Envelope Controlled Filter with control mods

[Fancy Lime]

Hi everyone,

I've been toying with different Envelope Controlled Filters for a bit. The original idea was to have "really flexible autowah". So in addition to an envelope mode I thought I'd integrate an LFO mode and a rocker pedal for manual mode. Wait, if we have an envelope follower anyway, why not control the LFO speed with the envelope and use the pedal to set the range? And of course a mode where the pedal controls the LFO speed and the LFO controls some parameters of the envelope follower, which in turn controls the filter sweep. And, and, and... Long story short within 5 minutes the thing I was planning to build was the size of a small car just to accommodate the half million knobs, switches, rocker pedals, motion sensors, radio antennas and flux compensators. So MAYBE this might have to be spread out over several different projects.

So I started with a simple old fashioned autowah. I quickly gave up the original plan to have low pass, high pass, band pass and notch mode after figuring out that resonant low pass is the only mode I would realistically ever be using. Notch mode would probably sound good in a foot or LFO controlled filter, but for envelope control I found it a bit boring. Next I gave upon the "down sweep" idea for pretty much the same reason. So whats left is a resonant low pass in "classic 70s autowah" mode. But I never liked any of the ones I have ever played. There was always something off. So I wanted maximum control: Sensitivity, Attack, Release, Range, Resonance. And since I wanted a lot of resonance, a Volume pot also seemed advisable. I went with the filter stage of the DOD FX-25 and tried different envelope followers. What I ended up with is a bit of a cross between the DOD FX-25 envelope section and the one of the Phuncgnosis with some extra bits and pieces stolen together from several threads in this forum, mostly here:
http://www.diystompboxes.com/smfforum/index.php?topic=77370.0

Anyway, here's the schematic:


It's really just a slightly improved DOD FX-25 with all the obvious mods and additions, better envelope controls and some tweaking of values. Here are the most important changes with comments. Hope this helps if someone wants to embark on a similar endeavor.

1: Added a second buffer on the input and a "send" and "return" path. One can either use this loop to add distortion before the filter but after the envelope follower or use the return as guitar in and the input jack as an aux in for a control signal (drum machine or something).

2: Increased the Resonance pot for more quack. This made it necessary to put diodes in parallel. They don't change the sound much (which I thought they would, actually) but they kill the resonant ringing. Who new?

3: Decreased the volume by half before the second input buffer and raised it back up on the output stage for more headroom. Was necessary because mod 2 caused clipping on very loud notes. The LEDs before the first buffer also help kill input clipping.

4: Added pre-conditioning to the envelope section (C13 and D5) to smooth the rippling a bit and make the Attack and Release controls work properly together.

5: Two on/off/on switches add caps for charging with the envelope. This arrangement allows 9 total cap values between 1µ and 33µ. Out of these only two combinations (23µ and 25µ2) are so close that one of them is redundant, meaning we have 8 useful settings. The low settings (1µ, 3µ2, 5µ7) are not classic autowah sounds but Moog-like synth sounds. Extremely cool. The medium values are Mutron and FX-25 territory, the very highest settings are a bit slow for my liking on bass (where autowah makes me staccato finger funk, if I want it or not) but might suit guitar sounds better. All in all an extremely effective control over the reaction time of the envelope follower.

6: Replaced the fixed 1M resistor that sets the floor of the sweep range with a pot. I can highly recommend this. It allows to take the beginning of the sweep just above the fundamentals of the stuff you are playing. This eliminated the speaker shattering FWOOMP! that you otherwise get on low notes when the Resonance is high and the Release is fast. Makes the extremely resonant settings much more usable.

7: The Attack and Release knobs usually suffer from too much interactivity with each other as well as with the Range and Sensitivity controls. So Getting those sorted out was a bit of a challenge. Two things turned out to be crucial here: Arranging the Attack and release knobs in a way that Transmogrifox suggested in a previous thread:
http://www.diystompboxes.com/smfforum/index.php?topic=118579.20
Thanks for that tip! When done like this, one can actually have independent Attack and Release controls that don't mess with the range too much either. The other important thing was to add a red LED (D8, could probably be replaced with 2 normal silicones) before the release pot. This is because the OTA control gates draw the envelope down to 2 diode drops above V-. When the Release pot goes to V- without the LED, then the Release becomes highly interactive with the Range control at low Release times. With the LED it is referenced to roughly the same level as the OTA gates (actually slightly above because it is D8+D9 above V-), making Release and Range independent. Now the Attack control is still interactive with the Sensitivity in that it not only controls the attack time but also the ceiling of the filter sweep (which is otherwise controlled by the Sensitivity). However, I don't actually find this very bothersome. The Attack can be used as a fine adjust for the filter ceiling in the "wah modes". In the "synth" modes, Attack and (to a lesser extent) Release shape the sound rather than the filter sweep. Well, they still control the filter sweep, but the sweep is so fast, it follows each string vibration. So Attack and Release shape the waveform. Did I mention I LOVE the synth sounds? Anyone up for a Herbie Han%^&* and the Headhunters cover band? Listen to the intro of Chameleon and you know what I'm talking about.

Despite all the controls, this thing does not have a million different sounds. But the ones it does have can be shaped exactly as I want them, which is the one thing I always missed with other autowahs. It just makes jamming to a funky beat so damn addictive. I hope to get this of the breadboard and into a box soon, this is so much fun to play! Itching to throw a bass octaver and overdrive in the loop. If only I wouldn't hate layouting (outlaying? How's that even called, dear native speakers?) so much...

Cheers,
Andy

[KarenColumbo]

Absolutely awesome!

[Fancy Lime]

Quick question:

Can someone explain to me how the LEDs D1 and D2 keep Q1 from clipping? The idea comes from Runoffgroove, who use it a lot. It does work but for the life of me, I can't figure out how. D2 pulls the gate down to one LED drop above ground but that can hardly be all, can it?. With D2 out, the gate sits at 3V, with D2 in at 1.4V (D1 in or out makes no difference to the voltages). D2 is normally on the edge of conducting and lights up a tiny little bit on the loudest peaks. For D1 to ever open, the gate would have to dive to -1.4V, which seems like an awfully big swing for a guitar-level signal.

Thanks,
Andy

[Mark Hammer]

They don't really prevent clipping.  They prevent it from clipping more.  Think of it as a very primitive limiter that avoids Q1 getting overwhelmed.

Given the high-ish (1.5V, usually) forward voltage, they'll really only stop the odd peak, but then we don't know how ugly the rest of the circuit could sound without that once of prevention.

[Fancy Lime]

True, it only kicks in at the very highest peaks and then the diodes clip audibly. So not ideal but significantly less awful than without. In either case I really have to punch the stings unreasonably hard with both pickups in series to get it to clip, but that happens sometimes, doesn't it? 

[Kipper4]

Did you try any other input buffers Mr lime?

Just curious.

[Fancy Lime]

Originally I had the op amp buffer IC1a as the input buffer, with the side chain branching of after that. The additional JFET buffer came on board when I decided I wanted a loop. Other then that, no other buffers were tested. I basically went with my default. The JFET buffer has always worked quite nicely for me, although I suspect a BJT in this position to perform pretty much identically. Not sure about input overload behavior, though. Do you have any experience with that?

Cheers,
Andy

[Kipper4]

I did use something like that with send return in this project.

http://www.diystompboxes.com/smfforum/index.php?topic=117301.msg1088961#msg1088961

[ElectricDruid]

It's true that I haven't seen those clipping diodes on the feedback path in (m)any effects circuits, but it's common enough on synth filters. It'd be pretty much standard for any state variable filter, all the way from the Oberheim SEM onwards, but you also see it on some OTA-based cascaded-stages 4-pole filters, like the Roland SH101.

Tom

[Fancy Lime]

Hey there,

little update: I finally figured out what the LEDs D1 and D2 do. D2 sits just on the brink of opening when there is no signal and pulls the gate bias voltage of Q1 down. The higher the positive swing on the input, the more the diode opens. On negative swings it closes. Thereby this configuration utilizes the relatively soft knee of the LED to move the bias voltage around by some amplitude dependent fraction of the phase inverted input signal. Neat! The resulting effect is clipping with a soft knee, so it goes from very soft clipping at low amplitudes to hard clipping at high amplitudes. I should be exploiting this more deliberately for overdrives... D1 still seems to server purely decorative purposes.

Cheers,
Andy

[n4m3less]

Hi everyone,

I've been toying with different Envelope Controlled Filters for a bit. The original idea was to have "really flexible autowah". So in addition to an envelope mode I thought I'd integrate an LFO mode and a rocker pedal for manual mode. Wait, if we have an envelope follower anyway, why not control the LFO speed with the envelope and use the pedal to set the range? And of course a mode where the pedal controls the LFO speed and the LFO controls some parameters of the envelope follower, which in turn controls the filter sweep. And, and, and... Long story short within 5 minutes the thing I was planning to build was the size of a small car just to accommodate the half million knobs, switches, rocker pedals, motion sensors, radio antennas and flux compensators. So MAYBE this might have to be spread out over several different projects.

So I started with a simple old fashioned autowah. I quickly gave up the original plan to have low pass, high pass, band pass and notch mode after figuring out that resonant low pass is the only mode I would realistically ever be using. Notch mode would probably sound good in a foot or LFO controlled filter, but for envelope control I found it a bit boring. Next I gave upon the "down sweep" idea for pretty much the same reason. So whats left is a resonant low pass in "classic 70s autowah" mode. But I never liked any of the ones I have ever played. There was always something off. So I wanted maximum control: Sensitivity, Attack, Release, Range, Resonance. And since I wanted a lot of resonance, a Volume pot also seemed advisable. I went with the filter stage of the DOD FX-25 and tried different envelope followers. What I ended up with is a bit of a cross between the DOD FX-25 envelope section and the one of the Phuncgnosis with some extra bits and pieces stolen together from several threads in this forum, mostly here:
http://www.diystompboxes.com/smfforum/index.php?topic=77370.0

Anyway, here's the schematic:


It's really just a slightly improved DOD FX-25 with all the obvious mods and additions, better envelope controls and some tweaking of values. Here are the most important changes with comments. Hope this helps if someone wants to embark on a similar endeavor.

1: Added a second buffer on the input and a "send" and "return" path. One can either use this loop to add distortion before the filter but after the envelope follower or use the return as guitar in and the input jack as an aux in for a control signal (drum machine or something).

2: Increased the Resonance pot for more quack. This made it necessary to put diodes in parallel. They don't change the sound much (which I thought they would, actually) but they kill the resonant ringing. Who new?

3: Decreased the volume by half before the second input buffer and raised it back up on the output stage for more headroom. Was necessary because mod 2 caused clipping on very loud notes. The LEDs before the first buffer also help kill input clipping.

4: Added pre-conditioning to the envelope section (C13 and D5) to smooth the rippling a bit and make the Attack and Release controls work properly together.

5: Two on/off/on switches add caps for charging with the envelope. This arrangement allows 9 total cap values between 1µ and 33µ. Out of these only two combinations (23µ and 25µ2) are so close that one of them is redundant, meaning we have 8 useful settings. The low settings (1µ, 3µ2, 5µ7) are not classic autowah sounds but Moog-like synth sounds. Extremely cool. The medium values are Mutron and FX-25 territory, the very highest settings are a bit slow for my liking on bass (where autowah makes me staccato finger funk, if I want it or not) but might suit guitar sounds better. All in all an extremely effective control over the reaction time of the envelope follower.

6: Replaced the fixed 1M resistor that sets the floor of the sweep range with a pot. I can highly recommend this. It allows to take the beginning of the sweep just above the fundamentals of the stuff you are playing. This eliminated the speaker shattering FWOOMP! that you otherwise get on low notes when the Resonance is high and the Release is fast. Makes the extremely resonant settings much more usable.

7: The Attack and Release knobs usually suffer from too much interactivity with each other as well as with the Range and Sensitivity controls. So Getting those sorted out was a bit of a challenge. Two things turned out to be crucial here: Arranging the Attack and release knobs in a way that Transmogrifox suggested in a previous thread:
http://www.diystompboxes.com/smfforum/index.php?topic=118579.20
Thanks for that tip! When done like this, one can actually have independent Attack and Release controls that don't mess with the range too much either. The other important thing was to add a red LED (D8, could probably be replaced with 2 normal silicones) before the release pot. This is because the OTA control gates draw the envelope down to 2 diode drops above V-. When the Release pot goes to V- without the LED, then the Release becomes highly interactive with the Range control at low Release times. With the LED it is referenced to roughly the same level as the OTA gates (actually slightly above because it is D8+D9 above V-), making Release and Range independent. Now the Attack control is still interactive with the Sensitivity in that it not only controls the attack time but also the ceiling of the filter sweep (which is otherwise controlled by the Sensitivity). However, I don't actually find this very bothersome. The Attack can be used as a fine adjust for the filter ceiling in the "wah modes". In the "synth" modes, Attack and (to a lesser extent) Release shape the sound rather than the filter sweep. Well, they still control the filter sweep, but the sweep is so fast, it follows each string vibration. So Attack and Release shape the waveform. Did I mention I LOVE the synth sounds? Anyone up for a Herbie Han%^&* and the Headhunters cover band? Listen to the intro of Chameleon and you know what I'm talking about.

Despite all the controls, this thing does not have a million different sounds. But the ones it does have can be shaped exactly as I want them, which is the one thing I always missed with other autowahs. It just makes jamming to a funky beat so damn addictive. I hope to get this of the breadboard and into a box soon, this is so much fun to play! Itching to throw a bass octaver and overdrive in the loop. If only I wouldn't hate layouting (outlaying? How's that even called, dear native speakers?) so much...

Cheers,
Andy

Hi and thank you for your contribution! the images of the schematic are no longer available! could you please upload the images of your schematic?

[Ben N]

I second that request to freshen the linky.

[suryabeep]

Me three!

[KarenColumbo]

Could be this one: