Twin Regulated Filament Power Supply

[mooreamps]

...

[skyboltone]

I've actually thought about that sheme. Good going. I suppose it all depends on making sure that the two coils are phased right other wise they'll buck. So what's the raw AC voltage of the two coils tied together?

Come to re-read it, maybe we're on a different page afterall. I'm thinkin series the two coils for 11.3 volts, then drive two seperate regulated loads off the resulatant bridge. No center tap. So you've paralleled the coils? What center tap? There ain't no center. The 5V coil has to buck the 6.3V. Has to.

Drawing please.
Dan H

[mooreamps]

...

[skyboltone]

Very, very kool. This will make one of my recycled transformers an EL-34 supply instead of a 6V6 or EL-84.

Thanks Mister
Dan

[David Root]

+1 on that! That is as good a definition of the Ah-Hah! moment as I've seen in a good while.

Somehow it rings a bell though, I will take a look in my books, I think I remember something similar in result but not the same in design.

[novosibir]

Only to save you some unnecessary efforts. You don't need neither regulated, nor unregulated heater DC in a push-pull power stage! Only insure, that you feed all pin 2's from the same heater wire and all pin 7's from the other.

Any possibly induced hum from the heater supply through the cathodes will be canceled out each other due to the push-pull operation of the power stage - the OT is doing the same with the remaining ripple after the 1-st filter cap - that's alike a humbucker pickup is operating.

A regulated DC heater supply for the preamp tubes tough is a good thing. In my amps I'm using 12.6V regulated DC for the heaters, fed into all pin 4's, while all pin 5's are grounded and all pin 9's unused. The result is absolutely hum free amps, even dimed not a tad of a hair of hum

On the attached schem's page you'll see some different ways to wire the heater supply. On the upper right schem it's an 'elevated center tap', what already makes an amp very very quite humwise, since the basically voltage potential of the heater supply is elevated much above the usual cathode voltages in the preamp tubes. Nearly no cross-talk from the filament to the cathodes!

On the lower right schem you see by principle, how I'm wiring the heater supply in my amps, although I'm using an adjustable regulator, which additional requires a resistor/trimmer as seen in another post some steps above in this thread. And of course '5' still has to be grounded.

On this photo you may see (and borrow the idea) on the right end of the board, how you might do it. Very close the big blue filter (10,000µF), the additional (black) after the regulator (1,000µF), and the regulator itself (LM317K) mounted to the chassis:

http://www.larry-amplification.de/briti ... pur096.jpg

Larry

[David Root]

I guess I was wrong, Mr. Moore, I didn't find your double bridge setup where I thought it was.

I've done DC both ways (hmm, perhaps I should rephrase that...), all tubes and just the power tubes. Generally speaking novosibir/larry is right about that.

The one case where I did use it on all tubes was a high gain circuit using 6SL7s instead of 12A*7s, not a good idea as the heater is not humbuckable like the 12A*7 and they hum like crazy on AC in a hi-gain situation. So I wired everything 6.3VDC as the PT heater tap could handle four 6V6s, a 6K6 and five 6SL7s, without a regulator, and it worked fine. It was mainly for convenience that I did that, though.

I too normally now use regulated 12.6VDC on 12A*7 hi-gain circuits, with the full bridge 4-diode rectifier. 6.3VDC works OK on low gain 12A*7 circuits and octal preamps.

One new point--some hi-fi types believe that AC heaters are better than DC, because AC supply gives an effectively flat potential plane to electrons leaving the cathode ie cathode current, so all electrons have the same exit velocity; whereas DC supply gives a effective potential gradient to those electrons, ie a gradient of different exit velocities.

They claim to be able to hear the difference. Of course we all know hi-fi types have ears superior to those of musicians--NOT!!

[mooreamps]

...

[skyboltone]

G-
Ok, I've ordered up some parts today. I'm choosing to use a pair of 8A 40V bridge rectifiers rather than horse around with individual diodes. They do have a max voltage drop of 1 volt. Also, I see you've used heat sinks. Have you calculated Pd? How hot do they get? In my application I would like to keep the footprint a little smaller than the board you've shown. In fact, it makes sense to remote the regulators to top of chassis with mica insulators and heat goop.

Dan

[nickt]

Apologies for the dumb question - what is the driver behind fully regulating pre and power amp heaters?

If there's another thread where this started I'd appreciate a pointer. Thanks!

[mooreamps]

...

[mooreamps]

...

[nickt]

The driver is to keep the filament voltage up when the line voltage sags. In small clubs like the QuarterNote, I measured the voltage on stage during one of the jam sessions, it was 103 volts. Even when I am working in my shop, if I turn on the floor heater, I can see the non-regulated filament power drop to 5.0 volts. The pre-amp almost dies when the cathodes are running that cold.

-g

Ah... wow... don't usually have that issue in Oz as we're 240V. Thanks for the info - now I see where you're coming from.

[Aharon]

G-
Ok, I've ordered up some parts today. I'm choosing to use a pair of 8A 40V bridge rectifiers rather than horse around with individual diodes. They do have a max voltage drop of 1 volt. Also, I see you've used heat sinks. Have you calculated Pd? How hot do they get? In my application I would like to keep the footprint a little smaller than the board you've shown. In fact, it makes sense to remote the regulators to top of chassis with mica insulators and heat goop.

Dan

I once mounted them on top the chassis. You have to remember the case of the regulator is hot, so you have to insulate it from the chassis....
Yes I have calculated power disipation. The part I selected is rated for 5 amps continous, and max load is 3.2 amperes. The part does get hot to the touch, but doesn't hurt anything. I did mount it on a heatsink because I will be shipping this one overseas.

Also, this will not work with just any power transformer. The 5.0 and 6.3 volt filament windings must come with a center tap, or all bets are off if your thinking about building one.



-g

What do you do with the center taps?.
Thanks
Aharon

[skyboltone]

G-
Ok, I've ordered up some parts today. I'm choosing to use a pair of 8A 40V bridge rectifiers rather than horse around with individual diodes. They do have a max voltage drop of 1 volt. Also, I see you've used heat sinks. Have you calculated Pd? How hot do they get? In my application I would like to keep the footprint a little smaller than the board you've shown. In fact, it makes sense to remote the regulators to top of chassis with mica insulators and heat goop.

Dan

I once mounted them on top the chassis. You have to remember the case of the regulator is hot, so you have to insulate it from the chassis....
Yes I have calculated power disipation. The part I selected is rated for 5 amps continous, and max load is 3.2 amperes. The part does get hot to the touch, but doesn't hurt anything. I did mount it on a heatsink because I will be shipping this one overseas.

Also, this will not work with just any power transformer. The 5.0 and 6.3 volt filament windings must come with a center tap, or all bets are off if your thinking about building one.-g

What do you do with the center taps?.
Thanks
Aharon

I'm with Aharon. They don't show on the drawing. The more I look at this thing the more I like Larry's solution at the lower right side of his drawing. With 11.3V input to the rectifier instead of 12.6. We're just looking to get at the P of the 5 volt winding. They way I see your scheme G- is that you're loading the 1100uf series capacitor at two places to get sufficient threshold voltage to overcome the losses of the rectifier and regulator and get the 6.3 at the output of the preamp tubes regulator. Cute. And no reason it won't work. I wonder though if you reversed the 5V and 6.3V windings so that you put the 5 volt winding into the center of the capacitor instead of the 6.3 if it makes any difference? I dunno. I know from experience that you can't get 6.3V out of a 5V winding alone. Too many losses in the rectifier and regulator. If you use a doubler you can get there but it seems like an aweful lot of trouble to avoid buying a filament transformer. Still, a nice starting point for my own experiments.