Dual Filter 'Mutron III'

[strungout]

Hey guys, been a while, I hope everyone is doing well and building!

I'll spare you the details, but I got back into building pedals a month or two ago.

I wanted to build Craig Anderton's Bi-Filter Follower, but I don't have the right optos for it and couldn't make them either. It was the one with Mark's mods. Now, instead, I decided to build a Mutron III work-alike (MBP's Naughty Fish). It uses VTL5C3s, which I have. Then I thought, what if I applied all of Mark's mods from the Bi-Filter Follower to a dual Mutron? Plus, I could get a lowpass sweep along with a highpass on the other filter, or bandpass, or whichever mix. Well, it needs some fiddling still, but it works!

I need some input from you guys.

Here's what I have right now:





Some things:
- The 'sounds like a 4-stage phaser' thing. I'm not sure I hear it, but should that part work at all in this circuit? (It's the upper part of the circuit with one inverting opamp and a non-inverted path.) The Bi-Filter uses two bandpass filters doesn't it? Like the Dr. Quack. So, it should at least work in bandpass mode...

- I'm not sure what value caps I should use to cover the most ground (or get the most dramatic effect) from one filter to the other. Marked Cx. I'm using 1nF at the moment. Let's call the upper filter Hi and the bottom one Low. Haven't covered the second switched pair yet. I guess I could just use one pair per filter and save two switches.

- I have an extra 1/2 opamp. Anything suggestions on what I can use it for?

- Mixing resistors (Rx)... well I guess I can just experiment with those until I find a good balance...

What do you guys think? Any improvements?

[Mark Hammer]

Daddy like! 

Not the exact same thing, but the complexity and flexibility reminds me of Marjan Urekar's "Fuzzy Balls" project, that is an EHX Bass Balls taken to the max.  http://www.synthdiy.com/files/2006/fuzzyballs1.pdf

The autoswept phaser-like sound would really only occur with the two filter circuits each set for bandpass, and staggered apart.  When the two filters in a Bi-Filter Follower are close together, it sounds like wah.  Pushed further apart, that's when it begins to sound more phaser-like.  That may or may not be attainable with your Range controls, in conjunction with the Hi/Lo switches on each filter section.

[strungout]

Hey Mark, glad you like it!

I've made some changes:




- Added a 5k Attack pot and reduced the series 330r to 100r.

- Tagged a 1M pot in series with the 47k draining the 4u7 cap for Decay control. It's what the Fuzzy Balls has and it works great.

- I changed the location of the Gain (Sensitivity) pot to right after the 2u2 that comes off the first opamp. I wanted to avoid raising the gain for the filters at the same time as for the envelope signal, which created some (not bad sounding) distortion. I don't know if I like it better thought...

- I changed the mixer/output stage to a non-inverting one. Unless I counted wrong, it would come out of the circuit inverted, as it was.

- For the filters, I found a 470p and 10n spread was the min/max for really hearing what's going on. To my taste anyway. Below 470p is really thin and kind of 'meh' to me. Didn't add much. Above 10n, the filter doesn't sound like it's making a full, satisfying sweep. The high strings just don't respond to normal picking strength either.

- Still trying to figure out what I can do with the extra opamp. Maybe I can boost the signal going through the envelope (without boosting the filters') and have an active 'more' control instead of a passive 'less'. Then I could add some caps above 10n and get some decent filter movement. As it stands, the Sensitivity knob lets the max amount of signal through. Might as well not be there!

Anyway, back to fiddling.

[Mark Hammer]

Glad you're finding tweaks that work to your satisfaction.

As enthusiastic as I initially was about being able to add variable attack time to autowahs, practically speaking, I find it tends not to add much.  It certainly does in the MXR Envelope Filter, but not much in this context.  Just as importantly, adding series resistance to increase Attack time eats up drive current in the envelope signal, which requires turning up the Sensitivity.  That is, altering Attack time to a degree that you can actually hear, is a two-knob operation, where adjusting Decay is a one-knob adjustment that you can hear, independent of Sensitivity setting.  As well, the rise time of a photocell is often not as easily shaped by an Attack control as we might like.

[strungout]

Hey.

So I used the last dual 1/2 to put a non-inverting gain stage before the 2u2 in the envelope follower. I'm surprised how well it works! Indeed a two-knob job.
I'm also getting more movement on the higher strings, which didn't give me much sweep travel before. At around 2/3, turn works well, so I still have some room to adjust if needed. As you said, I have to up the attack to compensate for the gain, but I'm good with that.

Now I just need my parts to build this thing. I'll probably try and record some samples, then. My breadboard is a rat's nest right now. Everytime I move something, something else gets nudged and touches something it shouldn't...

To be continued:

[strungout]

I've played with this more and found that actually, all THREE controls (Gain, Attack, Decay) of the EF interact. If I raise the Gain, I have to adjust de Attack and Decay values (or just one of them can be enough). If I raise the Attack, I have to 'push back' the Decay further. Lowering the Decay means lowering the Attack to maintain a discernable 'peak'...

Imagine I am moving the peak around while adjusting its 'peaking' length, by adjusting its upward and downward slopes. Everytime I raise the Attack I have to compensate by making the Decay longer so that there is a pre-defined amount of 'peaking'. Otherwise, there is no action in the filters at all... As if I'm canceling the peak, in some way...

Maybe this should be obvious, but... Anyone wanna lay some wisdom on me? Mark? Anyone?

I'll probably understand something. Or fart.

1 in 2.

[Mark Hammer]

The Attack and Decay resistances form a voltage divider.  If one reduces Decay and tries to get a longer Attack, you basically reduce the "volume" of the envelope voltage charging up the 4u7 cap.  Naturally, this requires one to turn the Gain up to compensate.  This is one of the reasons why I find that using variable Decay, on its own, results in less problematic functioning of the unit.  Not just this circuit, but almost any that  use a similar method of rectifying the signal.

[strungout]

Ah, I get it, thanks Mark!

For me, the more variety of sounds I can get out of an effect unit, the better. Even if it's more or less subtle... or if there's interaction between controls. It changes my way of playing and what I'll be inspired to play. If I'm using a moderate gain pedal, for example, I'll be playing rock stuff. Up the gain to high and I'm sliding into metal.

They say you have to "Play the effect". I guess I also let the effect 'play me'!

Update on the dual filters wah: It now has an EF for each filter. I have trouble hearing both sweeps, or I just find it kinda boring, so another EF will allow me to stagger the sweeps. Basically, I have two circuit boards and a third to mix them. I also added another output on each filter that will let me send each of the signals to a different amp. Because, why not? Haven't made the schematic yet... We'll see if that's a good idea. I'm just havin' fun!

(I remember doing something like this in my Nurse Quacky experiments. This project is that, but on steroids.  )