Using cascoded JFETS to bias op-amps into class A

[Dr.H]

Just been reading up on this method-can someone please confirm my understanding:

1. Gate of JFET 1 to -V
2. Drain of JEFT 1 to -V via Resistor R
3. Source of JFET1 to drain of JFET2

4. JFET2 source and gate to -V

Choice of bias current: Choose a current equal to 10% of the maximum ouput current of the op-amp (which it seems should have a buffer following it).

Choice of JFET-Select any N-channel JFET such that its Idss exceeds the bias current required (10% of Iout max)

Choice of R: Select such that V/R provides required bias current

Idss-Is it important to match these for the two (L and R) channels and if so how is this done?

Thanks for any help.

Ryan

[Dr.H]

Further thoughts and questions:

Seems that most DIYers use a cascoded JFET arrangement without any current setting R and simply use the JFET at its IDSS.

This seems to have the benefit of maintaining a high input impedance at the following stage (normally a buffer), which presumably would be compromised by the use of a R with the JFET.

However, it does seem that the use of an R would isolate the output of the amp from the high capacitance of the JFET, hence my question:

Anyone know of a low capacitance (10pf or lower) JFET with an Idss of 5mA or less?


For reference, the circuit I'm refering to is at

http://www5.head-fi.org/forums/showt...hlight=cascode

See the attachment in the post.

[jcx]

it seems cascode fet ccs loading has nearly a cult status among the headfiers

in http://www5.head-fi.org/forums/showt...639#post355639

i attempt to point out weaknesses in the the party line there and show some simple alternatives

a few references:

http://www.vishay.com/fets-small-signal/ssfanp

AN102 "JFET Biasing Techniques" Figures 7, 15, 16 and associated discussion should be required reading in addition to the obvious AN103 "The FET Constant-Current Source/Limiter"

[Dr.H]

Thanks, useful.