[patrick-turner]

More from me about wien bridge oscilator.

The idea of passing idc through fet does not make any difference to the nature of the effective Rd - it remains a very non linear resistance.

WB oscilators with lamps or thermistors are probably all capable of producing far less THD. But perhaps using current source from a bjt collector with PNP MJE350 to pass some dc through 3 x 12V/50mA lamps to warm them up so R = 100r, so 3 give R = 300r, and then NFB network total R = 900r, and that may work well. The main FB R = 600r approx, and would be cap coupled to lamps, and Idc controlled with rectified voltage to MJE350 base and current limited with say 220r emitter R between +20V and emitter.

"Grain of wheat" 12V lamps have about 27r at 20C and when hot are 240r.

I could get a type 327 lamp, 28Vx40mA = 700r at max on current, and maybe 240r at situation where you wan to use them in WB oscilator, sometimes with cathode Idc flowing through one at V1 cathode of a 2 stage tubes oscilator. I recall HP's first WB oscilator had this.

Patrick Turner.