A little amp with 10 Watt in Class AB with EXICON Laterals.
The power supply should be like 30-34 Volt DC.
So, a suitable transfomer can be 24VAC or 2x12VAC.
Simulation gives an output of 11.4 Watt with 33 VDC supply.
The distortion is rather low for such a basic amplifier. SPICE specify it as THD 0.0001%.
Bias of the MOSFETs is 200mA.
The amplifier is Two Pole Compensated, TPC.
The power supply should be like 30-34 Volt DC.
So, a suitable transfomer can be 24VAC or 2x12VAC.
Simulation gives an output of 11.4 Watt with 33 VDC supply.
The distortion is rather low for such a basic amplifier. SPICE specify it as THD 0.0001%.
Bias of the MOSFETs is 200mA.
The amplifier is Two Pole Compensated, TPC.
Last edited:
No, not absolute. Most time I only add a potentiometer.
But this time I added BD139.
U4 will increase biasing error, not reduce it. Lateral MOSFETs are thermally stable - also the bias voltage for some can be very small, less than a Vbe multiplier is capable of generating. Once hot U4 will starve the output devices of bias and might lead to oscillations.
Lateral biasing is easy, a single pot can do it.
The whole DC feedback system worries me - it seems to rely on hfe being fixed and stable and not verying between devices, none of which is true.
Lateral biasing is easy, a single pot can do it.
The whole DC feedback system worries me - it seems to rely on hfe being fixed and stable and not verying between devices, none of which is true.
Not absolute, but very good idea. As the supply drops under load, using just a resistor results in being under biased. The vbe multiplier won’t. If you were using a true ccs instead of a boot strap, then this effect would be eliminated.
Avagadro’s number of watts will not kill a simulated 2N5401. Don‘t even try 6.02 watts with a real one.
Of course it will have problems in real life. The simulator won’t tell you that unless you ask it (ie, turn those features on).
If you’re just occasionally clipping you might get away with it. But eventually it will catch up with you. Now add that 1N4148 and there may just be enough nonlinear capacitance there to double the THD. Simulators won’t capture all the causes of distortion either, unless the model completely describes the circuit and layout. So I never get excited about 4th or 5th decimal place results. The circuit has nothing really wrong with it, and probably sounds pretty damn good. But I don’t necessarily believe you’l get any .0001%, baker clamp or no baker clamp. But I’d put one in there, unless I had an infinite supply of expendable 2N5401’s. I can tell you I’d clip it enough to fry it.
