Phase inverter question

[Patrick M]

My amo has evolved quite a bit - kind of a 'living breadboard' of some sort. Started as a TTT inspired 6L6 fixed bias, with normal and overdrive channels, master volume and reverb. One of the first change was to go to cathide-biased 6V6 (like a 5e3 output stage). I can change from 6V6 to 6L6 by switching the transformer impedance setting (8Ohm for 6L6 - original configuration, and 4Ohm - with 8Ohm speaker - for 6V6)

After multiple changes, it has evolved in its the current configuration. The last (and maybe most significant) change has been the addition of a VVR. I got rid of the overdrive channel (which was picking some noise anyway, and the master volume. I like this a lot, as I can set the amp just right and control it from my guitar pickup settings (tele with 4 way switch), volume and tone. This is absolutely great, and I don't even use any pedals (except for the tuner pedal) anymore - I just like 'playing the amp' better.

Here is my main questions for the more experienced amp builders:
However, when overdriven too hard, sometimes it can be just a little 'farty'. I wonder if it is the concertino phase inverter (form I have read about it). Since I have a spare 1/2 tube, since removing the master volume, would it be better to concert it to a long tail phase inverter, like one the deluxe reverb? That would not be too hard to do, but I would like some opinions before doing one more change!

Current configuration:
Preamp with tone stack - one 12AX7
reverb driver 12AT7
Reverb recovery and summing stage: one 12AX7
Phase inverter: one half 12AX7 - (change to 12AT7 and long tail inverter?)
Cathode biased 6V6 PP output stage
Feedback on/off switch
VVR on the output stage

[J M Fahey]

Hi Patrick.
Please post the schematic.
Thanks.
PS: Don't think that one phase inverter type "is fartier" than another by itself, probably the amount of bass injected into it , coupling caps, etc. have a more direct influence.
Anyway many great builders around here will probably have something to say about your design.

[pdf64]

Agree with above - fartiness is generally resolved by lowering the value of coupling caps, maybe adding grid stoppers.
An LTP would add voltage gain (a bit over 20 for a 12AT7 with regular Fender BF values),making things worse, all else being equal.
What coupling caps are you using into and out of the existing phase splitter?

[Patrick M]

Here is the schematic as it is now.

[pdf64]

Are you sure that's the latest schematic? From what you've said (eg the drive channel has been removed, VVR added) it looks like a previous version.
The time constants of the RC at the power tube grid are too large, IMO, and may allow blocking distortion when overdriven. Try halving or quarter the cap value, 0.047 or 0.022uF.

[Patrick M]

Oops - wrong version

That is what happens when you have a 'work in progress'.

[defaced]

Are you sure that's the latest schematic? From what you've said (eg the drive channel has been removed, VVR added) it looks like a previous version.
The time constants of the RC at the power tube grid are too large, IMO, and may allow blocking distortion when overdriven. Try halving or quarter the cap value, 0.047 or 0.022uF.

Agreed. This is exactly where I would start. The next place would be the 0.047uF between the gain stage and the phase inverter triode. I'd drop that to 0.022uF. My amp has a cathodyne phase inverter, and when it falls apart it doesn't fart out, it just loses definition.

[Merlinb]

However, when overdriven too hard, sometimes it can be just a little 'farty'. I wonder if it is the concertino phase inverter (form I have read about it).

Most people find this can be cured by adding a very large grid stopper, 470k say, to the concertina splitter. In your case this would go between R46 and the grid.
See here for the reason why this works:
The Valve Wizard

[pdf64]

Have you looked at the cathodyne plate signal when the power tubes overdrive?
There's a massive blip when the power tube grid conducts and acts to bypass the cathodyne cathode, resulting in a gain boost. It doesn't seem to have any bad audio effect though.
I agree with defaced about that cap at the input to the cathodyne. The 1M grid feed resistor is bootstrapped, so it's effective value is greater than 1M. So 4n7 would be more than adequate. And a 100k grid stopper on the cathodyne - because there's no voltage gain, it won't affect the audio high frequency response, but will mitigate for blocking distortion.
As you've got that gain stage before the cathodyne, an LTP would have less voltage gain, so might be an option, but try the above first.
Also check out merlin's site on phase splitters.