I did this using my custom-built pedal prototyping test chamber (a lunch-box) which can be seen here (testing a different circuit): I used a 28v PSU (as per the Bandit pre-amp board) and selected some J201s from my "Fisherman's Bait" box (you can just see the corner in the above picture).
I use this box to keep pre-measured JFETs - they're sorted into compartments depending on Idss and Vgs(off), so I can see what I've got.
Anyway I chose the following to try (component refs are from the schematic in my last post):
- Q8 - J201 Idss = 0.77mA Vgs(off) = 1.0v
Q5 and Q6 - J201s Idss = 0.67mA, Vgs(off) = 0.95v (matched pair)
Q10 - J201 Idss = 0.70mA Vgs(off) = 1.1v
I figured using slightly lower gain JFETs for Q5 and Q6 would be good since I am paralleling 2 together (explained in my earlier post).
Also, I didn't add Q11 onward in this breadboard try-out, I just took the output to my practice amp from R(C26) and R40.
So this is what I found from playing around with the breadboarded circuit:
1. The pair of JFETs Q5/Q6 gave way too much gain for the classic Plexi sound I was after, this gain stage went straight into clipping with no crunch.
2. The gain with just one of Q5 or Q6 is better but needs a bit of a boost - I found a 47uF cap across R45 (to reduce negative feedback at audio frequencies) to be a good solution, but I think it needs to be switchable as I also like the low-gain crunch without it
3. The clipped sound is too fizzy and needs a touch more LP filtering after the Q5/Q6 gain stage
4. C66 = 180pF seems about right for the treble bleed cap (hard to tell until I use a full speaker cab)
5. The clipping sounds better with more bass so I changed my C19 to 4n7, and C24 stays as 470pF (not sure if this looses too much bass, again need to try with a speaker cab)
6. JFET drain voltages are around the predicted values (28v-5v ish for each)
So what I'm going to do on the Bandit pre-amp board mods is:
o Fit only Q6 with a J201 and short out base to emitter on Q5's PCB position. This is to address #1.
o Add a 47nF cap across R45 on a switch (as per #2) - I'll use a spare position of the 4 pole switch S2 which I'm re-wiring anyway. Need to think about how S2 will work
o Address #6 with an increase of C30 to 3n3 (from 2n2) - that gives a 3dB cutoff freq of R28+R(new) and C30 => 30k and 3n3 => 1.6kHz Looks a bit too low, I'll try it and see
o Fit 4n7 for C19 (the pre-stage2 bass boost)
Before I modify the pre-amp PCB for this circuit, I need to take a quick look at the post-dirt gain EQ (there's a bunch of Rs and Cs) and decide what I'm going to do with it.
I'll hopefully do that in my next post, and also sort out that new S2 switch wiring in the process.