EF86 with morph control in PP gigging amp

[Alex/Tubewonder]

Precisely, but I'd use a linear pot. IMO it gives a better gain vs knob rotation spread.

[tubeswell]

Sounds good thanks Alex. A new question has arisen

I compared the voltages of this with the EF86 SE amp I built a few months ago http://music-electronics-forum.com/t23443/, and I noticed that with the DC-coupled CF stage, the screen voltage seems to have gone up by about 10V (compared to the same HT with the same Ra and Rg2 and Rk without the DC-coupled CF stage). I wonder whether this is a side-effect of having the DC-coupled CF stage 'stealing' extra current through the EF86 Ra? (Not sure about this tho'?) - also this may be more exaggerated because its an ECC82 CF stage (higher current tube).

The voltage comparison:

EF86 without DC coupled CF stage with a 100k Ra, 560k Rg2, 680R Rk
HT = 209
Va = 92
Vg2 = 69
Vk = 0.99

With the DC-coupled CF stage (with morph control to CF grid) with a 100k Ra, 560k Rg2, 680R Rk (CF cathode 47k)
HT = 203
Va = 88
Vg2 = 83
Vk = 0.93

12AU7 cathode = 94

With the DC-coupled CF stage - morph control to CF cathode/output with a 100k Ra, 560k Rg2, 680R Rk (CF cathode 47k)
HT = 214
Va = 86
Vg2 = 80
Vk = 0.99

12AU7 cathode = 91

Seeing as how the CF stage alters the voltages slightly and the screen is now sitting at around 80V, I was wondering about trying a different stage set up for the EF86 to get the load line through the knee more, and I think I have it worked out as per the attachment, (56k load line with 470R Rk) but I would appreciate some feedback about it before I implement it, if you have a moment.

The calculated bias resistor to get -1V is 438R and I gather that if I pick 470R the bias point will be a bit cooler that -1V?

And should I change the Rg2 to 680k (if I do this) because I am wondering about how a 56k load will affect the tube current given the DC-coupled CF stage?

TIA
Pete

(The different schematics with measured voltages noted from the existing build are also attached)

[Alex/Tubewonder]

I wouldn't worry about these slight changes in operating conditions of the EF86:
primo: this is a tube amp, not medical tincture where any microscopic deviation from the recipe counts
secundo: as soon as you start turn the morph control towards CF cathode EF86 will begin entering varying degrees of UL operation and end up as triode without knees.
Let the wax in your ears be the judge

[tubeswell]

Thanks Alex.

Anyone else got any comments about the 56k load line idea? (and appropriate g2 resistance for such?)