Plexi-Drive JFET misery

[diydave]

Maybe this will help. This has never failed me yet.

Take a fet (to-92) and a breadboard.
Connect Gate to ground.
Connect Source to ground.
Take your multi-meter, set it to mA, and hold the black probe at the fet's Drain and the red probe at +9V.
What you read on the screen is the Idss. Note it down.

Now, Connect Drain to +9V and place a 10 Mega-Ohm resistor between the fet's Source and ground (keeping the Gate connected to Ground).
Take your multi-meter, set it to mV, and measure the voltage across the resistor. What you read is the Pinch-off Voltage (Vp) of your fet.

Say you get the following readings: Idss = 0.77 mA and Vp = 0.64 V (this is from an actual j201)
If I want to bias the fet midway, then I take Vgs to be 0.32 V.

We have this formula (google is your friend):
Id = Idss x { 1 - (Vgs/Vp) }²

So Id = 0.77mA x {1 - (0.32V/0.64V)}² gives 0.1925 mA (or about 1/4th of Idss - a rule of thumb so I've found).

Now we know the Id (0.1925mA) for a given Vgs (0.32V).
According to Ohm, this means a Source-resistor off (0.32v / 0.1925mA =) 1.6 k-Ohm (take 1k5).
To drop 4.5 V between +9V and Drain, we need a resistor off (4.5v / 0.1925mA = ) 23k Ohm (take 22k).
So to bias my fet, I need a 1k5 source-resistor and a 22k drain-resistor and I will get half Vp on the Gate and half Vs on the Drain.

I've numbered all my fet's and measured the Idss and Vp off all off them in this manner. If I want matched fets, I take the once with the closest specs. If I want to bias a fet in a certain way (midway, near Vp, near 0v, etc...), I use a spreadsheet which calculates the Rs and Rd for me.

Another way I tackle my fet baising:
I use the Source-resistor which is on the circuit's schematic. I measure the Id by placing the multi-meter between Drain and +9V (put the multi-meter on mA). Once I know the Id, it's simple math to calculate a drain-resistor which would drop 4.5 volts.