DIY Equivalent To Envelope Pedal

[robotboy]

Hey guys. I recently learned about a really cool pedal for which I haven't found any DIY equivalent. It's called envelope (made by toadworks), and basically, it's an effects loop that is activated once your guitar signal reaches a certain threshold. So, you can for instance run a fuzz through the effects loop and when you play softly there won't be any fuzz. Once you start strumming harder, the fuzz will slowly swell in. You can also set it to work in reverse, so the softer you play, the more "wet" the signal is. It seems like a similar effect could be achieved by mixing something similar to the paralooper with a vactrol. I have a bunch of VTL5C2 vactrols left over from my uglyface build that I'd love to use for something like this. Unfortunately, I don't have to know-how to design a circuit of this complexity yet. So, does anybody know of a similar DIY project or how hard it would be to achieve something like this.

http://www.virtualtoad.com/el/index.html

[Mark Hammer]

The Toadworks Enveloope has been discussed here recently, although I forget which thread.  I've only got a few minutes here so I'll leave the search up to you.

In principle, it is obviously feasible since thereis a commercial product that does it.  The trick would be to develop a way to slow things down.  Normally, envelope signals come and go just as quickly as your note does, and that's where the sticky point lies.

Imagine the following.  You pick a note hard and send it to the envelope follower.  The envelope follower has its threshold set such that it produces a kind of voltage-controlled pan, if you will, that pans the input signal to the fuzz circuit instead of straight through to some mixer stage.

So far so good.  You will note, however, that a fuzzed note will appear to sustain longer than its unfuzzed counterpart.  That is, the contrast between the highest amplitude point in a clean picked note, and what follows immediately after, is quite substantial, so the envelope signal goes from full tilt to diddley-squat faster than you would like it to.  Ideally, you would like for the panning aspect of this pick-controlled loop selector/panner to remain on the fuzz for a while after the envelope of the note you feed it has died down.  And THAT remains the technical goal of a well-designed loop selector of the type you want to emulate.

In principle, what you'll want is:

a) A decent clean input buffer that can serve as active splitter and send the signal in three directions.

b) An envelope follower stage.  Full-wave rectified is preferred by half-wave, well-filtered can probably suffice.

c) A mixer stage (inverting op-amp preferred) where the relative gain applied to the inputs is controlled by the resistance of the LDRs at its input.

d) A control voltage inversion stage.

So, the signal comes in and goes three places, one of them the envelope follower, one of them the input of the mixer, and the other a loop send.  The envelope folower is followed by an inversion stage that feeds the LED of the optoisolator for the "clean' half of the pedal.  The other EF output goes to the LED of the loop part of the circuit.  Without playing anything, the input resistance of the clean channel stays low (LED on) so its gain stays reasonably high (i.e., you hear it).  In contrast, the loop input of the mixer stage keeps a high resistance (i.e., gain is low).  Pluck a note and the LED states (and LDR resistances) change, except they do so smoothly, rather than instantaneously.  That's it.

Gotta run.

[StephenGiles]

Well actually I had a circuit published in ETI over 20 years ago that did something similar. I think if you played softly you got tremolo and if you played hard you got distortion.
Stephen

[robotboy]

The Toadworks Enveloope has been discussed here recently, although I forget which thread.  I've only got a few minutes here so I'll leave the search up to you.

In principle, it is obviously feasible since thereis a commercial product that does it.  The trick would be to develop a way to slow things down.  Normally, envelope signals come and go just as quickly as your note does, and that's where the sticky point lies.

Imagine the following.  You pick a note hard and send it to the envelope follower.  The envelope follower has its threshold set such that it produces a kind of voltage-controlled pan, if you will, that pans the input signal to the fuzz circuit instead of straight through to some mixer stage.

So far so good.  You will note, however, that a fuzzed note will appear to sustain longer than its unfuzzed counterpart.  That is, the contrast between the highest amplitude point in a clean picked note, and what follows immediately after, is quite substantial, so the envelope signal goes from full tilt to diddley-squat faster than you would like it to.  Ideally, you would like for the panning aspect of this pick-controlled loop selector/panner to remain on the fuzz for a while after the envelope of the note you feed it has died down.  And THAT remains the technical goal of a well-designed loop selector of the type you want to emulate.

In principle, what you'll want is:

a) A decent clean input buffer that can serve as active splitter and send the signal in three directions.

b) An envelope follower stage.  Full-wave rectified is preferred by half-wave, well-filtered can probably suffice.

c) A mixer stage (inverting op-amp preferred) where the relative gain applied to the inputs is controlled by the resistance of the LDRs at its input.

d) A control voltage inversion stage.

So, the signal comes in and goes three places, one of them the envelope follower, one of them the input of the mixer, and the other a loop send.  The envelope folower is followed by an inversion stage that feeds the LED of the optoisolator for the "clean' half of the pedal.  The other EF output goes to the LED of the loop part of the circuit.  Without playing anything, the input resistance of the clean channel stays low (LED on) so its gain stays reasonably high (i.e., you hear it).  In contrast, the loop input of the mixer stage keeps a high resistance (i.e., gain is low).  Pluck a note and the LED states (and LDR resistances) change, except they do so smoothly, rather than instantaneously.  That's it.

Gotta run.

Thanks for the ultra-informative reply. Unfortunately, my electronic skills are at currently very basic, and I'm still struggling to understand voltage dividers, biasing and other very basic concepts. I do appreciate the effort of your explanation though, and hopefully I'll come back oneday and understand what you just wrote. I might actually buy the real toadworks envelope because they seem like a cool company.

[troubledtom]

just got mine tn the mail. this is not a fly by night device. it's complex and compact.
  i think Z. would even scratch his head.
    more news latter.
           the co. does good work and it's in the USA.
i've done no test yet , but will in 2-3 daze.
                 out,
                      - tt

[Mark Hammer]

Steve,

Could we see the circuit?  There might be some nice springboards to other applications in there from the sounds of it.

[puretube]

food for thought from R.G.:
http://www.geofex.com/Article_Folders/VCA%20Applications.pdf
http://www.geofex.com/Article_Folders/panner.pdf
mix these with compressor-basics...