#Fet transistor as a switch drivers#
#Fet transistor as a switch series#

Additionally, and this is really important!: MOSFETs can be placed in parallel to increase a circuit's current-capability.Example: a TIP120 NPN BJT Darlington transistor can only drive about 5A continuous current, whereas the IRLB8721 N-Channel Logic-Level MOSFET, in the same physical TO-220 package, can drive as much as 62A.So, MOSFETs excel at driving high currents and high powers. For any given component package size, my personal experience in searching for parts indicates the best BJT transistors can only drive about 1/10 as much current as the best MOSFET transistors.When your power and current requirements ARE a dominating factor.Also (see the book reference, quotes, and example problem below!) BJT transistors can switch a touch faster than MOSFETs (ex: 15.3 GHz vs 9.7 GHz in 'Example G.3' below).So, in summary: the faster your switching frequency, the more MOSFET transistors lose their efficiency gains they otherwise naturally have over BJT transistors, and the more BJT transistors begin to be appealing from a 'low power' stand-point.Rapid switching of the Gate also significantly increases losses through the primary Drain to Source channel because the faster your switching frequency, the more time (or times per second, however you want to think about it) you spend in the Ohmic region of the transistor, which is the region between fully ON and fully OFF, where R_DS (resistance from Drain to Source) is high, and hence, so are losses and heat production.So, faster switching frequencies means more losses in driving the Gate of a MOSFET, as opposed to driving the Base of a BJT.

#Fet transistor as a switch drivers#

High Gate capacitances on MOSFETs with high-speed switching means you must have large Gate drivers and very high drive currents to a MOSFET (ex: +/-500mA), as opposed to the low drive currents to a BJT (ex: 50mA). This means each time you double the current you quadruple the power losses and heat generation in a part.

The high gate capacitances can involve rather large (up to hundreds of mA, for example, for a TO-220-sized part) momentary input and output currents, and power losses are proportional to the square of the current ( P = I^2 * R).

Charging and discharging their Gate capacitances repeatedly is like charging and discharging a tiny little battery repeatedly, and that takes power and current, especially since you are likely discharging that tiny little charge to GND, which is just dumping it and converting it into heat instead of recovering it.

MOSFETs start losing their efficiency gains the faster you switch them, because:.When your switching frequencies aren't too high.This makes MOSFETs ideally-suited to low-power applications, because you can make them draw a lot less power, especially in steady-state (ex: always ON) scenarios. BJT transistors, on the other hand, are current-controlled, so to keep them on you have to keep sourcing (for NPN) or sinking (for PNP) current through their Base to Emitter channel. So, you can just charge their Gate once and now you have no more current draw, and they stay on. When is a MOSFET more appropriate as a switch than a BJT? Answer: 1) a MOSFET is better than a BJT when: When is a MOSFET more appropriate than a BJT? Why am I continually being told that I should be using MOSFETs? There doesn't seem to be a great surplus of logic-level MOSFETs that can switch the current that cheap BJTs can (~600-800mA for a 2N2222, for example), and the ones that do exist (TN0702, for example) are hard to find and significantly more expensive.

#Fet transistor as a switch series#

A MOSFET also requires no resistor in series with the gate, but generally DOES require a pull-down resistor so the gate doesn't float when the MCU is rebooted (right?).

I do know that a MOSFET wastes no current on the gate, where a BJT's base does, but this is not an issue for me, as I'm not running on batteries. I've been repeatedly told, however, that N-channel enhancement-mode MOSFETs are a better choice for switches (see here and here, for examples), but I'm not sure I understand why. In my experimentation, I've used only BJTs as switches (for turning on and off things like LEDs and such) for my MCU outputs.