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

[nickb]

How about the asymmetric saturation of the inductor mentioned by RG?

I think it would be wise to avoid extrapolating too much from sample of one. As I understood it, just one of RG's pedal showed this effect. Also, I don't think it was quantified anywhere just how big the signal had to be before saturation became significant. It does, OTOH, give a good reason to put a boost pedal prior to the wah to get some of the extra harmonics.

[Helmholtz]

I think R.G.s results refer to one specific toroidal core Fasel inductor. Two of my modern Dunlop wahs have red Fasel (toroidal core) inductors. I replaced one of them with a home made pot core inductor and like it better. I don't understand the Fasel hype. I think it was invented by Dunlop marketing. Meanwhile their more expensive wahs sport a Halo type pot core inductor.
Definitely prefer my old 60s pot core wahs.
But maybe original Fasels (sometimes) were different.

Oscilloscope measurements of max. voltage-time area before distortion did not show noticeable asymmetry.

[Helmholtz]

On German Wikipedia I found the formula for the resonant impedance of a PRC: Zr = L / (R C), where R is the series resistance of the inductor.

The peak impedance of a PRC is real, i.e. resistive, because the reactive components cancel at the resonant frequency.
That allows to treat the resonant impedance and the 33k damping resistor R7 as two resistors wired in parallel.

Assuming an inductor series resistance of 100R, I get Zr = 500k for the treble peak and Zr = 20k for the lower or "bass" peak, where C is increased by a factor 25.
Wiring the 33k in parallel results in 31k and 12.5k respectively, meaning that the bass peak is much lower than the treble peak.

Now using a series resistance of only 20R gives Zr = 2.5M for the upper peak and 100k for the lower one without the 33k.
With the 33k in parallel results are 32.6k for the upper and 24.8k for the lower peak. Here the difference is only moderate.

The example shows that a higher inductor series resistance has a strong lowering effect on the "bass" peak, while the upper peak is not much affected and will always be around 30k.
The purpose of the 33k parallel resistor is to equalize peak heights, but that only works as long as the PRC impedance is much larger than 33k.

I have to add that the effective series resistance of the inductor will typically be larger than the DCR, as core losses add to the equivalent series resistance (ESR).

The formula for Zr also shows a dependency on the L/C ratio.
If e.g. a 400mH inductor is used with a C2 of 12nF, Zr will be 40% lower than with 500mH/10nF.
This results in the "bass" peak being lowered by 15% to 30% depending on L series resistance.


And now, if you'll excuse me, I must be on my way....(Hendrix "EXP").
Reason is my "Alter Ego" * urging me to play and listen.

* That's the other guy residing in me, an obsessive guitar player .

[Helmholtz]

It does, OTOH, give a good reason to put a boost pedal prior to the wah to get some of the extra harmonics.

One might think so.

But to me a wah sounds best when directly connected to the guitar. Overdrive or fuzz after the wah.
That's how Hendrix, Clapton and SRV used it.
Just use it with fresh strings. Testing wahs with worn strings can be frustrating.

[Helmholtz]

I did a few measurements to prove my inductor ESR theory of post '33.

First pic shows the lower peak with a number of different inductors in my project wah. The group of larger peaks consists of inductors with an L of 510mH to 630mH @ 1kHz and a DCR between 15R and 24R. The red curve belongs to a Dunlop yellow Fasel measuring 510mH/22R. The green curve shows a Roger Mayer Halo inductor having 350mH/53R.
Generally the peak height drops with lower L/DCR ratios as predicted. There's no indication of significant core losses adding to the ESR.




The second pic shows how wiring resistors in series with a Dunlop red Fasel inductors lowers the peak height:



Meanwhile I found that the "Variable Q" control of Dunlop 535Q wahs is a variable 1k resistor wired in series with the inductor. So it will essentially vary the height of the "bass" peak.

[Helmholtz]

An apparent oddity:

I was wondering about the influence of the impedance at the output of the emitter follower, which lies in series with the resonant cap, so I expected it to have a similar effect on "bass" peak height as the DCR of the inductor.
A rough calculation gave a value of 100R with the base grounded (high peak) and 170R with an hFe of 250, resp. 125R with an hFe of 670 for the low frequency peak.
(Emitter follower output impedance depends on collector current, hFE and base circuit impedance.)

But neither varying transistor hFE nor adding an external 100R series resistor with the cap had a significant effect on peak height or Q-factor.

I took me a while to remember that the Miller effect not only increases capacitance but lowers all impedance by the loop gain (+1).
So the effective series resistance of the resonant capacitance to ground will always be between 5R and 10R for the low peak.
And that explain the measuring results.

[vintagekiki]

Small contribution to the topic.

https://www.freestompboxes.org/viewtopic.php?t=15

https://patentimages.storage.googleapis.com/fa/0f/9e/6ad22f67158aae/US3749809.pdf

https://patentimages.storage.googleapis.com/d6/c5/42/6e6dba7d69a8f0/US3530224.pdf ¹⁾

[Mick Bailey]

I don't understand the Fasel hype. I think it was invented by Dunlop marketing.

Maybe it's one of those areas where because older wahs tend to use them quite a bit, they attract a mojo following for no other reason. I've had quite a few wahs with Fasel inductors that sound average, and one that I still have that's really nice. I think many people hang too much on the inductor being the magic ingredient, but I've played about with inductors endlessly and I've never had a wah pedal magically transform just because the inductor changed. One pedal that really sticks in my mind was a repair/mod to a 90s Dunlop wah that had just a basic ferrite inductor. These seem to be the ones that people swap out and discard because they have a poor reputation, but I left it anyhow. It still had a Clarostat 100K pot that was a little too noisy, so I switched it for a Hotpotz which completely spoiled the sound. Back to the Clarostat and I handed it back to the customer and it attracted quite a bit of attention on the local scene. I still get comments years later that "You're the guy that did that wah for Jim". It's nothing special - just a stock pedal with just a few component value changes - but for reasons I can't work out sounds exceptional.

[Helmholtz]

Some additional remarks.

Transistors:

As already mentioned the overall influence of transistor hFE is very small. Varying hFE between 250 and 670 changed peak height(s) and Q by little more than 5%.
And due to the clever bias arrangement, transistor idle currents don't shift much more than 5% either.

Some wah experimenters claim to hear differences between higher and lower hFE transistors. My measurements don't seem to support much audible difference.

My favorite wah, the 67/68 Vox V846, uses BC109B transistors.These have a typical hFE of about 300@2mA. At the lower collector currents in a wah (0.2..0.3mA) this will drop to around 240.

Contemporary Vox 847 and Dunlop GCB-95 wahs use MPSA18 transistors having a current gain above 500.


It might be interesting to know that the voltage gain at the peaks is between 7 and 8, which might cause (unwanted?) distortion in following equipment.
In my measurements I injected a 0.1Vrms signal to the wah input.

The total current consumption of the wah circuit is rather low at 0.5mA, meaning long battery life.


Buffers:

Some newer wahs have input and/or output buffers.
An input buffer makes the sound a little sharper - don't like it. It is sometimes used to avoid a true bypass switch, which in IMO is the better choice.
(My measured responses would correspond to the sound with an input buffer as I fed the wah from a low impedance signal generator.)
As the guitar's source impedance increases with frequency and adds to the series part (68k) of the input voltage divider, I expect the higher frequency peaks to be lower with a guitar directly connected. Thinking about modelling the loaded PU impedance in series with the generator for more realistic measurements/simulations.

An output buffer makes sense, when the wah is connected to a low input impedance device like a Fuzz Face.

[Helmholtz]

One pedal that really sticks in my mind was a repair/mod to a 90s Dunlop wah that had just a basic ferrite inductor. These seem to be the ones that people swap out and discard because they have a poor reputation, but I left it anyhow. It still had a Clarostat 100K pot that was a little too noisy, so I switched it for a Hotpotz which completely spoiled the sound. Back to the Clarostat and I handed it back to the customer and it attracted quite a bit of attention on the local scene. .

So you actually didn't change anything? Is that the message?

You probably know that some HotPotz-1s were made by Clarostat?

https://reverb.com/au/item/3287989-c...age-wah-pot-x1

[pdf64]

I'm inspired to get lose the input buffer on mine (and wire up a proper bypass switch)!

I seem to remember that some replacement pots are higher resistance; sorry if I missed it, but what effect does that have?

[Helmholtz]

I'm inspired to get lose the input buffer on mine (and wire up a proper bypass switch)!

I seem to remember that some replacement pots are higher resistance; sorry if I missed it, but what effect does that have?

See post #20.

Higher pot resistance like 470k might lower the middle range peaks, because of increased emitter follower output impedance.

The Roger Mayer board I'm using in my project wah has an input buffer. For my tests I disconnected the 68k resistor (R1) from the buffer and directly connected to the wah input.

[vintagekiki]

Small contribution to the topic.

http://thermionic-studios.com/wiki/index.php?title=Wah_Inductors

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

Wah Inductors.pdf

1) Anybody tried winding wah inductors?

https://archive.ampage.org/threads/4/pmgd/262457/Re_Anybody_tried_winding_wah_inductors-1.html

https://archive.ampage.org/threads/4/pmgd/262457/Re_Anybody_tried_winding_wah_inductors-2.html

https://archive.ampage.org/threads/4/pmgd/262457/Re_Anybody_tried_winding_wah_inductors-3.html

https://archive.ampage.org/threads/4/pmgd/262457/Re_Anybody_tried_winding_wah_inductors-4.html

2)
http://www.voxshowroom.com/ct/misc/wah/inductors/

[vintagekiki]

I remain in memory VOX by Sola Sound wahs, who worked somehow differently - more convincing than other wahs from that era.
It is probably a secret that they had a specific pot 70% of the walk was resistance 100k, the remaining 30% was 0 ohm.
And wah drive mechanism was different, so the change of tone was sudden, that's why they got the name crybaby.


https://www.diystompboxes.com/smfforum/index.php?topic=99248.0
About wah drive mechanism

https://www.freestompboxes.org/download/file.php?id=15699&sid=ed412fed2f129043d7edfd2643a6977f

[Helmholtz]

I remain in memory VOX by Sola Sound wahs, who worked somehow differently - more convincing than other wahs from that era.
It is probably a secret that they had a specific pot 70% of the walk was resistance 100k, the remaining 30% was 0 ohm.
And wah drive mechanism was different, so the change of tone was sudden, that's why they got the name crybaby.

I have 2 original Sola Sound wahs from the 70s. The mechanism only utilizes about 40% of the pot taper but requires very wide foot action.
This makes the use somewhat awkward.

Soundwise as well as regarding feel/control I much prefer the older Vox/Jen wahs.

I don't think Sola Sound wahs used the name Crybaby like some Vox/Jen wahs did.