Double floating Circlotron

[jane]

Double floating Circlotron.

The Circlotron and the cathode follower output stage have some desirable properties, for instance low output impedance. The drawback is the driving requirements. High clean voltage swing can be a challenge especially without feedback.

Traditional cathode follower output stage:


As we can see, the voltage swing over the primary of the OPT is ep=(e1+e2)*A, where A is the gain in the tubes, normally 0.8-0.9. If we want a voltage swing of 200V on the primary side of the transformer the input signals need to be 200V/0.85=235V.

But what if we let the input signals swing with respect to each tubes cathode and not to the ground?



In this case the gain is mu*Zprim/(Zprim+rp) witch ends up with a reasonable requirement for the driver.
Example: Output tubes 6550, mu=8, rp=1300, Zprim 2.5k:
Gain = mu*Zprim/(Zprim+rp) = ~5.2x.
If the target still is 200V primary swing the input (e) now just have to be 200/5.2 or ~38V. About 200V less than the traditional cathode follower.


Next the Cirlotron:


The input signal is here with respect to the center of the output transformer. This reduces the requirements for the driver but still the input signal need to be at least the half of transformers primary voltage.

Double floating Circlotron:


About the same requirements as for the floating cathode follower, but in this particularly case the output tubes are wired as pentodes.

So to the question: Is there something wrong with my argumentation or will this be a standard Cirlotron with very nice input requirements?

Jan E Veiset

[Wavebourn]

As soon as you apply input voltage strictly between grid and cathode it is not a cathode follower anymore. It is a "common cathode" output stage, where plate (anode) load and a power supply are swapped; anyway they are in series.

On the last picture output tubes are not wired as pentodes. And the difference is, each tube sees a full winding of a primary VS half like in the first example. Too complex PS for better potential symmetry (less strict requirements to primary winding).

[Yvesm]

Originally posted by Wavebourn
As soon as you apply input voltage strictly between grid and cathode it is not a cathode follower anymore. It is a "common cathode" output stage, where plate (anode) load and a power supply are swapped; anyway they are in series.

Yeap ! No free lunch

Yves.

[Wavebourn]

Originally posted by Yvesm


Yeap ! No free lunch

Free cheese. In a mousetrap.


2 seperated power supplies, well isolated in wide frequency band from everything including each other... How much?

[smoking-amp]

How about this single supply circlotron:
http://www.diyaudio.com/forums/showt...796#post228796

Don

[Wavebourn]

Originally posted by smoking-amp
How about this single supply circlotron:
http://www.diyaudio.com/forums/showt...796#post228796

It is called cathode NFB

[jane]

Originally posted by Wavebourn
As soon as you apply input voltage strictly between grid and cathode it is not a cathode follower anymore. It is a "common cathode" output stage, where plate (anode) load and a power supply are swapped; anyway they are in series.

I should have seen that when I worked out the gain for the "floating CF" the same way as for a plate follower. I suppose I just wanted it so bad to be a CF that it blurred my mind.

Jan E Veiset

[Dave Cigna]

Originally posted by smoking-amp
How about this single supply circlotron:
http://www.diyaudio.com/forums/showt...796#post228796

Don

I think the McIntosh Unity Coupled stage is a more elegant solution. I like to call it the Iron Circlotron. Instead of coupling each plate to the opposite cathode through big honking capacitors we use the output transformer. I've left the secondary winding out of the drawing for clarity.

Conceivably, a dual winding CT choke could be used (as drawn, without secondary winding) and the output take across the cathodes (or plates, same thing) for OTL operation.

-- Dave