PY 81 as power rectifiers?

[Amie]

I have an idea. I will lose the 15volts and connect the centre tap to chassis, so the 15 volt section just runs the heater?

A

Edit.. Sorry Jeffrey, I estimate about 50mA

[Geordie McBoyne]

Only one disadvantage with the PY81 is the longer warm up time. Although in some applications that could be useful as a delay device.

Geordie McBoyne.

[ppppenguin]

Keeping the heater supply separate from the HT is a good idea. At 50mA HT there should be no problems with halfwave. Do you know what sort of resevoir cap you'll use? Or alternatively how much ripple you can tolerate. To a fair approximation CV=It. C is the cap, V is the ripple voltage, I is the DC current and t is the cycle period. 20ms for fullwave, 40ms for halfwave.

For 50uF and 50mA that's about 40V of ripple unless I've made a magnitude error. For most purposes you'll need a smoother cap with resistor or choke to get that down to something respectable.

[Terry]

I have an idea. I will lose the 15volts and connect the centre tap to chassis, so the 15 volt section just runs the heater?

Great minds think alike! *

I was just going to post that but, going through the other previous posts, I see that you beat me to it!

* Before someone else says it, there is another, contrary adage: fools seldom differ!

[Mike Watterson]

The transformers have a 230volt primary, 230 volt 330mA secondary tapped at 15 volts, the tapped part of the winding can't be separated..there is also a separate centre tapped 12volt 2A winding.

The tapped 15volt section gives just over 17VAC when running the heater of a PY81.

I've seen old designs that did indeed have the rectifier between the transformer and chassis. Some valve and one was a test instrument with a metal rectifier.

It's also handy if you want a -ve voltage.

[ppppenguin]

Before there were TV boost diodes like the PY81 energy recovery in line output stages was a bit tricky. The theory was easy enough - Blumlien developed the idea in the 1930s. The problem was that most topologies had high pulse voltages on the cathode, too much for the valves of the period. So there were clever topologies a bit like Amie's proposal, to keep the cathode near chassis potential.

[Amie]

Keeping the heater supply separate from the HT is a good idea. At 50mA HT there should be no problems with halfwave. Do you know what sort of resevoir cap you'll use? Or alternatively how much ripple you can tolerate. To a fair approximation CV=It. C is the cap, V is the ripple voltage, I is the DC current and t is the cycle period. 20ms for fullwave, 40ms for halfwave.

For 50uF and 50mA that's about 40V of ripple unless I've made a magnitude error. For most purposes you'll need a smoother cap with resistor or choke to get that down to something respectable.

Bear with me here,

One idea (not mine), I have some 6w mains transformers that when loaded with a 16 ohm speaker will give a not to bad a match for the pentode section of a PCL 83 or similar, across half the primary.

So, what I intended trying is to feed the rectified HT via a 47uF (reservoir) cap to the center of the two primary windings in series, one primary to go to the anode of the output pentode, the other to power the rest of the set via another 47uF (smoothing) capacitor.

Why? Well if you look at the output stage of the Grundig 80U that I repaired, it uses a similar system. Now I thought it was a way to provide the rest of the set with nice flat DC without the expensive choke. But, the transformer in that set is very small, and the core is interleaved, not gapped, (I Think that it's original), why Grundig did it that way I think is to balance the standing SE current in the output valve with the current in the rest of the set, (I measured it, they're similar), by that means they were able to use a smaller component for the transformer, AND, provide "lossless" smoothing.

If I'm right, then that's a pretty cool piece of design work.

I can't fathom how well that idea will work with the random parts I have to hand, but it may be fun to cut and try.

If anyone has any insights, then please chime in.

Thanks ,
Amie.

[ppppenguin]

I'm no expert on the subject but I think you're right. I've certainly heard about hum cancellation and DC balancing techniques. I have no idea how critical the balance needs to be for either purpose.

[Amie]

I have an idea. I will lose the 15volts and connect the centre tap to chassis, so the 15 volt section just runs the heater?

Great minds think alike! *

I was just going to post that but, going through the other previous posts, I see that you beat me to it!

* Before someone else says it, there is another, contrary adage: fools seldom differ!

Our maths teacher is bald and claims that grass never grows on a busy road......one of the lads in the class pointed out that it rarely grows on wood either....

[Amie]

I'm no expert on the subject but I think you're right. I've certainly heard about hum cancellation and DC balancing techniques. I have no idea how critical the balance needs to be for either purpose.

I think that it could only improve it, not make it worse.

It follows that it could allow the ungapped transformer to not saturate its core if there is a ballance current too, crudely aproximating push pull...

I will experiment tomorrow...

Amie