"Greg Mayman" wrote to "Mike Ross" (09 Jan 03  08:22:00)
 --- on the topic of "I/F"

 GM> I suggest you set up the FET as a shunt attenuator, set the gate
 GM> voltage for 6dB attenuation with about a 1v RMS input signal, and
 GM> measure the output on a good distortion analyzer. Then add the
 GM> audio feedback from the drain to the gate and see how the
 GM> distortion decreases.

1V rms?! You aren't joking, I guess not. A J-Fet's junction will be
biased on by 0.6V so imagine 1.414V peak. No wonder you're trying to use
feedback on the poor thing!  Really, I wouldn't apply more than about
20 or 30 "milli"-volts AC to it. This keeps the signal within the
junction saturation current region beyond which non-linear distortion
becomes a problem with small signals. Remember, though the channel
resistance can be modulated, a J-Fet is still a unipolar device.
However, I'll try your suggested circuit and see if it helps.

 Mike
 ****

... Thomas Edison invented the "Light Emitting Resistor"
--- Blue Wave/DOS v2.30