I've been working on that dirt channel schematic - this is a work in progress so bear with me - this is what I can now re-draw those tranny stages as (I've kept all the component references as the original Bandit PCB)
- Re-drawn Bandit dirt stage with track cuts and joins and component swaps done
Hopefully you can see where I'm going with this if you compare it to the Wampler Plexi Drive schematic I showed in the last post.
So by swapping the transistors for J201s, making some track cuts, adding some link wire, changing one transistor and one capacitor for resistors (not a mod I would normally recommend
) I can change the Bandit dirt channel into something very similar to a Plexi Drive which I hope will sound a whole lot better.
I've still got to sort out component values, and I've not yet used that four pole switch S2 to best advantage (so far it is only going to switch in an extra cap in one position).
While we're talking about switch S2, what I've done is parallel a cap in one position - this will add more bass frequencies into the distortion stage (Q5+Q6). I've got further ideas for this switch though.
For component values - here goes:
R61, R55, and R40 are just keeping the tranny inputs at ground => 1M
Let's start off with the JFET biasing resistors having the same values as in the Plexi Drive, namely R51, R45, R57 = 1k and R26, R29, and R28 = 15k
I'll just say a bit about what those resistors do. In each case the resistors convert the JFETs drain current into a voltage. In the case of the resistors in the source leg (to 0V) they provide a positive voltage on the source pin relative to the gate pin, so the gate sees it as a negative voltage. This means for a little drain current there is a small negative voltage on the gate, and as the drain current gets bigger there is a bigger negative voltage on the gate. Since these are N-channel JFETs (to get technical) more negative gate = less drain current, so there is negative feedback - which means the gain stage sort of limits its own gain (that's how most amplifiers work folks).
If I chose JFETs with the right Idss and Vgs off then those 1k source resistors will work just fine.
For the 15k drain resistors (which go to Vdd) these give a voltage at the gain stage output again which is proportional to the drain current. In the original Plexi Drive the supply was 9v so they were probably chosen to set the voltage at the JFET drains, with no input signal, to be around 4.5v (which is what you'd probably want). In my case that 4.5v is the voltage below my Vdd which means I'll sit at +28V (the Bandit's supply for the preamp) minus 4.5v = 23.5v. That means when my signal swings up it will hit the rail pretty dang quick. When it swings down though it has plenty of volts before it goes so low the JFET turns off. All of that means r
eally asymmetrical clipping which ain't no bad thing. I'm not intending the Q8 stage to be clipping either.
One further complication. I put 2x JFETs in parallel for Q5, Q6. If they're matched, that will double the drain current of one, however there's plenty voltage headroom down the way (towards 0v) so I'm going to leave R29 at 15k for now. I'm going to breadboard all this anyway before committing the changes to the PCB.
Another thing - these drain resistors affect the
tone. That's because each resistor forms an RC filter with the gain stage output cap. So fiddling with the bias resistance will also affects the EQ => affects tone.
By the way (I'm really digressing now
) that's why you shouldn't just stick any old JFET in a distortion pedal circuit and then tweak the drain resistor for the drain voltage you want. It works Ok, but as you change the resistor value you change the EQ of that stage and you'll get different tone - maybe not what you really wanted if you;re tryng to do a pedal copy'
Digression over
For all the caps. I'll start with the values in the Plexi Drive and tweak for tone. So that means:
C27 = 22nF, C25 = 1uF, C30 = 2n2
I'll leave the stuff around Q11 as it is on the Bandit board since its just a unity gain (that means gain of 1) follower to give a high impedance into the next bit of circuit.
C19, C24, and C66 are really going to affect tone so I'll just play with them on the breadboard, starting with C19 = 470pF, C24 = 2n2, and C66 = 180pF (the Bandit original value).
C19 and C24 as said earlier will affect how much bass feeds the clipping stage, while C66 is the treble bleed cap around the gain pot so you get more treble when the gain is down low.
I need to re-draw this think again at some point as the schematic is getting really messy now!!