What are the pros and cons of BJT versus FET transistors?

For pros and cons of BJT versus FET transistors first we needs to know whats its use in circuit and then we can understand benefit and losses of BJT vs FET transistors.

Use of Bjt: this is the best device in most analog applications ‘linear’. switches, logic, etc. more suitable for cmos. However, getting access to bjt well on modern processes is less common. integration is a process technology that moves into cmos and bipolar more niche.

  • Bjt has an exponential relationship: ic = is*exp( vbe / vt )
  • Cmos is approaching because the channel lengths are decreasing on the modern process nodes .. but that is not quite the case yet. I predict that it can reach exp () but not exceed it.
  • This great signal characteristic is strongly non-linear (although many say that the bjt is more linear, this is actually incorrect). this high derivative results in a high transconductance. transconductance (gm) is an excellent tool in analog circuits. it provides gain, speed (more than gm helps with feedback compensation) and it does it with a lower bias current (smaller devices).
  • The exp () characteristics give simple and reliable equations for hand analysis .. not really a reason to use the device, but it’s easier to manage.
  • Lower noise than cmos. better gm, better noise.
  • Minimal flicker noise. cmos is notoriously bad for glittering noise especially at the finer geometries.
  • Better output impedance.
  • Much higher intrinsic device gain.
  • Better match (unlike the previous entries), the bjt has a lower entry referred to as vbe offset. the reason why some people see as more sensitivity is that transconductance is so great, small offsets lead to larger current bias deviations. It’s easy to manage. I would prefer to have the low input shift compared to a bad match with a bad GM to stabilize it.

Uses of Cmos: the only real choice for logic, but it has some nice analog advantages as well. the biggest boost for cmos is integration with digital applications. integration pushes all ic development into cmos for digital content. only more niche apps still use bipolar.

  • Switches. cmos makes great switches.
  • Door leak. cmos does not have a door current (the modern process nodes start to show more door leakage, but essentially zero by the bjt standards). this is important in switched capacitor circuits.
  • Triode region is well behaved. If the bjt enters the saturation region, the base current can become very high, which can affect the bias circuits and block the circuit. positive feedback. As a result, Bjt bias is more complex and increases the risk of unplanned transients in the circuit.

Pros for BJT:

Bjt on physical dimension and same price can usually give you much higher speed, because they have little input capacitance.

Bjt can give you much higher profits. just take a bunch of components and compare them, and you’ll find bjt gives you better gain characteristics and therefore requires fewer gain steps.

The bjt amplifier stage is much more linear than the MOSFET amplifier stage, since the gain is independent of the bias voltage. this gives better loyalty.

Bjt is capable of handling higher output current for signal output and can have lower output impedance. in amplifiers intended to drive a low impedance input load or provide large amounts of power, this is a huge advantage. many high quality op amps are made with bicmos process using bjt pair for output buffer stage.

Cons for BJT:

Bjt is a device that is operated at this time rather than operated with voltage. most of the time this means higher power consumption.

Bjt’s has an essentially reversible input-output relationship by biasing the terminal opposite to the intended design. what this means is that a sloppy amplifier design can more easily end up with positive feedback and therefore become unstable. I’ve seen this before in one of my previous designs at school, and I do not know it. (I did, but not long later).

Bjt has a lower input impedance. this means the higher output impedance device may not match the input bjt.

Bjt’s are harder to scale.  fets are easy.

Related: bjt has significant variation of transistor to transistor, while fets not so much.

It’s harder to make a good current mirror than bjt than fet.

Pros for MOSFET:

They are very easy to measure. want exactly half the current? just make the gate half wide.

High input impedance.  at low frequencies, essentially infinite.

Because the output is controlled by input voltage instead of input current, they consume can consume very small power. cmos logic winning reason over other process technology is power consumption, combined with scalability. in the cmos circuit, the gate to the power gate is consumed only when the switch circuit states, while other logic forms consume power to remain in the active or inactive position.

Considering most digital circuits are cmos, using fets for analog remote circuit design, it is much cheaper than using bjt because it does not require a series of step additional process.

Easy enough to make the MOSFET almost identical. simple transistor-to-transistor matching.

MOSFET is easier to design a stable circuit with. unless you add feedback, they do not suddenly act upside down and become unstable.

Cons for MOSFET:

Not as high as loyalty as bjt, since the gain will be slightly different as you increase the input voltage (that is, it will produce very weak harmonics).

Input capacitance. the higher the gain, the greater the input capacitance thanks to the miller effect.

Can’t drive a low-impedance load very well.

Low gain per section, which often means more amplification stages are required for higher gain, even when using advanced design techniques. each amplification stage adds noise – that is, you can never get a better signal-to-noise ratio at the output than on the input.

This list is not comprehensive. it’s just the best I can do from the top of my head.

Recent Updates