[sdiy] FET buffer followed by Darlington buffer

[JH.]

http://www.jhaible.de/varislope_filter_phaser/varislope_filter_phaser_sch1.pdf

> What advantage does using the Darlington pair in addition have over just using the JFET?

FET gives high input impedance, BJT gives low output impedance.
Wouldn't exactly need a _darlington_, but as it's already there ...
Also: DC shift of n-channel JFET is partly compensated by DC shift of npn darlington.

But all this wasn't the reason for choosing that configuration at all. What happened was this:

I built the Varislope Filter/Phaser for the JH-3 Modular years ago (http://www.jhaible.de/tonline_stuff/hj1.jpg), and started with 
the simple darlington buffer configuration. This created ugly thumping at low cutoff frequencies (like Harry said), and it was 
terribly noisy, too. Then I made some changes to improove the noise, and I found that putting BF245's on the solder side was the 
easiest way to cure the thumping without cutting too many traces.
And the whole thing was still a tad noisy, but otherwise did sound awesome (to my ears), espcially as a phaser. Kind of gritty with 
high signal levels, and very flexible because of the ability to spread out the pole/zero-pairs instead of having them fixed.

When I decided to make that PCB project last year, I carefully cloned my own circuit, down to the every detail like coupling RC time 
constants, in order not to loose that "vintage" sound I had partially created by accident. :)

JH.