2
\$\begingroup\$

What is the main difference between a mosfet transistor and a BJT transistor.

I know both are used for amplification and can be used as switches.

But where would you use one over the other and what are the benefits of each one ?

\$\endgroup\$
  • \$\begingroup\$ You mean MOSFET (NMOS or PMOS) and BJT (NPN or PNP) transistor. You can edit the title. \$\endgroup\$ – DKNguyen Jan 27 at 20:19
  • \$\begingroup\$ Just google "difference between MOSFET and BJT transistors" \$\endgroup\$ – Shadow Jan 28 at 10:20
8
\$\begingroup\$

This question is really broad.

  • MOSFETs switch faster as switches (they transition between zero and full conduction faster). BJTs are faster (have higher bandwidth) as amplifiers devices (BJTs can move around within the "partially on" region where linear amplification occurs faster.

  • BJTs need continuous signal current to control. MOSFETs do not (only voltage and transient currents to charge the gate-source capacitor). This is a strike against BJTs in high power circuits where your control current is proportional to your load current, or in amplifiers where the source impedance is very high and can provide very little drive current.

  • MOSFETs behave as a resistance when fully conducting. BJTs behave as a diode. This is a benefit for lower voltage MOSFETs (<200~500V) since the voltage drop for resistance is less than the voltage drop for the BJT which means it runs cooler. But as you construct the transistors to tolerate higher voltages, the MOSFET resistance increases to a point where the voltage drop is larger than that of BJTs and the BJT becomes more efficient.

  • MOSFET minimun gate voltage to get it to conduct is higher than the minimum base-emitter voltage required to drive a base current into a BJT to get it to conduct (at least for now) . Important if you are using very low voltages to control a transistor switch.

  • MOSFETs have a parasitic anti-parallel body diode (i.e. they can only block current in one direction). BJTs do not. This can be good and bad. In some circuits with inductive loads, it means you might be able to get away without an external flyback diode and instead rely on the body diode MOSFET, even though you really should have an external one optimized for the task.

  • MOSFETs can current share (put in parallel to share a load current) more easily than BJTs without thermal runaway.

  • BJTs have higher transconductance so tend to work better as amplifiers.

  • BJTs are lower noise. Better for sensitive amplifiers.

  • For integrated circuits, BJTs are easier to match (provides numerous advantages for analog circuits), but MOSFETs can be fabricated are smaller (required for high density digital circuits).

\$\endgroup\$
  • \$\begingroup\$ When used as a switch in saturation a BJT can conduct down to 100-200mV, it is not like a diode. The body diode of a MOSFET is in the wrong place to act as a normal flyback diode, although luckily most MOSFETs are rated to go into avalanche safely so can often be used without a flyback diode. Few discrete MOSFETs are suitable for linear operation - they can have significant safe operating area restrictions if they are used that way. MOSFETs can have very large gate to source and gate to drain capacitances that can be a problem to drive. \$\endgroup\$ – Kevin White Jan 27 at 22:12
  • \$\begingroup\$ @KevinWhite 100-200mV, but that is a fixed voltage with current right? And does not increase nor drop with current? That's pretty diode-like. As for body diodes being used for flyback that usually for half-bridges. \$\endgroup\$ – DKNguyen Jan 27 at 22:34
  • \$\begingroup\$ @KevinWhite Body diode in MOSFET is in the right place for a bridge circuit. BJT can reach low saturation voltage, but current gain goes to zero as saturation voltage reduces, and Base current causes a residual voltage. So not exactly like a diode - unless it's a Darlington - but not like a MOSFET either. \$\endgroup\$ – Bruce Abbott Jan 29 at 6:02
  • \$\begingroup\$ @BruceAbbott - Yes, when you have a bridge or half-bridge the body diode can be exploited, but not in the case of a single device. Yes, there is a residual voltage in a saturated BJT (which can be as low as a few millivolts with some transistors) but the device can pass reverse current - so it is not at all like a diode. The current gain does not go to zero as the saturation voltage is reduced, even with reverse current there is some current gain (less than 1 with modern transistors but could be as high as the normal gain). \$\endgroup\$ – Kevin White Jan 29 at 15:14
1
\$\begingroup\$

Without discussing the differences, I suggest you visit diyAudio.com, and find the thread "simplistic NJFET RIAA", wherein the zero input bias current of some Hitachi NJFETs are crucial in interfacing to Moving Coil vinyl record audio cartridges (no DC blocking needed). To avoid hot-gate-electron aging issues that cause Vthreshold shifts and thus GM shifts and thus channel-channel mismatches, the source-drain voltage is greatly reduced, in a 35 volt preamp circuit, by using NPN bipolars to provide cascading of the NJFET and also greatly reduce the input capacitance (no Miller Effect on the LARGE NJFETs), so the preamp has excellent use of the cartridge's precious high-frequency energy.

Also, for excellent OUTPUT_TO_INPUT crosstalk via the Ground Plane, the Ground system consists of two large regions separated by a thin region used to accept return currents from the RIAA pole-zero components.

To achieve this in bipolar ---- you'd need input Darlingtons (and lots more thermal noise), or DC_blocking capacitors at the most sensitive (largest signal-swing, before compensation occurs) node.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.