Semiconductor Devices: Class B Distortion Sims

[@LousyPainter]

You're a very good instructor. Thanks.

[@christopherventer6391]

Curious if it would be better (or how much better) to use the base and emitter of two more of the same transistors instead of diodes at the front end.

[@christopherventer6391]

I simulated this just now, actually (in a different software, which unfortunately doesn't give me a THD figure) and it seems like using the BE junctions of two transistors instead of 1n914s is more tolerant of higher R1 and R2 values without crossover distortion (up to 100k seems to look pretty good, after which point I start to get obvious clipping).

[@ElectronicswithProfessorFiore]

@@christopherventer6391 Yes, that would be the case, and in integrated circuits we often make matching "diodes" by just creating a transistor and shorting the base-collector junction. The practical problem with a power amplifier is that you'd be using comparatively expensive power transistors as diodes (and at a relatively low current and power level, as well). The usual route is to use diodes as shown, and then apply some global negative feedback to the circuit to reduce what distortion remains.

[@christopherventer6391]

@@ElectronicswithProfessorFiore That makes sense! In the meantime I turned my worst-performing test (BAT54) and an OPA and wound up making a wavefolder lol