inexpensive PT's on Ebay Woodward Shumacher

[stingray_65]

 Found this at one of the HiFi forums

http://cgi.ebay.com/NOS-POWER-TRANSFORMER-120-Watt-Tube-Amp-7019-/320586931410?pt=Vintage_Electronics_R2&hash=item4aa4785cd2#ht_500wt_716

he're their discussion, a couple have bought and tested these out

http://www.diytube.com/phpBB2/viewtopic.php?t=4398

$71 will get you 3 of these

Ray

[mcrracer]

Can anyone give an example with values of a good doubler circuit that would work with these trannies?
What B+ range could be expected? Thank

[mresistor]

Evidently Schumacher is still in business  ShumacherElectric.com

[eleventeen]

mc, here's a link to a standard voltage doubler that I have used and I know works. Two caps in series, two diodes in series. One side of the HV winding to the junction of the caps, one side of the HV winding to the junction of the diodes.

http://www.coolcircuit.com/circuit/voltage/index.html

In my experience the B+ output is just "a tad more", about 10% more than double the AC input as read on an ordinary voltmeter, not an RMS meter. I hesitate to give the calculation, but the use of a doubler circuit allows caps that have lower voltage ratings than they'd have to have if they spanned rail-to-rail. Hence, there is a tendency to use slightly higher value caps, and the output voltage is responsive to this. With 50 ufds, that's where you'll see input +10% or so. If you use 20 ufds, lower.

If each half is 160 VAC, using 50 ufds, you'll see 350-360vdc on the output.

[FYL]

mc, here's a link to a standard voltage doubler that I have used and I know works.

Standard half-wave doubler. Works OK, but remember that max DC current will be halved re. to a standard half-wave, ie. if the PT is sized for 300 mA, you'll only get 150 mA.

[mcrracer]

mc, here's a link to a standard voltage doubler that I have used and I know works. Two caps in series, two diodes in series. One side of the HV winding to the junction of the caps, one side of the HV winding to the junction of the diodes.

http://www.coolcircuit.com/circuit/voltage/index.html

In my experience the B+ output is just "a tad more", about 10% more than double the AC input as read on an ordinary voltmeter, not an RMS meter. I hesitate to give the calculation, but the use of a doubler circuit allows caps that have lower voltage ratings than they'd have to have if they spanned rail-to-rail. Hence, there is a tendency to use slightly higher value caps, and the output voltage is responsive to this. With 50 ufds, that's where you'll see input +10% or so. If you use 20 ufds, lower.

If each half is 160 VAC, using 50 ufds, you'll see 350-360vdc on the output.

Thanks!

[eleventeen]

Standard half-wave doubler. Works OK, but remember that max DC current will be halved re. to a standard half-wave, ie. if the PT is sized for 300 mA, you'll only get 150 mA.

Good, important point. My own point was that it (the linked ckt) actually works. By the way, for EACH of the diodes, I use two diodes in series, 4007's. And, one can use balancing resistors (150K - 220K) across the caps.

These particular transformers (the topic matter of this thread) have a completely separate bias winding, which is nice. Lacking a bias winding, it is possible to derive a bias voltage using a feed from one of the HV legs (as in Bogen amps) while using a voltage doubler, but I was not able to get such to work quite the way I wanted it.

[Geezer]

Would be great for using a FW Bridge recto & running 4x EL84's @ ~~315v B+

[FYL]

Would be great for using a FW Bridge recto & running 4x EL84's @ ~~315v B+

Or even medium bottles in gloriously inefficient class A.

A pair of 6L6Gs would give around 20 watts in fixed bias, slightly less with cathode bias. Nice brown sound, pre-tweed.

[PRR]

> about 10% more than double the AC input

The no-load DC should be 2*1.414 times ACV. "Double" what you'd get with a non-doubler such as half-wave or FWB.

If you are getting 2*1.1*ACV, you need bigger filter caps or a bigger transformer. The doubler does need frighteningly huge caps.

> as read on an ordinary voltmeter, not an RMS meter

If your DC isn't awfully riddled with ripple, its RMS is "the same" as its average.

I happened to have a simulation of an awful doubler handy. 100V peak to peak ripple (482V to 582V). The Mean is 532, the RMS is 533. For RMS, the ripple peaks count more than the ripple dips. For amplifiers, what counts is the ripple dips... any output needing more than 482V will be hashed-up with buzz. 532 average and 482 clean is not a good plan. We normally want MUCH less ripple than this.

For a transformer like above (235VAC and tested to 200mA in a Doubler), two 470uFd caps gives 1%pp ripple. If caps were expensive we could explore smaller values... but these only need to be 350V parts and 470u 400V caps are readily available in snap-cans.

> I use two diodes in series, 4007's.

One GREAT thing about this doubler is LOW diode voltage. 350V peak (1.414*VAC). A single 4007 each side is ample.

OTOH the current pulses are brutal. The 200mA output is awful close to a 1A diode's rating. Anything larger, use fatter diode.

> one can use balancing resistors (150K - 220K) across the caps.

The raw doubler is self-balancing. (if the wall-power wave is symmetrical) How can the caps not have the same voltage?

If you need another stage of filtering, those caps may need balance resistors. Actually if you don't load the half-voltage output much, there is an elegant way.... but not posting that today.

> running 4x EL84's

Or maybe ten? It has the B+ oomph to warm ten 12W plates.

[eleventeen]

I stand corrected, twice: I measured the actual voltage doubler I have in use.

102 volts AC input yielded 272 volts on the output using dual 40 ufd electrolytic caps. Very close to 1.4 * double the input volts, and..

given that I have a couple of what I erroneously called "balancing" resistors across the caps (150K each) PLUS a 330K bleeder resistor across the whole thing (I will no longer build power supplies for tube circuitry and their associated B+ levels without at least SOME form of bleeder resistor)

the output is very close to 1.4 * double the input volts.

[moonbird]

Thanks for the tip on this PTs - I picked some up!!

Sorry for the nubie question - but which voltage doubler (there are two) are you all referring to on the web page referenced above. The one at the top or the second one called a "cascaded" doubler?? What is the difference between these? When you you use one but not the other? thx.

[DummyLoad]

If you are getting 2*1.1*ACV, you need bigger filter caps or a bigger transformer. The doubler does need frighteningly huge caps.

one rule of thumb i use with a Vdblr. double the values. e.g. if you'd use 1000uF/A then with a doubler use 2000uF/A or even 2500uF/A for heavier loading. that goes for the VDC rating of the caps as well. i guess at some point in the 60's the caps got cheaper than the iron, of course, with the help of silicon. 

--ISO

[eleventeen]

moon, if you're asking me, the Vdoubler at the very top of the page: Two caps in series, two diodes in series. One side of the HV winding to the junction of the caps, one side of the HV winding to the junction of the diodes.
 
Output = 2 * 1.4 * AC volts on the secondary. Lock and load, when I have 102 VAC on the tranny secondary I get 272 VDC as the highest high DC output. This is with twin 40 ufd caps. I have a 330K bleeder resistor across the whole thing and one 150K across each electrolytic cap, which are superfluous. But those resistors *DO* knock down the output volts just a tad.

Why do I use that one?

1: Probably because I know it works and I can predict the output volts.

2: Because (in the first ckt) no cap "sees" the entire output voltage, this enables using lower voltage rating = cheaper caps.

3: I myself am not especially fond of configurations where the entire through-current of a highish current power supply has to go through a cap. Each and every electron supplied by doubler #2 (cascade) has to go through C1. Might be a superstition or a silly preference, but I just prefer not to do that.

[PRR]

> Why do I use that one?

Because you can.

The "other one" can be worked with a grounded source, and can be extended to triple, quad, sextupler. But you have a floating source, and 2X is plenty.

[eleventeen]

Don't triplers, quadruplers, etc; suffer from serious current-capacity loss?

[darryl]

Don't triplers, quadruplers, etc; suffer from serious current-capacity loss?

A doubler will supply ( approximately ) twice the voltage of a bridge rectifier/filter, but the available current must be halved. This ratio applies to all voltage multipliers, so a quadrupler, for example, provides 4 times the voltage, but only 1/4 the current.

There's no such thing as a free lunch... :sad7:

[eleventeen]

Understood. No free lunch expected; but my readings on triplers. quads, etc; (which I have never used nor experimented with) seem to suggest that ampacity declines beyond the rise in voltage output. By what fraction or percent, I don't know.