Adding servo you'll loose most benefits of jfet stage. In this case it'll sound like simple opamp with clipping diodes.
Simpler tunung but poorer sound.

I don't understand what do you mean by "explain the graphs".
But the processes is the following:
When grid current begins, constant amount of voltage at inter-stage coupling capacitor rises.
This lead to creating virtual negative biasing of the control grid. Stage operating point shifts closer to cutoff.

Accurate emulation of grid current (shift of operatin point) often means more than a shape of clipping.
Look how tube stage respond on "burst" input signal.



K1 - Anode voltage
K2 - Input voltage
K3 - Grid voltage
K4 - Cathode voltage
Output signal  "floats" in time at a constant input level.
Average current varies almost 2 times.
Amount of changes depends on interstage circuitry (impedance, time constant etc).
That's why guitar players often say that tube sound "live" and the sound of semiconductors "dead" (uniform).
This refers not to the color of harmonics but to reaction on the sound picking.

Mini head project - LND150 based JCM800 preamp + 5W PA.

Quote from: J M Fahey on September 11, 2012, 12:35:00 PM
PS: as a side note: your actual physical construction is *so* good, that I thought those were not pictures but some CAD design, go figure  :cheesy:

You're right, it's rendering from SolidWorks.

40W transformer output PA Module.

Partially SMD version.




All through hole version.



Information about both versions  there.

Don`t forget another important thing - specific "voices" of different speakers.
One speaker maybe good for vintage sound but poor for modern.
http://www.youtube.com/watch?v=jWK0sa7tlfI

Simply select "AC Analysis" instead of "Transient" in "Edit Simulation Command" dialog.
Put voltage source at circuit input. Set "AC Amplitude" = 1 in advanced dialog for it.
Run simulation & see frequency response in any point of your schematic.

Simulator is very useful when you creating "skeleton" of your schematic.
It reduce time to adjust component values.
Final tuning of course must be done in "hardware".

I use LTspice, it`s free of charge, and good enough if you get adequate component models (especially for vacuum tubes).
It also allows to pass through circuit WAV files & record result.
Note, processing 20 sec WAV file through tube preamp ENGL530 takes about 30 min.
http://www.linear.com/designtools/software/#LTspice

Comparision records (made by my friend) of 6V6 PP PA & 5W fet PA:
http://milas.spb.ru/~kmg/files/projects/fetpa/mp/samples/ModS%20-%206V6PP%20vs%20KMG%20FET%205W.rar
They recorded using MOD_S custom amplifier using its own PA & 5W fet PA connected to "Send" jack.
Frequency response of both PA was equalised using Presence & Depth controls.

Recordings are unnamed (numbered) so you can try to estimate "who is who".

QuoteHave you tried any dual rail designs with complimentary pairs of lateral FETs into a tx?

I tried this prototype, but it has zero temperature drift & I finished experiments with it.
http://milas.spb.ru/~kmg/files/projects/fetpa/nt/FetPaNt.pdf
Also you can try to read this article
http://translate.google.com/translate?hl=en&sl=auto&tl=en&u=http%3A%2F%2Fgtlab.net%2Fgtlab4%2Farchives%2F30

bassetrox, you can use LTspice simulation files from there
http://milas.spb.ru/~kmg/files/projects/fetpa/mp/sim/

Assembled micro PA:

Assembled 5 W PA