Circlotron questions

[SpreadSpectrum]

For a while I just thought the circlotron looked weird and overly complicated, but recently I have opened my mind and started taking a look at it. I just have a couple of points I wanted clarified.

1. I see a lot of people referring to this circuit as a 100% feedback arrangement (a cathode follower), but it seems that tying the plate supply to the cathode of the opposite tube would result in only 50% of the total plate swing being seen in the grid circuit, so it seems that drive requirements would be half of that for a cathode follower output stage or similar to a McIntosh style amp?

2. I just need an output transformer with half the turns ratio as normal to build one of these, since the entire primary is still driven when one tube cuts off?

[Zero Cool]

Or you could build OTL! Atma style!

[Yvesm]

1. Yeap !
2. Yeap !

The main drawback is having the two PSU "swinging" at signal rate . . . and in opposite directions !
You have to design them with lo capacitance between them and to ground, this includes the transformer windings.
If not, it's like adding few nanofarads between plates.
Of course, an OTL version driving a few ohms load doesn't care about even a full microfarad shunt cap

In short, Circlotron moves some problems from the OPT to the PSUT but in opposite direction: you MUST have an hi linkage inductance and lo shunt capacitance between HV secondaries and others.
Using two PSUT puts this inductances and capacitance in serie, better than nothing !

Yves.

[banat]

Originally Posted by Yvesm

1. Yeap !
2. Yeap !

The main drawback is having the two PSU "swinging" at signal rate . . . and in opposite directions !
You have to design them with lo capacitance between them and to ground, this includes the transformer windings.
If not, it's like adding few nanofarads between plates.
Of course, an OTL version driving a few ohms load doesn't care about even a full microfarad shunt cap

In short, Circlotron moves some problems from the OPT to the PSUT but in opposite direction: you MUST have an hi linkage inductance and lo shunt capacitance between HV secondaries and others.
Using two PSUT puts this inductances and capacitance in serie, better than nothing !

Yves.

Hi Yves.Must say this maintransformer two separated secondary capacitance have some `interference` only in old style Circlotron Tube Amp,with Tube rectifier PSU(EV for example),in this case each secondary have center tap conected to two,source`(oposite Circlotron cathodes),betwen two cathodes is Load(OPT),here some AC(sonic) power exist in main transformer secondary.
But any way even here major AC(sonic)power travel thru Electrolytic Capacitor(Smoothing Filter)not thru main transformer(Here is Your missunderstading).With modern SS bridge rectifier we have two separated main transformer secondary(for two phase output balanced bridge) without center tap and when we use modern SS bridge rectifier only `inverse` capacitance is junction capacitance of each SS diode in rectifier bridge(40-80pf) in series with main transformer both secondary capacitance(resulting in minor comon mode capacitance).Remember AC(sonic)power travel thru the Output Power Tubes,Smoothing capacitor and Load.DC component only suply output Balanced bridge,To increase AC(sonic)power travel thru the each smothing Capacitor(and linearise fiter capacitor transfer characteristic and whole Amp) I always install SS clamp diode(UF5408) parallel with each two Smothing filter(electrolytic Capacitor),important thing is to keep SS bridge rectifier snubber blocks low as posible and conect both secondary in same manner to both SS rectifier bridge( this second is to avoid noise&hum).All the Best

[Yvesm]

Hi banat,

We don't speak about the same problem !
"Nanofarads" that usually appears between secondary windings of the same transformer are not hidden by any rectifiers unless you put a choke in serie with BOTH wires.
Of course, individual transfos improve the situation.

But I agree with the utility/necessity of some snubbers blocks.

Yves.

[SpreadSpectrum]

Thanks, Ives. I was pretty much sure I was right but had read in several places that it was 100% feedback, but didn't see how that could be true.

Zero Cool, I know it is not fashionable to say here, but if I were to build an OTL circlotron I would use transistors! I live in a warm place and don't need a space heater.

[Nelson Bass]

Originally Posted by SpreadSpectrum

Thanks, Ives. I was pretty much sure I was right but had read in several places that it was 100% feedback, but didn't see how that could be true.

Zero Cool, I know it is not fashionable to say here, but if I were to build an OTL circlotron I would use transistors! I live in a warm place and don't need a space heater.

The Amazing Fet Circlotron by Mike Rothacher

[banat]

Tubes Circlotron(OTL)classA/AB sounds FAR Better than SS Circlotron (OTL)classA/AB.Heat delivery is Allmost the same for both and depends from bias class only(admit SS have no filament).Any way Circlotron topology is right choice(SS or Tube)for excelent sounding power Amp(I prefer Tubes).Good Luck

[SpreadSpectrum]

Surely a tube OTL circlotron would require banks of power tubes to get a low zout. Are you saying it is possible with a pair of output tubes? That must require an unholy amount of feedback?

I think I would prefer to just use a pair of KT88s with a transformer.

[banat]

Here is old Electro Voice A50 Circlotron amplifier with 2x 6550,OPT primary 1,2Kohm,fixed biased near to class B(for long output tubes live).Specs are quit good for early 50`but for 2010 too.Of course PSU cood be much simplier with modern SS device(dirt cheap).Like the way of variable voltage/curent feedback in this Amps which allow to find optimum damping factor for diverse Load(speaker).For further info you can visit great Net site:Circlotron History Page . For pure OTL Circlotron Atmasphere S30 are good example,only 5x6as7 and 3x6sn7 per chanel(2x30w/8 ohm,2x45w/16ohm,overall feedback only 2db).For further info found thread on this site:`What Tubes For Tube Amp`(here Mr Karsten(Atmasphere) published M60 schematic)