A „Vox-Wah“ project, some circuit analysis and measuring results.

[Helmholtz]

A while ago I decided to build a Roger Mayer RM9090A wah board into a nice 60s Vox organ volume pedal.
I already have a collection of wahs, my favorite one being a '68 Vox V846 with a „halo“ inductor, which I use as a reference.
When comparing my new build to the old Vox wah, the results were rather disappointing: Almost no wah-effect on the bass strings and a somewhat dull sound in treble position.
I did some measurements showing a weaker low frequency peak and a somewhat lower frequency treble position peak, also being narrower than that of the reference.

That's when I realized that I needed a better understanding of the Vox circuit and the interactions of the components.

I had already read the original patent as well the excellent analyses by R.G. Keen and ElectroSmash:

http://www.muzique.com/schem/wah-patent.pdf

http://www.geofex.com/ "The Technology of Wah Pedals"

https://www.electrosmash.com/crybaby-gcb-95

https://www.electrosmash.com/vox-v847-analysis

(Please read! Especially the patent.)


But I felt that I needed to dig a little deeper.

So here's a brief summary of my findings:

The Vox wah circuit is a frequency variable (tunable) resonant (band-) filter followed by a gain stage.
The filter is wired as a voltage divider, the series part being the 68k input resistor and the shunt part is a parallel arrangement of an inductor and a „variable“ capacitance, a parallel resonant circuit (PRC).
The signal voltage across this PRC is fed to the input of the gain stage, a single transistor common emitter amplifier.
As the impedance of a PRC varies with frequency, so do input and output voltages of the amplifier.

The pedal's output is taken from the collector of the first transistor (gain stage).

Now what makes the PRC tunable, resp. the capacitance variable?

The trick employed is called the Miller effect.

So, before I continue (probably tomorrow) please read this: https://en.wikipedia.org/wiki/Miller_effect

Please understand that my motivation to continue will depend on the feedback.

[nosaj]

Well a lot of what is in there is way over my understanding. But one question regarding lead dress. Is the miller effect in play in amplifiers by Ken Fisher ..using lead dress to create hidden capacitances? Or is that an entirely different effect?

nosaj

Hopefully RG will pop in

[Helmholtz]

Well a lot of what is in there is way over my understanding. But one question regarding lead dress. Is the miller effect in play in amplifiers by Ken Fisher ..using lead dress to create hidden capacitances?

Not familiar with Ken Fisher amps.
I will gladly answer questions about Miller capacitance in amps in the amp theory&design section.

[Mick Bailey]

I've had similar issues with Boomerang wah builds where the sweep acts more like a tone control. I built two boards at the same time using identical inductors, the only difference between them being one had the same value but different brands of capacitor. Here's the sound of the 'good' one in use, the other still in my possession because I never got it to sound like a Boomerang wah. Sure, it wahs - but not the same sweep or tonality. I had the good one back and matched the transistors and checked the component values but it's still not there.

https://www.youtube.com/watch?v=WT8Ll-zAfTA

[mozz]

Just bought a empty wah shell so I 'll be following. I wound a few inductors and even bought some ferrite shells but other projects took over. One of these models had a polystyrene cap or am I thinking of another pedal?

[Helmholtz]

So here's the second part of my circuit description:

The Miller effect means that any capacitance (C) between output and input of an inverting amp acts as if there was a capacitance of C(Av +1) wired across the amp's input, in our case between ground and the base of the gain stage. This is an NFB effect.
Av means the voltage gain of the amp and the capacitance C is called Miller capacitance.

So to make the input capacitance variable, the gain needs to be variable.

It is obvious that the collector output cannot be used as it's fixed gain.
That's why there needs to be a second stage providing a separate output with a volume pot between the stages to vary the gain.
In the Vox circuit this is realized by an emitter follower, which also has the advantage of a low output impedance.

Measurements of several wahs showed that the max. gain between emitter follower output and the base of the input transistor is around 27.
This means that the resonant capacitance can be varied from 10nF (zero gain) to 280nF with a typical 10nF feedback cap.

As the resonant frequency varies (inversely) with the square root of the frequency, the peak frequency can be varied by a little more than a factor of 5.
With a typical 500mH inductor, this means a (theoretical) frequency range of 440Hz to 2.3kHz.

In reality the range is a little smaller, as the travel of the pot is not fully utilized. There is also some influence of the taper of the pot.

The actual value of the inductor can be calculated from the top frequency (pot at zero).


More to come about resonant impedance vs Q-factor, real world inductors, pot tapers, transistor gain, seemingly unnecessary parts and the effect of some component value changes on the measured frequency response - when I find the time.


Topic related questions welcome!

[Helmholtz]

Just bought a empty wah shell so I 'll be following. I wound a few inductors and even bought some ferrite shells but other projects took over. One of these models had a polystyrene cap or am I thinking of another pedal?

Thanks for your interest.

Yes, those 10nF Ducati caps in early Vox V846 wahs are supposedly polystyrene caps. But I want to concentrate on measurable effects.
So please don't expect me to comment on sonic advantages of polystyrene or "tropical fish" caps.
Apart from that I encourage everyone to do their own comparisons.

[vintagekiki]

Small contribution to the topic.



http://www.geofex.com/article_folders/wahpedl/wahped.htm

1)
https://delicious-audio.com/the-wah-pedal-inductor-the-secret-to-your-favorite-wah-tone/

https://www.sweetwater.com/feature/dunlop/wah-pedals/index.php

https://www.reddit.com/r/guitarpedals/comments/2oftdt/electrical_comparisons_of_different_crybaby_wah/
Electrical comparisons of different Crybaby wah inductors

https://www.diystompboxes.com/smfforum/index.php?topic=91111.0
Wah Inductors and magnetism

2)
https://www.marshallforum.com/threads/wah-inductors-pots-and-other-secret-sauce.77144/
Wah inductors, pots, and other secret sauce

https://www.youtube.com/watch?v=x1N12FArkF4
Wah Pedals: Does the inductor affect the tone? A comparison

[Helmholtz]

For in between, a few pics:

My project Wah on the right, left a 68-70 Vox V846 with "film can" inductor.


Top view, order reversed.


This is the simple jig I use to measure pot tapers. Regulated 10VDC connected to pot, output voltage taken every 30° of rotation. Good enough for my purposes.

[vintagekiki]

http://www.voxshowroom.com/ct/misc/wah_record/vox_wahwah.mp3

http://www.voxshowroom.com/ct/misc/wah_record/voxwahwah_record.jpg

[Helmholtz]

Here are the frequency responses of 2 original Vox V846 wahs and the project wah with the Roger Mayer RM9090A board at the extreme positions.

[Helmholtz]

In the following I will refer to the schematic and component designations as used in the ElectroSmash article here: https://www.electrosmash.com/vox-v847-analysis

Transistor biasing:

The Vox circuit uses an interesting and clever biasing arrangement which utilizes DC voltage feedback to stabilize the collector current(s) against transistor variations. In fact, doubling the current gain (hFE) of Q1 changes the collector current and potential by less than 5%. The effect on voltage gain is even less.

The bias for both transistors is derived from the collector potential of Q1.
For Q1 the bias network essentially consists of the voltage divider R6/R8. C3 (4.7µ) grounds the signal across R8. This is necessary to avoid undesirable signal NFB.
The DC voltage across R8 is then applied to the base of Q1 via the low DCR of the inductor in series with R2.
The bias for Q2 is provided via R5, which connects between the collector of Q1 and the base of Q2.

[nickb]

The 6dB drop in gain at the low end between the versions is striking. Were there any circuit or component differences between them? It would be instructive to disable the PRC in all of them and do the sweep again.

A problem with this type of wah is the uncertainly regarding the inductor specs and the cost thereof. One alternative would be to a voltage controlled filter. This could be done cheaply with an LM13700 OTA. The attraction is low cost, excellent tolerance from unit to unit and the ability to easily tweak the frequency range and sharpness (Q). Another advantage of the VCA approach is that it easily lends itself to triggered auto wah or can be driven by a low frequency oscillator.

[Helmholtz]

The 6dB drop in gain at the low end between the versions is striking.

Yes.

Were there any circuit or component differences between them?

Yes, I'll refer to that later.

[nickb]

Yes, I'll refer to that later.

Oh, you teaser...