clean blend that doubles as optional buffered bypass?

hi all,

i'm interested in the idea of adding a clean blend to an existing distortion effect, and i'm wondering if it's possible to also use it as a switchable buffered bypass option.

with the help of this very useful tutorial by @jesuscrisp:


i feel pretty certain that i'm going to go with a dual op-amp buffer and active blend setup, though i'm not sure yet if i'll be doing one that goes from 100% dry to 100% wet version, or just the purely additive one. i do have some questions for the forum though:
1) are there particular advantages and disadvantages to placing the input buffer before or after the split? should i buffer the input right off the bat and then split passively into clean and distorted paths, or split first and then have the clean path alone have a unity gain op amp buffer? the first stage in the distortion circuit is a non-inverting op amp boost of +10.4dB with a 15hz high pass, which should serve as the buffer for that path.

2) the way i envision doing the buffered bypass -- and if there's a better way of doing this please let me know -- is to have the 3PDT stomp switch still switch between on and bypassed, but a second toggle switch changes whether the bypass path is true bypass, or send the signal back into the effect, and then out the end of the clean blend path before the clean/distorted mixing section. the question for me is whether it's possible to get an isolated clean signal to the output through this method without the distorted signal "leaking" backwards through the mixer and into the clean signal. i've attached some crude concept diagrams of the switching scheme that will hopefully make my idea clear.
 

Attachments

  • Screenshot_20241228-224357~3.png
    Screenshot_20241228-224357~3.png
    176.5 KB · Views: 8
  • Screenshot_20241228-224357~4.png
    Screenshot_20241228-224357~4.png
    189.9 KB · Views: 8
  • Screenshot_20241228-224357~5.png
    Screenshot_20241228-224357~5.png
    193.7 KB · Views: 9
  • Screenshot_20241228-224357~6.png
    Screenshot_20241228-224357~6.png
    202.4 KB · Views: 8
  • Screenshot_20241228-224357~7.png
    Screenshot_20241228-224357~7.png
    270.9 KB · Views: 8
The "ON, TBP" and "BP, TBP" diagrams are the same.
The "ON, BBP" and "BP, BBP" diagrams are the same, save the additional text regarding leakage.
THUNDERAXE BYPASS.png
What's being bypassed? 🤷‍♂️


[EDIT] 💡Ahhh, nevermind. I finally clicked on one of the screenshots and saw the full pic.
I would say that in the last diagram, there's no way around having the buffered bypass with the clean&distortion separate — you have to have some way of separating the two if you want just clean on buffered-bypass.

The following was written before the lightbulb came on, but all still holds true...



Please post a schematic, whether complete or just the dirt-circuit.

I'm not seeing what the advantage is of sending the "buffered" bypass back into the circuit's "clean" path.

Why wouldn't you just have a bypass that either sends the signal through the dirt-circuit (with added clean-blend) ie engaged,
or bypassed one of two ways:
1) Buffered,
2) Unbuffered.

What could be cleaner than completely bypassed (whether buffered/unbuffered)?


I'm working on a boost with buffer, that can be true-bypassed or bypassed while retaining the buffer in the signal path.
Aion's Cygnus (Cornish G2) is the closest I've found so far to that bypass topology, yet uses a much more complicated switching method than what I'm intending to use on my boost-buffer build.


AION CYGNUS BYPASS.png



Maybe the Cygnus' bypass has something to it that may in turn help you.
 
i'll look into the cygnus switching and see if i can wrap my head around it, but in the meantime:

if i was really committed to the switchable true/buffered bypass, obviously the easiest thing would be to just give it its own switchable buffer op amp dedicated to the bypass path. but honestly i don't care enough about that feature to do so, i just think that if the clean buffer can double as a bypass buffer, it would be a cool bonus feature to throw in there. so my idea was, when the pedal is in buffered bypass state, to route the signal through the clean path buffer and then out before the clean signal gets mixed in with the distortion (the green/blue switch).

however, since the signal is still also going in parallel through the distortion path, and then downstream getting mixed together with the clean signal, the question then is, will the signal from the distorted path "leak" back through the mixer into the clean path when it's in that state? are you saying it's unavoidable the way i have it set up?
 
Last edited:
okay, i traced the paths, it's pretty much the same idea as mine except with two differences:

1) the switch between buffered and true bypass is a 4PDT, and the other two poles switch in the two resistors at the bottom when it's in buffered bypass mode. i don't really understand what those do and why they're necessary, any help would be appreciated.

2) there isn't a clean blend path for the distorted signal to potentially "leak" back through, so that isn't a concern here.
 

Attachments

  • Screenshot_20250114-091558~2.png
    Screenshot_20250114-091558~2.png
    107.7 KB · Views: 7
  • Screenshot_20250114-091558~3.png
    Screenshot_20250114-091558~3.png
    107.6 KB · Views: 7
  • Screenshot_20250114-091558~4.png
    Screenshot_20250114-091558~4.png
    206.9 KB · Views: 7
  • Screenshot_20250114-091558~6.png
    Screenshot_20250114-091558~6.png
    108.6 KB · Views: 7
Wow, great work.

I was trying to wrap my head round the extra couple switching-bits — being for the resistors all makes sense now, thanks for working all that out, your latest screenshots really helped me better wrap my brain around the Cygnus' switching.

R38 and R39 are setting the output impedance of the buffer when it is used in BP-BBP mode.
Electrosmash explains that resistor pair in the analysis of the Microamp.
The analysis says they also limit the output current and protect the op-amp that precedes them.





If you use some isolating resistors on the outputs of both the CLEAN and DIRT, then you shouldn't get bleed-through. (I think/hope — depends on what the rest of the circuit looks like / is doing).
Here's an example of the isolating resistors, as at the end of this modded ROG SplitnBlend (100Ω to 1k for R1 and R2 is suggested as being sufficient), from this thread:

ktsme6h.jpg





I don't know if there's unavoidable leakage or not, I'd need to look at the full schematic of your project,
and even then, nothing beats breadboarding it and finding out definitively.
 
the pedal i'm looking to do this to is the sour grape (keeley rotten apple) but i don't think that actually matters so much as the way i'm planning on wrapping a clean blend around it, so my plan was to have an op amp buffer up front:

1000009772.png
though, as mentioned in the first post i don't know if it's better to put it right up front, and then split into the clean and dirt paths, or split first and then put the buffer only in the clean path.

and then in back, an active op amp mixer, either R.G. keen's panner that is equal volume 100% clean to 100% wet:

1000009782.png

or else, if i decide i don't ever need to turn down the distorted signal relative to the clean, i might go with a purely additive mixer at the end:

1000009582.png

in both cases i'm pretty sure it's only the ratio of the resistors that matters, so if it would help, you can add one or even two zeros to their values.
 
Should the clean tapped off U1A come after C4 (and possibly R12)? 🤷‍♂️
I do see that U2A goes straight into a pot, too, though.

Might need a cap to block any DC trying to sneak through U1A and U2A, but I don't really know.


Another consideration... Overall the circuit(s) are non-inverting — until you get to U2B. Is there a reason you're inverting the signals there?


Looks cool so far. I've got the Sour Grape PCB as well; I think I picked it up on sale but only saw demos of the Rotten Apple afterwards that failed to impress me. The clean blend could change that.

If your clean-blend/buffer idea works out well for you, then that very well could re-insert my SourGrape PCB back into the build queue — so you might say I have a somewhat invested interest in this working well for you. 😉
 
for me the clean blend and optional buffered bypass are bonus features that are nice to have but not crucial to me enjoying the pedal. what i liked best about the rotten apple/sour grape was the clarity, i could play complex chords through it at maximum gain and they stayed fuzzy without getting muddy. but there were a couple of things i disliked:
1) the tone section -- i generally dislike big muff tone controls with their inherent mid scoop, and i find every AMZ-style mid-adjust control to be pretty ineffective. this one was no exception. and besides which if you're going to have a second tone control, whether it be a knob or a switch (and i at first turned it into a knob which i preferred to the switch, but still, didn't love it) then you may as well have bass and treble controls and be able to control the mids through them -- want a mid scoop? turn bass and treble up. flat? keep them fairly flat. mid hump? turn bass and treble down and turn up the volume. i spent a LONG time playing around with various james and baxandall tone controls until i came across the neve 1073 shelving EQ, which allows for second-order shelves, and that let me dial in exactly the curves that sounded best.
2) the relative lack of low end compared to the original op amp big muff. i wanted to bring some low end back without sacrificing the clarity that i loved so much. i was at first planning on doing this by just setting the tone section to boost the bass a little at the default setting but then adding a clipping cap actually delivered exactly what i was looking for -- allowing a decent amount of low end to bypass the clipping diodes, so the low end stays loud but not distorted.
3) the sweep of the gain knob -- with a B100K pot it spends like 95% of the sweep doing overdrive and light distortion, and only turns to fuzz in the last 5% of the sweep. switching it to C100K is a massive improvement but even it could use a little more refinement. a 5C taper (50% of the sweep is 5% resistance) would probably be ideal. tayda doesn't carry those so i'm thinking of using a C250K with a 180K fixed resistor across it, which should turn it into a 5C100K pot.

then the clean blend is nice for dialing a bit of your dynamics back in or using it as a bass distortion.
 
Last edited:
Should the clean tapped off U1A come after C4 (and possibly R12)? 🤷‍♂️
I do see that U2A goes straight into a pot, too, though.

Might need a cap to block any DC trying to sneak through U1A and U2A, but I don't really know.
my conclusion (and the breadboard agrees) was that i didn't need caps after U1A or U2A because the circuit continues to be biased to +4.5V up until the output. grounding the two pots to VREF rather than GND maintains that bias, otherwise i would had to either put a cap between each pot and GND (two extra caps), or on either side of each pot (four extra caps) to isolate them from the +4.5V bias. putting C4 between U1A and the CLEAN pot wouldn't cause any harm but it wouldn't really do any good either since the other side of the cap is also biased to +4.5V from U2B.

as for R12 that's to put a little series resistance between U1A and the output jack when it's in buffered bypass mode, just so the op amp isn't connected directly to the output jack.
 
Great list of mods for me to try! Thanks!
(Still so much for me to learn!)

I'd be using mine primarily on bass, so those mods are right on target for me.

I'm smacking my head over the clipping-cap... duh, why didn't I think of that!?!


Awesome! You've renewed my excitement for this circuit. Merci beaucoup.
 
I'm smacking my head over the clipping-cap... duh, why didn't I think of that!?!
it didn't occur to me either until i saw that the keeley moon pedal (the successor to the rotten apple) has one. as soon as i tried it out it was just like... oh there it is, everything i've been trying to achieve through other more complicated means...
 
the potential downside to the clipping cap is that without the diodes to keep the level in check, at a certain gain level the low end will start clipping your op amp. this is mitigated by the cap right before the GAIN/BUZZ control, which forms a variable high-pass filter that really only comes into effect at the highest gain levels. it's a delicate balancing act but one worth playing around with.

these were the values that worked best for me, but you might want to try playing with the value of the clipping cap -- lower values will raise the cutoff frequency of the diode bypass. the green russian and triangle big muffs which are famous for their fat low end use 47nF clipping caps, which is an order of magnitude smaller than what i ended up choosing. i just found that going any smaller than 470nF brought out too much of the muddy 100-200Hz region, but YMMV.

the other thing to play around with would be the cap before the GAIN/BUZZ pot. in that case larger values mean more low end at higher gain settings, and i had settled on 1.5uF just to let a little more low end through at higher gain, before i tried the clipping cap and found that it did the same thing but slightly better. if you're looking for a big fat fuzzy fuzz, i liked values as high as 3.3uF for this one, but in the end concluded that it took away from the clarity that i liked so much about the pedal in the first place and went back to the original 1uF.
 
Yeah, the Russian 47n was where my mind went when you mentioned the clipping-cap — and that's when I started kicking myself. Seems so obvious once pointed out, to extrapolate that transistor Muff concept to the op-amp. Makes me think, now, I should try adding a clipping cap to some other circuits, like so many YATS for instance.

re pre-GAIN/BUZZ cap...
Largest box-film I've got are 2µ2, so I'd try to stick to that — even though WIMA has some 4µ7. While I want bass-freqs to shine, you've put me on a quest for clarity with my bass plugged into this circ. Really looking forward to tweaking this.
 
i used electros when trying out different caps and they all worked just fine, and the original op amp big muff used a 4.7uF electro. any reason you'd want to limit yourself to box films?
 
No concrete reason, I just try to keep the signal-path as clean and robust as possible.

I once read about box-film superiority in terms of noise-level, reliability and life-span — never heard any difference myself.
(I'm just an impressionable noobie-amateur)

A pristine path is great for audiophile-gear, EQs and compressors, but not really necessary for a dirt circuit. 😸
 
Back
Top