Hi there,
recently, Joe Davisson's wonderful EZ-250 came up (
https://www.diystompboxes.com/smfforum/index.php?topic=122849.msg1169521#msg1169521), which I really like. But as with any low to moderate gain distortion, I asked myself: How would that sound with more gain? So I decided to dual-bootstrap it. Regular old impedance bootstrap on the first transistor (Cornish style), gain bootstrap on the second. I originally intended to do the latter like it is done in the Jordan Bosstone (direct coupled bootstrap) but that messed with the bias too much with the additional direct coupled source follower up front. So I went with an AC coupled bootstrap. Anyway, here it is, the "250 biturbo":
Q1, Q2 = 2N5088; Q3 = 2N4403; any low noise, high gain NPN should be fine for Q1 and Q2; the Q3 position is not picky either, 2N2907A, 2N3906 and a bunch of other should work fine. Depending on the transistors, R10 may need to be adjusted to get around half supply voltage at the collector of Q2.
I really like the sound of this thing. Smooth, thick distortion, transparent enough to be the "default sound" in many styles of music, yet dripping with character. As presented in the schematic, I would consider it a "desert rock" or "desert blues" type of distortion but by changing the clipping diodes, you can change the character quite drastically. Noise level is surprisingly good. I was unable to make out any noise that was not completely swamped by the remaining hum of a humbucker (so, pretty quiet then). What I like about the high input impedance is that this thing reacts extremely well to the guitar volume pot. No dulling or other changes in the sound as you turn it down. Just the desired effect on the distortion level. If you work with the Vol pot of your guitar a lot, try this on for size.
The interesting "trick" in the design is that the clipping diodes are placed between the gain transistor Q2 and the bootstrapping transistor Q3. This way, they do not only clip the signal directly but also limit the extend of the bootstrapping action as soon as the clip, which reduces the gain available from Q2. This keeps the transistors themselves from ever running into the rails and clipping harshly. If you want that lovely transistor clipping, remove the clipping diode arrangement (D2, D3, C7) or move it between C9 and the volume pot (C7 can be omitted in that case). Gain is controlled somewhat unconventionally by adjusting how much bootstrapping takes place. The rev-log 50k pot provides an OK range here, but C20k or C25k may give a better sweep over most of the range with only marginally higher minimum gain. I just happened to only have a C50k at hand.
I went with the back-to-back 2V7 Zeners for clipping diodes after some experimentation. They give a very nice soft-knee distortion, meaning you get strongly diminished higher-order harmonics compared to hard-knee clippers. This is even more pronounced for the Zeners than for back-to-back mosfets (BS170) with the gates and drains tied together. You can, and should, of course use whatever clipping arrangement you like. With the regular old anti-parallel 1N4148 pair, you get *a lot* of distortion, possibly too much. red and blue LEDs sound OK to my ears. Asymmetric (red LED + 1N4001) was pretty good too. Do experiment here but I definitely encourage to give the 2V7 Zener pair a try, I absolutely love it. Goes very gently into clipping, none of that annoying "brittle fizziness around the edges" that plagues many a diode clipper. It does not give crazy amounts of distortion though, due to the high clipping threshold (2.7V+0.7V). At low gain this also makes a very nice, slightly dirty boost with a lot of volume on tap (I will build me one like that without a foot switch as an always-on, first-in-chain dirt booster / sound maker / moderate driver). Remember though that at the position where the diodes are in the schematic, you need C7 for DC-decoupling (aka AC-coupling) in order to keep the diodes from upsetting the transistor bias. For that very special transistor distortion, leave out the diodes completely (C7 is not needed in that case either, obviously).
If you feel the need for a tone control, replace R12 and C9 with Mark Hammers Stupidly Wonderful Tone Control. I found it unnecessary with the Zener clippers but with harsher clipping diodes it will be useful.
The anti-pop resistor R1 should be left out if you build this without a switching circuit, like I will.
A word of caution: This was intended mostly as a proof of concept, regard it as work in progress. Some values were chosen because a particular resistor or cap happened to be lying around near the breadboard, so those still need optimizing. Updates will, hopefully, follow shortly.
And finally: Yes, all of this could be done even easier, even cheaper, even less noisy and just all-around better with a single opamp. If I were in the business of producing for-sale pedals, I would use opamps too. This is more of a design study to see if I could also do this discreetly and how to do it. It is meant to showcase some useful techniques that are fun for the tinkering DIYist. Plus, designing with bare-bones transistors is somehow more fun to me than with opamps. Maybe *because* it is often less predictable and sometimes more difficult to get what you want.
Cheers,
Andy
