A few comments.
First, a 6N2P (12AX7) is pretty wimpy to drive the output tubes. IIRC, the grid leak resistor for a 6L6GC should not be less than 100K. There is also some miller capacitance to consider.
A better solution would be to use something like a 6SN7, 6N1P or 12AT7 in an long-tail pair. Use equal value plate resistors (27K works well for a 6SN7 at 400V) with a CCS on the tail. If you can tolerate the idea of using a little silicon, the IXCP10M45S is an excellent little device for this purpose.
The 12AU7 will work in a LTP here as well, but it isn't as linear as the 6SN7.
well I've used a 6N2P for driving 6L6G's (not GC's, 6P3S's which are 6L6G's (my mistake). without any problems with using 220k grid leaks (fixed bias) and I've seen many amps do the same.
and if your interested I moved to a paraphase PI, I liked it much better. anyway, I'll make sure to try lowering the grid leaks and swaping to a 6N1P, I guess I was hoping it wouldn't be a problem and skipped the most important step - reading the datasheet.
Anyway, thank's for suggesting me that.
nvm, I can't read xD. you said "not LESS, than 100k". then I'm fine, I guess
also, I don't (at least didn't mean) to drive the into saturation (class AB2), I was running them in AB1, so it doesn't draw any current. so 6N2P is just fine.
with bigger toobs that might require smaller grid leak resistors, maybe it would be a problem, and running the output in AB2 then ofc a 6N2P would be a problem, maybe barely it could supply the current for 2 toobs but then you would want to use something like a 6N1P (or any tube with a decent plate disapation)