Semiconductor Devices: Class B Bias Stability

[@simonyoungglostog]

Fab! I have vague memories about using matched pairs of BJTs to achieve a smooth output. So a twin 3904 for the top half where the temp-co variance self-cancels. For clarity (again, thinking out loud), with BJTs being cheap, we could get a great match mounting the BJTs together (for heat stability) and tying the collector of the biasing BJT to the base and using it as a diode (same for the PNP side). The amp I designed uses a VBE multiplier (I recognise that part of the circuit).

[@ElectronicswithProfessorFiore]

Along that line, I have seen designs that use biasing diodes mounted to the power transistor heat sink as a means of thermal feedback.

[@simonyoungglostog]

@@ElectronicswithProfessorFiore I have the prepared heatsink for my amp design. It looks like I drilled enough holes for the two final transistors with their corresponding drivers plus the VBE transistor. (It was a mono design).

[@JohnWiku]

Great lesson!

[@pdeepakjayan1993]

Is there any reason for mentioning the current should be small(2 to 3 percent) How did the ammemeter measure 2.31mA for 10v..Is there any equation for finding the current.

[@ElectronicswithProfessorFiore]

It's kind of like idling the engine in a car. If you do that, touching the gas pedal gives you instant response. If the engine is turned off, it has to be restarted, resulting in a delay. But you don't want to idle the engine at 5000 RPM as that would be wasteful. You just need a modest amount.