Direct coupled pentode to triode AC cathode follower

[ChrisM]

Working on a design that has a 5879 pentode in it. This tube has a plate resistance of ~2M. This translates to a very high output impedance compared to a standard 12AX7. As a result having a tone section after the pentode does not make a whole lot of sense.

After the 5879 pentode there is a effects loop consisting of a 12DW7 with the 12AU7 portion of the tube doing the CF and the 12AX7 portion of the tube doing the recovery. I am planning on putting the tone control right after the CF but before the effects loop.

So what I am not thinking of here? What is there to consider when direct coupling a pentode stage to a CF stage?

[tubeswell]

Small signal pentodes usually have a lowish plate voltage, so there is no problem with DC coupling to a CF stage.

I've DC coupled a 5879 to a 12AU7 before (in a circuit using Merlin's morph control but with the triode coupling mode buffered from the CF) - schematic attached (with voltages. Voltages would be the same for a conventional CF buffer stage in this regard)

[ChrisM]

Ah perfect, that is exactly what I was going to try.

What was before the 5879? How did this design work out for you?

One question about the morth control, why did you connect it back to the cathode and not the grid? Don't understand your comment about buffering the triode mode...

[Alan0354]

What is the effect of Morph control to the sound? It almost like a negative feedback but to the screen grid instead.

[pdf64]

CFs are generally used before fx loops to achieve a low impedance drive, thereby mitigating signal degradation.
However, a tone stack generally has a high impedance output, so you may wish to review such an arrangement as part of the design rationale.

[tubeswell]

Ah perfect, that is exactly what I was going to try.

What was before the 5879? How did this design work out for you?

One question about the morth control, why did you connect it back to the cathode and not the grid? Don't understand your comment about buffering the triode mode...

I've tried this as a V1 stage and as a second stage. (grid load resistor for the 5879 was omitted from the schematic sorry - you can use 1M)

[tubeswell]

What is the effect of Morph control to the sound? It almost like a negative feedback but to the screen grid instead.

its mostly a form of gain control

[teemuk]

Looks like its "morphing" between pentode and triode modes.

[ChrisM]

CFs are generally used before fx loops to achieve a low impedance drive, thereby mitigating signal degradation.
However, a tone stack generally has a high impedance output, so you may wish to review such an arrangement as part of the design rationale.

Good point

I was planning on using a simple single knob tone control. I haven't decided on what tone control exactly though, will have to consider the output impedance though...

Maybe the pentode should not be direct coupled to the CF. Maybe I should have a high impedance tone control in there as to not degrade the CF...

[Alan0354]

What is the effect of Morph control to the sound? It almost like a negative feedback but to the screen grid instead.

its mostly a form of gain control

I kind of figure that as the screen grid control the plate current also. It's a negative feedback. But what is the advantage of using this instead of simple volume pot or NFB to the grid like the standard way?

[tubeswell]

What is the effect of Morph control to the sound? It almost like a negative feedback but to the screen grid instead.

its mostly a form of gain control

I kind of figure that as the screen grid control the plate current also. It's a negative feedback. But what is the advantage of using this instead of simple volume pot or NFB to the grid like the standard way?

For the most part there is very little sonic difference to a gain control in a guitar amp. But if you use a linear taper pot or connect the log taper up backwards (so that with full CCW rotation, the pentode is in pentode mode), this makes more of subtleties that would otherwise occur in a very narrow band of the pot range (whereby you get screen compression and/or ultra-linear operation depending on the reactance of the screen bypass cap and the other cap with the pot). I found this out by accident BTW the first time I hooked this circuit up.

For a fuller explanation, read pages 95-96 of Merlin Blencowe's pre-amp book (1st edition)

[Alan0354]

For the most part there is very little sonic difference to a gain control in a guitar amp. But if you use a linear taper pot or connect the log taper up backwards (so that with full CCW rotation, the pentode is in pentode mode), this makes more of subtleties that would otherwise occur in a very narrow band of the pot range (whereby you get screen compression and/or ultra-linear operation depending on the reactance of the screen bypass cap and the other cap with the pot). I found this out by accident BTW the first time I hooked this circuit up.

For a fuller explanation, read pages 95-96 of Merlin Blencowe's pre-amp book (1st edition)

Thanks for the explanation, makes a whole world of sense now. You kind of go between pentode and triode by the setting on top of volume.

[tubeswell]

^Yes^. With the pot setting such that the .1uF and .47uF are in series, the plate is AC-coupled to the screen (providing the feedback loop). Using caps in this ('1-to-5') ratio in series pretty much keeps the frequencies that are being fed back to the screen at virtually full bandwidth.

If you want, you can tweak the extent of screen feedback by decreasing the bigger cap - turning this into a tone control.

(Edit -Using the buffer to drive the feedback loop (as per the above schematic), you could afford to devise a pretty elaborate tone control in conjunction with the screen feedback loop, merely because of the excellent load bearing characteristics of the buffer stage. There you go Experiment away to your heart's content. Untested ideas attached in this regard)

[teemuk]

I'd be tempted to replace the "Morph" pot with a dual-ganged one, and wire the other half of it post the gain stage to compensate the gain differences between pentode and triode modes. That way it will more effectively blend between "sonical" triode and pentode characteristics, without their discernible gain difference. Without such compensation the pot is indeed mostly just another volume control.

Oh, since use of "special" tubes or "triode/pentode" modes is pretty pointless if the particular gain stage only runs in its linear area of operation it's adviseable to locate this kind of gain stage to place in signal path where it's most likely to distort. Using it as the first gain stage, not likely to clip or distort due to low input, kinda negates any sonical benefits of this circuit.