Power Supply Question

[RichardJones]

This is a strange request.

I am looking to build a 170vdc 2.5A power supply. It will be supplying 6AS7's in Class A.

Can I use an isolation transformer for the power supply?

I do have some 200vdc 7500uF caps... can I use these? Or is it better to parallel smaller better ESR caps than these?

Should I go with a CLC stage? or even a CLCLC ?

ripple reduction is a must.

Any ideas? I don't really want to use any voltage regulators....

Thanks!!!!

 

[SY]

Class A will help a lot. Hopefully, you'll be doing push-pull, which will greatly increase power supply rejection so that you don't have to do too much in the way of L (heavy, expensive).

In any case, an isolation transformer should work pretty well, assuming it's got a sufficient current rating so that the voltage doesn't get dragged down. If you've got a decent amount of capacitance (and it sounds like you do), and you're doing a capacitor input filter, you'll get your 170V with no problem.

 

[RichardJones]

seperate caps...

Actually its SE Class A.

Each 6AS7 section will draw about 100mA. I would rather buy smaller value caps and connect them at the 6AS7 pins.

They would have to be rated at 200vdc... but what capacitance?

Also..... should I then do a common intial capacitor to a common inductor then to individual caps?

hmmm.......

 

[SY]

Ahhh, a half-assed circuit!

Distributed capacitance has to be done very carefully to avoid oscillation. Pay particular attention to ground returns.

 

[RichardJones]

what would work better???

OK... then I am open to suggestions....

What would work better?

 

[dhaen]

Richard,

Are you aware of PSUD2?

http://www.duncanamps.com/psud2/index.html

Power supply simulator. Free, small, works!

Cheers,

 

[SY]

What would work better?

Push-pull.

If you're determined to make a 2nd harmonic generating heat machine, use a centralized PS with separate SHORT leads to each tube socket and a star ground. Use antiparasitic resistors in grid and plate leads. 6AS7/6080 outputs will take off if you don't use good layout practice.

 

[dhaen]

If you're determined to make a 2nd harmonic generating heat machine,

Yes, why not cancel out all the pleasant harmonics so you can concentrate on the unpleasant ones

No wonder Mrs Y doesn't want an orchestra!

Seriously, PP is a lot less bother .. I agree

Cheers,

 

[EC8010]

If you are not prepared to use a regulator, you will find that you need CLC at least to get the ripple down to an acceptable level. Back of an envelope calculations say that your ripple needs to be <14mVpk-pk to be 90dB down.

I'm afraid that you really pay for SE in the power supply.

 

[RichardJones]

SE it is!!.... maybe

Unless I throw in a SS regulator.... and have small local bypasses on each tube?

If I do PP.... I would need 26 6080's per channel.

Do you think the constant power draw and ripple cancellation is worth this effort????

I would cut out a large coupling cap (4700mf / 200vdc).

One channel would draw 170vdc 5A. ugghh!!!!

And no... I CANNOT change the amount of current draw or voltage....

 

[EC8010]

Local bypasses are for dealing with RF, not hum.

How much power is this behemoth required to deliver?

 

[SY]

John, Mrs. Y thinks music is a lovely thing to have on in the background, but quietly enough so as not to interrupt the television. And for gossakes, without those horrible woofers that shake her recliner whenever I play anything with a bass drum.

It's amazing that we've been able to stay married all these years.

 

[RichardJones]

power???

umm... i estimate around 25W per channel into 8 ohms... about 5% efficient in SE.

 

[dhaen]

Richard,

Is this beast OTL? Why such low efficiency?



SY,

I sympathise.
I am presently negotiating for the permanent removal of a large vase that "appeared" on one speaker. "But where else can it go?"

Cheers,

 

[RichardJones]

Yes

It is an OTL... but not a Circlotron design.

 

[fdegrove]

DUH?

Hi,

If I do PP.... I would need 26 6080's per channel.

Lessee...

For an OTL design going Class A is not require since crossover distortion is pretty much absent.

So 26 6080s run in Class AB would get you 300W into a 8 Ohm load and about 550W into a 16 Ohm load.

The heater of the 6080s alone would consume 2.5 A * 26 = 47 A per channel.

With a PP design all you'll need for the bipolar supply are those 7500 uF/200V caps and a high amperage diode bridge.

Regulating such a supply requires a reg the size of an amplifier, in other words you'll need about one 6080 to regulate another.

For, say a 100W OTL amp you'll need about 10 6080s run at +/- 150 V.

You could regulate the B+ of the input stages, which I'd recommend.

Cheers,

 

[dhaen]

Frank,

He only has 170V DC to play with....
Unless he voltage doubles..

Cheers,

 

[RichardJones]

heres the deal

I want to do SE.

I need a 170vdc power supply capable of 2.5A cont. current.

low ripple. (can 20db of feedback from output to input tube reduce this PS ripple?).

I'm thinking of a pentode input stage.. with an unbypassed cathode resistor.....

I just need to know if I should go the paralleled caps at each tube... or just a few big caps.....

and whether SS regulation is a must?

Thanks!!!

 

[fdegrove]

PSU QQ.

Hi,

I don't really see how you're going to be able to get a clean supply starting with an isolation PT.

Assuming the secondary is 120VAC, the highest possible voltage is going to be 168 VAC and even that is an optimistic calculation.

For an SE amp a clean, ripple free supply is a must.
A voltage doubler is no solution either since you'd only have half the current rating of the PT available.

Also, the 7500 uF/200 may turn out to be too much of a good thing as well.

You're facing some dilemma here, I'm afraid.

Cheers,

 

[SY]

With a bipolar pass element, he'll only lose 5 volts or so, assuming the raw supply is stiff. And the feedback will reduce the effects of ripple by a factor of 10.