input PSU before and this looks like a good place to do one, so I need a bleeder resistor after the choke across the first cap of size ?? 10K to 20K ??. Do I put the 80uF cap after the first choke? I can split the supply into dual mono from there on. Do I put Teflon .1 caps in a Y shape before and after the first choke? Is this audible or overkill? I have plenty of teflons, so it's no sacrifice.

I'd welcome your ideas on the best PSU to cook up out of these ingredients. I want it to be really good so I don't need to change it for a good while. Andy

[andyjevans]

I should add that I'm planning on using damper diodes - I have 6DT4s and 6BY5G. The single 6BY5G would save space, but I'm happy to use the 6DT4s. Andy

[Eli Duttman]

It seems that the 335 V. taps get you about 300 V. of B+, with choke I/P filtration. Of course, I'm ignoring losses.

Go with the 6BY5. Each plate can handle 175 mA. of DC. That's PLENTY for a preamp.

How about using the 9 H./100 mA. and the 4 10 H./100 mA. chokes? Use the "beefy" 9 H. part after the rectifier. The 1st filter cap. does not have to be big; so, use a pair of the 10 muF. parts. A 9 KOhm bleeder resistor across the the pair is indicated. Follow the initial LC section with paired LCLC filters (1/channel). 10 muF. is more than adequate in the intermediate filter cap. positions. Use the 80 muF. parts for the final filter caps. You can go to the 375 V. taps of the power trafo safely, since all of the filter caps. are 600 WVDC.

The scheme I've outlined will be very heavy. However, ripple will be very low and regulation quite good.

[andyjevans]

The scheme I've outlined will be very heavy. However, ripple will be very low and regulation quite good.>>

Yes - it's pretty much what I was thinking of doing. How light can I go on the caps? You're suggesting L>20uF>L>10uF>L>80uF. The two 80uFs might be hard to fit - that's why I was thinking of a single one after the first choke. I'll have to look at all this in PSUD. I can't find any way of putting a bleeder resistor into PSUD - is this correct?
Thanks for the helpful response. Any further ideas very welcome, ANdy

[Eli Duttman]

If the 80 muF. parts are simply too large, space wise, use paralleled 10 muF. parts. I like lots of energy storage in the final position, but you should be fine with a value below 80 muF. Quite satisfactory, if unspectacular, PSUs have been made using a CLC filter constructed from 2X 8 muF. caps. and a 10 H. choke. The filter we are "kicking around" is BETTER.

[Miles Prower]

Hey Andy: Rectifier Applications Handbook

[andyjevans]

That's a pretty good web page - thanks! I'm off to read Morgan Jones on the subject - always enlightening. Andy

[SY]

Andy, for a circuit with good PSR (which you have), class A, and a 20mA draw, you really don't need anything exotic. Regulation isn't really an issue- this isn't an AB output stage. Low noise and ripple are quite easy. And if you do your layout and grounding properly, all the exotic diodes, snubbers, chokes, and suchlike won't make a blessed bit of difference in the actual noise and ripple on the PS line. Using very, very simple and conventional circuitry, I managed to get noise and distortion of my preamp below -100dB ref 2V.

Morgan Jones has some interesting waveforms in his power supply chapter, but he begged the question a bit by not showing the effects of his snubbers and high speed diodes on the actual rail voltages. I'm guessing there weren't any. And that's not a random guess; I did a bunch of rectifier measurements, substituting different types into a standard circuit, and didn't see any of the claimed differences where the rubber meets the road: the rail voltage.

[andyjevans]

Using very, very simple and conventional circuitry, I managed to get noise and distortion of my preamp below -100dB ref 2V.>

Hello SY - good to have your thoughts. I wonder could you tell me what in practice you'd call a very simple circuit? I have a 6BY5G to rectify, and various chokes and polypropylene caps - can you suggest an 'adequate' PSU schematic? Thanks, Andy

[SY]

Take a look at what I did in the Heretical's supply. The exotica factor is near-zero, which some might complain about, but the bottom line is a quiet, low-distortion preamp. I would particularly commend your attention to the RC filtered version, where the filters take up a few cubic inches and cost me about $10. I think it's worthwhile to go through the exercise of specifying a preamp S/N, then back-calculating to see what the power supply must do.

Using a CC diff amp was a good idea- common mode noise is efficiently rejected. And PSU noise is common mode.