12SJ7 Screen By-pass cap calculation

[Paul Barker]

Having trouble with my Thevenin calculations.

If I put just a screen resistor from B+ to Screen, (B+330v desired screen 100v, expected screen current 1.5 so screen resistor 150k)

Then I have to bypass for my frequency of interest the screen impedance (do we call this rg2?) and Rg2.

For rg2 I calculate 24k, which parallel with Rg2 is 20689R to bypass that at 5hz we need 1.5uF

Now here is the thevenin question.

If I use a voltage devider (as I am assuming it is a closer thing to using zenners or VR tubes without going the whole hog), I suppose I have to parallel Rg2 and rg2 with the g2 to ground resistor. Right?

I am finding by the way some more accurate curves for the 6sj7 at the voltages I want and am now predicting better what gain to expect. for the previous project of driving the 6080 CF output stage the mu I achieved must have been 50, which was pitifully poor.

Anyway armed with hopefully accurate curves I can see an operating point of:

anode voltage 100v screen voltage 100v
transconductance (which you need to calculate gain) 1.57
anode resisatnce 0.7 megohm
With a load resistor of 77k
direct coupled = 1megohm

I get a gain of 100

I can live with that but I'll just put a simple dht as the output valve.

I'll be running the pentode at 3mA so I can slew 45 6b4g px4 that kind of thing.

[Paul Barker]

Oh forget it. Plenty of bass with 0.47 in it. Should be -3db at 15hz but it's the only high pass filter which encroaches into the coveted 20hz zone.

I got more gain fiddling with it and now it is actually driving the 6080 SRPP output stage (output parafed from Cathode of top valve).

The 6080 is worth keeping in there because of the beautiful bass quality I am getting. Something I don't normally get.

The damping factor is the main reason. But also the OPT's are wound for no phase rotation down to 20hz.

[IslandPink]

It's probably affecting the bass timing a bit, if it's at 15Hz, but if it sounds OK to you, that's the main thing.
Incidentally, James gave me a handy rule of thumb for pentodes in normal operation : the g2 impedance is roughly 4 times the value of the anode impedance when the valve is triode-connected .

[Paul Barker]

thanksMike. I'll add a 1 uF cap inthatcase.