JFET vs BJT

[ele1200]

Hi guys,<p>You all sound smart and all and well, i just started electronics and was wondering why you would use a JFET vs a BJT. I know the electronic difference (in construction) by why!!!!!<p>thanks

[russlk]

The JFET has high input impedance, so it is useful when you don't want to load the source. It is a depletion mode device, which means it is normally "on" and you have to reverse bias the gate-source junction to turn it off. That mode is similer to a vacuum tube. The BJT input is a forward biased diode which is a low and non-linear impedance. Circuit tricks, such as putting a resistor in series with the emitter to raise and linearize the input impedance are used.
The output impedance of the JFET and BJT is similar, both are pretty much current sources.<p>The MOSFET is different in that it does not have a gate junction and it is an enhancment mode device. That means it is normally "off" and you have to forward bias the gate-source to turn it on.

[billdar]

In the digital world, the choice between BJT and
FET is a choice between speed and powe (heat).<p>BJT's switch much much faster than FET's, but
require much more power.<p>FET's are on the other side of the scale, taking
little current to switch but quite a bit longer
to turn on/off.

[bobsRAC]

Another really important design consideration is the tolerance of the manufacturing process. FETS and JFets in particular have very poor manufacturing tolerances, on the order of 10 times worse than BJTs. FETs also have poorer noise characteristics.<p>To get an idea, browse through the data sheets for some op-amps (you can find hundreds of such data sheets on digi-key.com). You'll notice that the input offset current and voltage for BJT input stage op-amps are much lower than for JFET input stage devices. This is because the transistors used to construct the op-amp are much more closy matched on the BJT devices.<p>Converslt though, if you need a device that has a very high imput impedance, a JFET input stage is far superior.<p>As an example, JFET instrementational amps can be found that only draw 3 femptoAmperes (that's 3*10^-15 Amperes), whereas most BJT instrumentational amps draw currents in the micro to milliAmpere range ( that's 10^-6 or thereabouts). This corresponds to a billionfold increase in bias current. This increase may load the circuit that is driving the input in some cases.
The flip-side is the input offet parameter. BJT devices have input offsets down to single-digit microVolts, whereas FET devices have input offsets in the millivolt range. This is a 1000-fold increase in the offset error, reducing precision.

[bobsRAC]

Also... here's a fun one...<p>Look up the data-sheet for a JFET and look at the data sheet for a typical N-ch JFET the MPF102 (I've got a link below). Look for Idss and look at the min and max values. You'll notice that the min is 2mA and the max is 20mA. This means that some large percentage of the devices (likely 3 std deviations or 99.97%) will be within this range, and the proability of being within one std deviation (or in this case 40% of the mean value) is certainly no more than half (OUCH!).
MPF102