Crybaby Replace inductor with Gyrator -- works? Yes

[Transmogrifox]

I had been wondering how successful I could be replacing the grounded inductor in the Crybaby with a gyrator.  Some initial Spice simulations showed that a textbook gyrator could not be used to get sufficient resonance for anything more than a dull "bwah bwah" sound.

This was confirmed by my searching, and behold, I found I was not the first to think this a cool experiment:
http://www.diystompboxes.com/smfforum/index.php?topic=22757.0;wap2

Reading that post I saw a link to a John Hollis circuit making an attempt at this.  By reports of the first link, this worked but was noisey:
http://www.hollis.co.uk/john/phantomwah.jpg

Well, I didn't let that stop me.  Two things I determined:
A)  The crybaby stock circuit is not sacred.  I can change components to interact with the gyrator.
B)  A textbook gyrator has its limitations with respect to getting a high-Q LC resonance, particularly with BJTs used as emitter followers (a constraint I arbitrarily put on myself to make this fun)

I realized I could modify the standard gyrator circuit to make a little higher feedback gain and virtually eliminate the feed-forward effect present in a BJT follower:

Gyrator


It really truly works.  If you try this, please note that the component "R2" is extremely sensitive to component variation.  It may be of value to build it with a 1k resistor in series with a 500 ohm trim pot, and tune by ear to taste.   It's a very narrow margin between "not resonant enough" and unstable.

Next part is to harmonize the crybaby with this circuit to make the interaction sound good.  I quickly found it was hard to make a drop-in replacement inductor like this just because this circuit emulates an inductor in series with a 1.5k resistor, significantly lowering the Q, so resonance must be adjusted elsewhere in the circuit.  Give here, take there, I finally got a contour I like with Spice simulation, so part B was:
Plug a wav file into Spice and listen to the Sim <--this is a really cool feature of LTSpice I discovered.  Love it

Well, of course, there's nothing like the real thing.  After I built the first prototype, these circuits went through a lot of iterative Spice-breadboard-Spice-breadboard cycles until I got something sounding good -- but it was noisey.

I had a chat with one of the lead hardware engineers at work and he gave me some good pointers regarding BJT noise, so a few component tweaks, and the noise is not much more than a stock Crybaby -- the last fortress overtaken

So here are the mods I came to for the Crybaby circuit itself:
Crybaby Mods


This doesn't sound like a normal Crybaby.  The main difference is the signal output location.  Notice I took it at a different place.  This was partly intentional to get a little more low frequency content so it isn't as thin sounding.  I'm not totally sold on this, but this tap location prevents having to provide an extra resistor divider for bias to the emitter follower.  In either case, the output buffer is highly recommended because if the next pedal/amp/circuit after the Crybaby loads it down too much, this circuit will go unstable.  Unfortunately by the nature of the balancing act on the gyrator positive feedback this circuit is easily pushed into instability.  With the output buffered, this removes unexpected impedances from the closed-loop system and at least keeps it from going unstable when plugged into the wrong type of load.  

If I do anything to further emulate the stock crybaby low end attenuation it will by in the form of changing the output capacitor after the buffer.  I was also experimenting with a "contour" stage:


I present the contour stage for the concept.  I have only simulated it and listened to a wav file processed in simulation.  It basically does what I would want it to do.  Might build it as a separate device to into the signal chain separately.  I think it could use a pot to adjust the hard-wired low-end attenuation, and one for frequency sweep.  It's just a Sallen-Key high pass.

I probably will have to set up a sound cloud account or something if people want to hear audio samples.  For now maybe it is just fun to consider that it is possible to replace the expensive inductor in a Crybaby with a few discrete components.  

As for really improving the sound of a wah pedal, I'm not sure I would recommend this.  It sounds better than the stock GCB95 I started with, but I think I could have had as much improvement without removing the inductor.  The problems with the stock GCB95 can be solved with the plethora of mods suggested across the internet.

I just wanted to tie my hands with some constraints and find out if it can be done effectively.  Here I present one way the inductor can be replaced with cheap components....maybe would be of more value to a commercial entity trying to save a few bucks on parts.

By the way, there is nothing magical about my use of pnp transistors (as you typically would see an emitter follower using an npn).  I just happen to have about 100 2N5087's laying around and not many 2N5089's, so I just used what was convenient.

[Digital Larry]

I'm impressed with your analytical skill and creative impulses.  I imagine the possibility of small PCB modules allowing different frequency ranges and whatnot via component tweaks.  Or a stereo wah.  Or a dual wah with different peaks.  All those kinda things.  Woo hoo!

[Transmogrifox]

... I imagine the possibility ... All those kinda things.  Woo hoo!

That's the spirit! 

This gyrator circuit and output buffer all fit on a piece of perf board the size of the inductor footprint if components are tightly packed and keystoned.

With surface mount components it would comfortably fit into the PCB real-estate occupied by the inductor.  In either case it all fits very comfortably into a crybaby shell.

[joegagan]

wow! great work. i need to read up on gyrators in general.

[Galego]

I'm impressed with your analytical skill and creative impulses.  I imagine the possibility of small PCB modules allowing different frequency ranges and whatnot via component tweaks.  Or a stereo wah.  Or a dual wah with different peaks.  All those kinda things.  Woo hoo!

A dual wah with different peaks could act like a formant filter, no? Kinda like this: http://img.photobucket.com/albums/v317/StephenGiles/EHtalkingpedal.gif

[midwayfair]

Great work! I'd be exceptionally interested in making clips.

Keep in mind that, even if you're not "improving" the original sound, a couple transistors, caps, and resistors are a heck of a lot more cost-effective than an inductor.

[Transmogrifox]

here are a couple demos:
A bit through a distortion/amp sim
https://soundcloud.com/transmogrifox/gyrator-wah-distorted-demo

A cheesey elevator "cool jazz" attempt (badly)
https://soundcloud.com/transmogrifox/gyrator-crybaby-demo

[WhiskeyMadeMeDoIt]

That sounds pretty dang good! Good job man.

[joegagan]

really a great sounding wah! congrats, this is an exciting development. kudos.

[Digital Larry]

A dual wah with different peaks could act like a formant filter, no? Kinda like this: http://img.photobucket.com/albums/v317/StephenGiles/EHtalkingpedal.gif

I think so.  I think if you ran them in series like that one that one of them (the higher one) might better be a low pass, so that you're not trying to boost with one where the other one is cutting.