Would this biasing arrangement work?

[fryingpan]

As per the picture.

[antonis]

Make R1 + R2 a 5k trimpot and pray(*)..

(*) Always advisable when dealing with JFETs..

P.S.
I presume you know that JFETs bootstraping isn't so effective compared to BJTs..
(less than about 15M input impedance in parallel with  a 5 to 10 pF capacitance for your circuit..)

[fryingpan]

Otherwise, there's this circuit (that I adapted from an R.G. Keen buffer):



Advantages: significantly cleaner, less sensitive to JFETs, basically no current draw. Disadvantages: less drive capability as is (needs a buffer at the output to be able to drive less than 30kohms or so), negligible distortion (which may "sound good").

[antonis]

The later is totally different design than the former..

It's a CFP with negative feedback amount (gain) set by 1 + (R8 + U1)/R5..
Here gain raises output impedance (hence the limited drive capability) so it should need some compromise..

P.S.
May I ask what you intend to design/build..??

[fryingpan]

The later is totally different design than the former..

It's a CFP with negative feedback amount (gain) set by 1 + (R8 + U1)/R5..
Here gain raises output impedance (hence the limited drive capability) so it should need some compromise..

P.S.
May I ask what you intend to design/build..??

A simple booster. I know I could do it with an opamp (even with a lowly 4558, but the 4580 is cheap too, at least with bootstrapping to keep the input impedance at least over 1Mohm). With much better PSRR, by the way. Still, I want to do it with discrete devices. Where's the fun?

(This was borne out of a personal need. I play bass, and the cheap boosters you can buy are all designed for guitar, and they all cut the bass frequencies. I need a booster because I have a few basses that have a series/parallel switch, with a significant change in volume when switching between the two).

[antonis]

R.G.'s circuit can have a max gain of +12dB which might be considered OK for a simple booster..

You can of course add a directly coupled BJT(*) output buffer (you might need to trim Q1 Collector voltage) of 1 to 2 mA working current..

(*) Strongly recommened due to its output impedance much lower than JFET's one..

P.S.
In case you're discrete design fun lover (that makes two of us) I'm willing to suggest some - more complicated - strictly bi-polar configurations..

e.g. https://www.diystompboxes.com/smfforum/index.php?topic=130344

Combine Q1/Q2 & Q11 stages with altering R5/R6 according to your gain taste..
Also, R41, C21, R43 & C23 can be ommited, although not recommended to do so..

[R.G.]

R.G.'s circuit can have a max gain of +12dB which might be considered OK for a simple booster..

As shown, yes. The complementary feedback pair can be tinkered. There's enough raw gain there to get more effective AC gain if you want it. The tricky stuff in the complementary feedback is that it is a little constrained by the DC bias conditions. The JFET really needs its source at a certain voltage above ground, and the bipolar really needs its collector at about half of the DC supply. The resistors to get that constrain the gain.
But you can play gain games by bypassing some or all of the resistance on the JFET source to ground with a big enough capacitor.
There's a lot of transconductance with the JFET and bipolar in series. You can get quite a bit of voltage gain with even a modest collector resistor on the bipolar.

[fryingpan]

+12dB (4 times) is plenty for a clean booster. Also because I want the booster to be able to take any kind of guitar or bass signal. With very hot pickups (even passive pickups), you might even get 2Vpp. Both designs accomodate this. The first design, if run clean, adds about 2% THD (strong 2nd harmonic, slight 3rd and 4th), the second almost none. Both are almost inaudible for guitar and bass. (If anything, 2% THD adds some slight """""warmth""""").

The nice thing about R.G.'s design is that it is really minimalistic and works well (technically, better). It can also be made to distort a bit, by fiddling with resistor values.

(Yes, I do know that bootstrapping a JFET doesn't work as well as with opamps or certain bipolar configurations. A Darlington or compound pair with 1:1 feedback should probably be able to make a 500Mohm input impedance out of a 100k resistor or so).

[antonis]

The nice thing about R.G.'s design is that it is really minimalistic and works well (technically, better).

As fas as I'm informed, nobody(*) ever blamed R.G. for flawed or deficient design..

(*) OK, I know a forum member but let it be..

[fryingpan]

Anyway, I built it. (Well, "build" is a bit excessive...). I added a switch to reduce the drain resistor to either 1.2k or slightly less than 1k. Gain goes a bit down but it's still well over unity (like, it goes down to 10dB, fine for a booster), but you do get some slight and strong distortion respectively. No diodes at all (of you ignore the intrinsic diodes within a JFET or a BJT)! It's also mostly low order harmonics (very soft clipping). It could be a good "block" for more complex circuits.

[antonis]

Drain resistor value sets the operating current for J1..
(set by voltage drop across Q1 Emitter resistor + V_E-B..)

By lowering its value 20 times you're getting too close to J1 I_DSS (maybe you even hit it) so distortion is unavoidable, despite the NFB good intentions..