MOSFETs are unipolar devices, i. H. There is only one type of carrier in action, generally electrons. BJTs are bipolar devices, i.e. H. Both electrons and holes conduct electricity within a BJT. \ ‘B \’ stands for bipolar.
The MOSFET can therefore switch on the time scale of the opening / closing of its conduction channel, which is determined by the speed of the change in the electric field via the gate oxide. This is in the nanosecond range for typical MOSFETs.
BJTs. On the other hand, only on the time scale of the “recombination lifetime” of the charge carriers, i.e. H. Electrons and holes to be switched. BJTs carry electricity (when switched on) by flooding their interior with many electrons and holes. They are of the opposite charge type – positive and negative. Before switching off completely, all these carriers must recombine with one another (positive and negative recombinations in charge-neutral states). This time scale is on the order of microseconds. Regardless of the size, the switching speed of the BJTs is therefore in microseconds compared to that of MOSFETs in nanoseconds.
A BJT is more like a constant saturation voltage, so the losses are only Vsat x I. You can see that at high currents in MOSFETs the losses increase exponentially while the BJT losses increase linearly.
The MOSFET resistors are getting lower, so the crossover point for BJTs that are better gets higher. However, BJTs require significant base drive power, which offsets any benefit from loss of power. In practice, BJTs are only used up to a few watts because they can be cheaper. After that, MOSFETs win.