PI instability

[catalin gramada]

Hello. I tried to put a volume pot in front of the pi stage from image and have a issue to put the potentiometer to minimum level position. When turn off the pot at around 10 ohm resistance to ground the stage start to oscillate. I tried a series resistance in series with the input cap. that solved the problem to close the pot at minimum level but now have instability in an other position of potentiometer The oscillation is sort of buzzing sound. What sort of issue could be and how can be cured.please? It is a grounding problem maybe? Thanks.

[ATTACH=CONFIG]42004[/ATTACH

[pdf64]

I would add grid stoppers, mounted at the tube socket terminals; Gingertube provided a formula that indicated 2k2 for a 12AT7.

[catalin gramada]

Hi. Thanks. I.ll give a try. I don.t understand... Fender used similar circuit. the only difference they float the ground PI point over 100 ohm resistor means nfb load, maybe this did a difference...?

[catalin gramada]

allright. I solved the problem doing exactly what Fender did. I just raised the PI ground point in the top of 100 ohm resistor (maybe some other small value works but I did not bored with. 100 ohm is small enough) even I did not intend to use the NFB here. Now I can put the MV pot down without any problem. Still I wish to know why those parasitic coupling appears with the PI grounded directly to signal ground path, please, Thanks

[pdf64]

Did you try grid stoppers (they may be considered as basic good practice, and if missing, should be the 'go to' method of improving a system's margin of stability).
The Fender operating point looks to be colder than yours, eg you have a 150 ohm cathode bias resistor, they tend to use 470 ohms.
The 12AT7 is suitable for use in RF applications.
Note that Fender had trouble maintaining stability in their amps, hence various changes to circuits over the SF era.

Also there's the rest of the system to consider, unless just that LTP is oscillating, ie without anything else connected to its input and output.

[mozz]

I have seen some 12at7's howl, the ones that had single micas, the double micas thickness don't seem to have this problem.

[catalin gramada]

Yes I know. I try to use as much current as convenable to drive some capacitive loads (pentodes triode straped). You are right regards 12at7 stability. For instance the datasheet asked for 1M grid leaks which I found unusable. Values as low as 470k did not works. I choosed 330k with a small reserve. 270k pretty safe.(but think it is a current dependant issue) Regards Fender I wonder about heater to cathode limit they large overpassed 120 to 170v range whilst datasheet mention no more than 90v max limit

[pdf64]

The grid to cathode resistance is a max limit, not a minimum.
So you can go as low as the preceding stage can drive sufficiently.
The cathode bias grid to cathode resistance limit for 12AT7 doesn't seem to be noted in most manufacturer's info; I found this one which advises 0.5M ohms http://www.mif.pg.gda.pl/homepages/f.../1/12AT7WA.pdf
I think that most Fender models comply with a 90V cathode to heater voltage limit on their LTP.
Rather it tends to be cathode followers where it gets exceeded.

[catalin gramada]

late edit: I put some large 22k resistor sticked directly to the socket and changed the operation point up to 2.5v Now it works even with 1M Rg1. Thanks for ideea.