Cathode Follower Push Pull

[abbey road d enfer]

I get your point, balance is a non- issue with CF.

I see you have a switch for SE operation. In order for it to work, you need to disconnect the bottom tube. If you don't it will be in parallels with the top one and drop level by 6dB and possibly add distortion..

 

[abbey road d enfer]

It does, but I doubt you need better balance here (your PI is perfectly balanced already). The CF output stage is very limited in power, and that resistor is stealing some power, but I haven't done the sums so maybe you're not losing anything important.

According to simulation, this resistor steal about 90V from the B+, which leaves about 120V of clean cathode swing, for an output level of 18+ Vrms into 600r, or +27dBu.
With the resistor replaced with a link, the max clean output is a whopping +32 dBu into 600r (about 1.5W).
Of course these simulation results should be confirmed by experiment.

 

[solkatten]

I see you have a switch for SE operation. In order for it to work, you need to disconnect the bottom tube. If you don't it will be in parallels with the top one and drop level by 6dB and possibly add distortion..

Yes its dropping 6db, its there as a colour option, as is pentode in the first stage who is also adding distorsion/colour to the pre.
Yes the tail resistor was a bad choice in my case. Thanks for the feedback.

 

[merlin]

I see you have a switch for SE operation. In order for it to work, you need to disconnect the bottom tube. If you don't it will be in parallels with the top one and drop level by 6dB and possibly add distortion..

The output transformer is not gapped, so it will saturate if you disconnect the second tube.

 

[Walter66]

Maybe I will try the PP CF in my 4in1 pre it could be arranged like this:

The input 100K resistor is of no value in practise. Did the same on my tube amp with a 250K pot and it had nearly no effect. Signal volume is regulated with the pot, no need for additional input resistor.

 

[merlin]

Yes its dropping 6db, its there as a colour option, as is pentode in the first stage who is also adding distorsion/colour to the pre.
Yes the tail resistor was a bad choice in my case. Thanks for the feedback.

You'll get way more gain in SE mode since the PI will be converted into a gain stage. Thought I'd mention it, but maybe you're already prepared for that.

 

[solkatten]

The 100K to the grid of the concertina is to prevent grid current during clipping condition...
In my present 4in1 pre the level drops by 6db when a ground one of the grids in the PP-stage and force it in SE-mode, but the concertina still sees a balanced load of 2x 235K.

A better way to implementing the se/pp control and make it swepable.

 

[abbey road d enfer]

The 100K to the grid of the concertina is to prevent grid current during clipping condition...

What for?

In my present 4in1 pre the level drops by 6db when a ground one of the grids in the PP-stage and force it in SE-mode, but the concertina still sees a balanced load of 2x 235K.

In this last schemo, it is not the case. When you switch to SE, the cathode is grounded via a 220n capacitor, so the gain increases at mid and high frequencies.

This is the smartest way of implementing a se/pp control

Please allow me to disagree.

 

[solkatten]

What for?

In this last schemo, it is not the case. When you switch to SE, the cathode is grounded via a 220n capacitor, so the gain increases at mid and high frequencies.

Please allow me to disagree.

The cathode grounded via a 220n cap but it also ses 220k resistor to ground thats 3,3 Hz. But that was true for the earlier version, with cap direct to ground.
Picture in post 27

 

[solkatten]

according to valve wizard concertinas can have freq. dubbling effect when they is clipping hard a big value grid stopper prevents this
The Valve Wizard

 

[abbey road d enfer]

The cathode grounded via a 220n cap but it also ses 220k resistor to ground thats 3,3 Hz
Picture in post 27

OK, didn't see that.

 

[abbey road d enfer]

according to valve wizard concertinas can have freq. dubbling effect when they is clipping hard a big value grid stoppet prevents this
The Valve Wizard

OK, I see. This is a concern only when the PI is required to deliver high voltage swing. And this is probably the case here since the output stage has no gain.

 

[solkatten]

concertinas has a bad rep. in guitar amps due to this nasty clipping behavour

 

[abbey road d enfer]

concertinas has a bad rep. in guitar amps due to this nasty clipping behavour

True when the output tubes are 6V6/6L6/EL34, but nice with EL84/6BQ5.

 

[Mr. Moscode]

And as Abbey pointed, if you really want big output it could better to connect a high ratio transformer to the anode and rely on the tubes great voltage swinging capabilitys, than rather poor current capabilitys.

Assuming the circuit is driven with tubes then you still have the high voltage swing capability. I like low impedance driving the xfmer, although I have never built this one.

FWIW McCintosh used a scheme where both plates and cathodes went through 2 output transformer windings for a power amp.

 

[abbey road d enfer]

Assuming the circuit is driven with tubes then you still have the high voltage swing capability.

That puts the burden on the stages that drives the output tubes. In a standard Williamson arrangement, the PI is not capable of driving the output tubes grids full swing.
If you want the full swing of about +/-400V for a PP of 6L6's, it takes a driver stage capable of such a swing, so basically powered from 800V B+.

I like low impedance driving the xfmer, although I have never built this one.

It certainly has its benefits, but also its constraints.

FWIW McCintosh used a scheme where both plates and cathodes went through 2 output transformer windings for a power amp.

That's correct. However they used an equal load on plate and cathode, which halved the driver stage's swing requirement.
Still, the driver stage is using floating rails as to not exceed the driver tube's plate-to-cathode voltage.

 

[solkatten]

Assuming the circuit is driven with tubes then you still have the high voltage swing capability. I like low impedance driving the xfmer, although I have never built this one.

FWIW McCintosh used a scheme where both plates and cathodes went through 2 output transformer windings for a power amp.

I read about the uni-coupled power amp
Intresting... its a thing you could try if you have 2 simular PP-trafos. One goes to cathode with grounded CT (or B-), the other cross-coupled to the anode, then join the secondarys and voila!!! I might try that I got 2 UTM-3580 trafos....

 

[abbey road d enfer]

I might try that I got 2 UTM-3580 trafos....

Just remember to seriously increase the drive voltage...

 

[Mr. Moscode]

That puts the burden on the stages that drives the output tubes. In a standard Williamson arrangement, the PI is not capable of driving the output tubes grids full swing.
If you want the full swing of about +/-400V for a PP of 6L6's, it takes a driver stage capable of such a swing, so basically powered from 800V B+.

The use of a follower is to add current to a signal.
Isn't this a 12AU7 a small signal tube? It will have heater cathode voltage limits.
I know this is an experiment but isn't the idea to to use it as a preamp output stage? Doesn't the xfmer step up the voltage?

 

[abbey road d enfer]

The use of a follower is to add current to a signal.

Yes.

Isn't this a 12AU7 a small signal tube? It will have heater cathode voltage limits.

Yes. In particular that's why McIntosh use medium power triodes as drivers, combined with floating supplies.

I know this is an experiment but isn't the idea to to use it as a preamp output stage?

Yes, still a preamp is supposed to have a maximum output level of about +20dBu (about 8V rms or 20V peak-to-peak). With a 4:1 xfmr, that implies an output swing of 80V, which in turn implies driving the grids with 80V swing. It is indeed feasible but somewhat drifts from the standard arrangements used in Williamson-type designs.

Doesn't the xfmer step up the voltage?

Actually the OP specified a 4:1 xfmr.
Most equipment is designed to be capable of driving one 600 ohms load. It is not entirely justified in a world where the input impedance of most equipment is 10 or even 20kohm.
The bridging principle requires the output Z being <10 times smaller than the input Z.
In order to drive a 600r load, small signal tubes need xfmrs.
Of course one can choose to stay away from the bridging concept, that's what many audiophile products do.
Good for people who have just recently discovered the virtues of balanced connections.