transformer and transistor question

[nigelwright7557]

Originally posted by Nico Ras





As you can see that bias can be adjusted quite accurately using a single pot of about 1k and 25 turns. I run mosfets at about 250 mA bias current and there is no crossover distortion.

The fixed resistor is set a little higher and you should never encounter cross-over distortion in this setup.

You can also see from the graphs that the system is stable under varying temperature conditions. It clearly shows that at higher temperature, the bias backs down.

This remains a very simple and robust little amp.

Nico

The trouble is the MOSFETs gate switch on voltage can vary from batch to batch. I found on mine I needed 3.5 volts to switch on a MOSFET (or 7 volts for the pair.)
Also vertical MOSFET's suffer from thermal runaway for its best to use a transistor in the bias circuit bolted to the mains heatsink.
Or to get around the heat problem a large heatsink and a couple of fans so it never gets warm !

[Nico Ras]

I am not sure why they would thermally run away with a heat sink selected for the application. If you remain within the DC SOA, then the FET should be characterized by a negative temperature coefficient, i.e. the hotter it gets the less current it will pass. It acts as varying resistor and its properties follows that.

Obviously when you exceed the thermal limits by not using a suitable heat sink, then things can go horribly wrong.

There is no harm done by using a Vbe multiplier provided you can find one that matches the temperature characteristics of the MOSFET - this is not always easily achieved.

In my humble opinion, I think designers use Vbe multipliers in MOSFET designs to get away with substandard heat sinks.

Let him experiment and decide what is best for his application.

Nico