JFET Gain

[Paul Marossy]

Maybe a dumb question, but how do you determine the gain on a jfet outside of a circuit? I haven't messed with jfets too much. I know that transistors will tell you by looking at the Hfe number. I seem to remember something about Vgs, but that doesn't make sense to me... the data sheet I am looking at just tells you the maximum voltage rating. Can anyone help me to understand?  :oops:

[R.G.]

JFETs are hard. That's one reason that almost all electronic gear is made from either silicon bipolar junction transistors or CMOS. The silicon BJT is predictable - at least, it is easy to get devices that you design around the variations. CMOS offers another avenue to designing around the variation especially in logic circuits.

The magic number for JFETs is gfs - the forward transconductance, which is usually expressed in Siemens or milli-Siemens. A Siemen is the inverse of an ohm.

An Ohm is one volt per ampere. A siemen is one ampere per volt. A milliSiemen is one milliampere per volt.

Looking at a common JFET's data sheet, (2N3819, http://www.vishay.com/document/70238/70238.pdf) we see the specification for the minimum gfs on the second page as 2 to 6.5 mS, or two to 6.5 milliamperes per volt.

That "volt" is Vgs change, so you get a transconductance in the drain-source channel of between two and 6.5 milliamperes for every volt of change in Vgs. Vgs0ff is specified as -3 to -8V, so you have some range in there to work.

High gain JFETs have high transconductances - high current flow per change in gate-source voltage.

The forward gain of a JFET is best expressed as Vo = - Gfs*Vgs*Rd, where the voltages are both the change in voltage, not the absolute voltage. If you have a JFET with Gfs = 2mS, and change  Vgs by one volt, then you get a change of 2ma in the drain conduction, and that change is multiplied by the drain resistor to get the output voltage.  If Rd is 5K in this example, then the output voltage changes by Vo = -2mS*1V*5K = 10V, and the gain is ten - assuming that the power supply voltages and other circuit details are set up to allow that change, of course.

This is just the tiniest hint of JFET stuff. High gain JFETS have high gfs,  (usually) low Vgsoff, and a reasonable Idss, which is what the JFET conducts when the Vgs is 0, and the highest current that the JFET can usually conduct. Using a source resistor lowers the gain in the same way it does for a BJT - the voltage from the drain channel raises the source to subtract some of the Vgs signal voltage away.

JFETs usually have about an order of magnitude lower gain per device than a BJT. You have to want their special characteristics to put up with the variation and oddities and use them.

Ask where it's not clear.

[Paul Marossy]

Thanks for that great explanation, RG.
I understand how that works a whole lot better now!