QuoteFor instance, imagine you had a pedal with three paths: dry, wet-1 (flange), and wet-2 (phase). The dry is fed to the standard mixer stage, and so is the wet......BUT....you get to blend combinations

not quite "effect", but 3 different pre-amps:
http://www.pure-tube-technology.de/Tube-Meister%20MM1-1.jpg
(I sat for weeks over that one...)

esp. there is no "low output impedance" left and right of the tone-pot...

I see a "damping" (or say: shunting") of the feedback-path varying between the cw & the ccw section of the tone-pot by means of the vol-pot towards ground.

To check out, if some (pos.) feedback is involved here,
one should disconnect the 2 stages, buffer each, and re-combine them additively...
(don`t wanna disturb anyone here, but Mark: You know what I mean...).

I do believe you are correct Brother Barmentloo.  There IS feedback.  Even without much knowledge, a close listening just now through something a little bigger than a 3" speaker reveals the following:

1) Overtones are *really* easy to get on the ow strings at high gain (yeah!)

2) The blend pot crackles where the others don't.  (in this case crackle maybe not so OK)

3) When I rotate the tone/blend control all the way over to the first stage, the signal becomes a lot cleaner (even though it is still distorted).  At higher gains, the difference between the noise and what seems to be intermodulation or some sort of motorboating, coming from the second stage vs the first stage is bigger.

4) The gain in the second stage is fixed.  The overall signal level hitting the LEDs is fundamentally dependent on the amount of gain in the first stage.  When you turn down the gain all the way on the first stage so that it is simply a gain of 3 (which is actually pretty clean, still, and one of the reasons why the blend thing is a nice feature) there is still a noticeable distortion from the second stage.  Not searing, mind you, but more than you'd expect with such a high clipping threshold and a gain of "only" (:lol:)around 100.  I wouldn't doubt that a little feedback to the input helps out.

I am assuming that one issue here is a lack of a DC blocking cap between the diodes and the 6k8 resistor that taps the signal at that point.  That will still permit some feedback, but at least what I gather is a DC offset finding its way from the output of stage 2 back to the input willl be eliminated.  I put in a 1uf NP cap and it doesn't seem quite so bad (crackle seems to be gone) but the motorboating is still there when the gain is increased substantially.  If anyone has some ideas about how to eliminate it, other than "layout", I'd be grateful.

CN,
The noise, from a hiss standpoint, it not that bad at all when you figure how much louder than bypass the signal can be.

err, Brother Hammer: my statement:"seeing a feedback-path" was not at all meant as criticism.
In fact the (safe) zone between (controlled) pos. feedback and oscillation (eeeeek), is a nice field to graze...  
(what about converting the 2nd stage into a "clipped Wah" ?)

Wanted to know (ever nosey...) if that`s intentional, accidential, or un-noticed...

B.T.W.: my dealing with "metal" musicly, nowadays is reduced to singing "Whole Lotta Love" in our band.

Somewhere around here was a thread talking about using the stereo MXR micro amp pc from GGG or Tonepad, not sure which.  HMMMMMMM

What is the center freq and deapth of the notch filter.  That could come in handy in a number of places?  

Brother Barmentloo,

Be as nosey as you like....my long range plans are to stick with my day job and lose as little money as possible on my hobbies.  No industrial secrets here!

Absolutely NOTHING intentional about introducing feedback.  I am enough of a simpleton to have believed the signal "knows" where I want it to go, and never ever thought about the bidirectionality of the blend circuit.  On the other hand, I'm glad you pointed it out since it is a principle that would have never occurred to me.  Perhaps a fixed resistor in parallel with the output pot (to operate as additional damping/attenuation of the feedback through the blend/tone pot) will be a useful circuit change/option...or not.

As for the motorboating, that seems to only occur with the panning all the way to the second clipping stage at highest gain settings.  I imagine shielding of the leads to the blend control may help to alleviate this.  I have it build into a plastic box at the moment, with a sheet of copper shim along the underside of the top of the box to serve as a ground plane.  That may not be enough for the particulars of this unit/design.  A DC blocking cap doesn't seem to be a solution to the motorboating, unless it is a small enough value to roll off lows below 200 or so, which is not really a solution since the lows are kind of nice.

I *will* get these problems licked though.....with perseverence and maybe with a little help from "nosey" people. :wink:

High gainers sharing boxes can get noisy.
 I was able tofind ' noisy wires, by just moving them around in the box..or just putting my finger on them sometimes increased or decreased in any crease changed the noise.
 IME ground planes do 'something' but I don't know exactly 'what'...lol...and I'm not all that too inclined to worry about figureing out why ground planes do what they do...or how to place them or where they should be for max noise supression.
 I'ts a wood box with an Obsidian and a LPB in it. the Obsidian has the grounded copper sheet 'unde-round' the sides [a short/wide U shape at cross view]. With the Obsidians gain knob up [past like ~7]and the LPB on, a whirring motor noise appears.

against motorboating, usually the best remedy is to de-couple the "B+"s of the 2 involved stages.
Since a dual-opamp seems to be used, this is not possible here.
You could try `n find how the circuit behaves with 2 single opamps.
Or have a guess `n check out if 2 individual Vbias dividers for the
2 stages might help....

I can try that.
I will say that even with gain maxed, there is no evidence at all of motorboating when I pan the blend control all the way to first-stage output.  All motorboating comes from the second stage with high gain.  On the other hand, the Vref for stage 2 appears to be coming from the output of stage 1 (i.e., there is no resistor to Vref at the input of stage 2), so maybe some sort of decoupling IS called for.

yes, you need to insert a cap "between the 2 10k Rs", just "left of where the 10k and the 6k8 join", to isolate the 2nd (new) Vref....