How to best add a volume pot on the output of an Ibanez WH10 V2?

[Josh_bb]

Hi, another day another wah...

I have an Ibanez WH10 reissue (v2) that i am modding.

they lose ALOT of top end when off so I guess the solution for that has to be true bypass, (my original 1990 version does not seem to suffer so much)

Also, when the wah depth control is on full you get quite a lot of boost, too much really, I would like to add a volume control so the output volume is tweakable.

Would a pot work instead of the output resistor? say middle lug to out lug 1 to earth (or should that be V-Bias?) and lug 3 to where the output resistor was. 

Is that too simple? will it mess with output impedance and screw things up for the next pedal in the chain? Should the volume be before the output buffer instead? Not sure how best to do that. 

Any tips would be greatly appreciated.

cheers

I have attached a schematic of the WH10 V2

[antonis]

An easy way is to replace R21 with a 100k pot..
(lug 3 to C15, wiper to R22 & lug 1 to GND - C15 is large enough for HP filtering even at low volume settings where OUT LPF shouldn't face any practical issue ..)

Output impedance SHOULD be upset (the upper pot part resistance should be set in series with Q4 1/g_m where lower pot part resistance should be set in parallel with whatever input impedance is plugged in, dominating it..)

On the other hand, placing a Volume pot just before Q4 should result in Gate bias upset..

Strong coffee is urgently needed..

[duck_arse]

perhaps a variable R11, to reduce the gain of the stage feeding the wet to the out, without upsetting the "dry" level, what with the fet bypassing and all.

[.... antonis and his coffee ......]

[antonis]

Hmmmm...

It should work fine as long as U1A output (as it is) isn't hitting the supply rails (no distortion at all)

Otherwise, lowering gain shouldn't lower Volume (in signal voltage form)

[..... stephen and his beer .....] 

[Mark Hammer]

Superficially, replacing R21 with a 100k pot seems reasonable.  But I'd be concerned about its interaction with R22/C20.  At first glance, I'd worry that anything under max volume would roll off highs.  As shown, R22/C20 doesn't roll off until well above hearing range.  Perhaps if you simply eliminated C20, the problem would be solved, and you could use a pot in place of R21.

[Josh_bb]

Superficially, replacing R21 with a 100k pot seems reasonable.  But I'd be concerned about its interaction with R22/C20.  At first glance, I'd worry that anything under max volume would roll off highs.  As shown, R22/C20 doesn't roll off until well above hearing range.  Perhaps if you simply eliminated C20, the problem would be solved, and you could use a pot in place of R21.

hi
I already pulled c19 c20 and c21 as they are not in the v1 version of ths pedal. They  seem like high values and tone sucking in there own right possibly.

So maybe a 100k lin pot directly in place of R21 would work to attenuate the output without messing anything else up? just 2 of the lugs
connected?


Attached is a schematic of the original WH10 v1 for comparison

[antonis]

So maybe a 100k lin pot directly in place of R21 would work to attenuate the output without messing anything else up? just 2 of the lugs connected?

With just 2 of the lugs it shouldn't work like Volume pot (voltage divider) but as variable resistor..
(it will overload total Source resistance (R20//R21) especially at low settings..)

It might work in case of placing it in series between C15 & R21/R22 node (to form variable voltage divider with R21) but it should be overkill..

Just break the node of C15/R21/R22 and replace R21 with 100k pot (Log if handy) wired with lug 1 to GND, lug 2 to R22 left leg & lug 3 to C15 right leg..

[duck_arse]

just so you follow that I'm not quite following - if you put a pot to control the wet output which is too loud, and then bypass the wah [fet switching], the "dry" signal will be lesser volumed as well, passing thru an R21'd pot. wouldn't it?

[where is the beer, antonis? can i have cider instead?]

[Josh_bb]

just so you follow that I'm not quite following - if you put a pot to control the wet output which is too loud, and then bypass the wah [fet switching], the "dry" signal will be lesser volumed as well, passing thru an R21'd pot. wouldn't it?

[where is the beer, antonis? can i have cider instead?]

i'm confused now, i know it can be done as i have 'seen' modded ones just noone has ever posted how they did it

[antonis]

Mr. Duck tells you that if you only want "wet" signal with adjustable Volume level, you have to place respective device - pot, obviously(*) - BEFORE wet & dry signal mixing point..
(which point is on Q4 Gate..)
(*) Refer to posts #3 & #4..

Maybe a 10k pot (with 1k resistor in series with lug 1, if you like no "mute") on IC01 OUT (pin 1) should do your job..
(IMHO, 10k value is a good compromise between neither significally upseting C11/R14 cutoff frequency nor IC01 Out drive..)


[cider served, Stephen, what about apple pips..?? )

[Josh_bb]

Mr. Duck tells you that if you only want "wet" signal with adjustable Volume level, you have to place respective device - pot, obviously(*) - BEFORE wet & dry signal mixing point..
(which point is on Q4 Gate..)
(*) Refer to posts #3 & #4..

Maybe a 10k pot (with 1k resistor in series with lug 1, if you like no "mute") on IC01 OUT (pin 1) should do your job..
(IMHO, 10k value is a good compromise between neither significally upseting C11/R14 cutoff frequency nor IC01 Out drive..)

cheers,

the dry out will be essentially lost as i will disconnect the socket and use the hole for the volume pot.
Your right i dont want it to completely mute, I just need to be able to dial down the output gain a little when needed

[antonis]

...to dial down the output gain...

Don't say that word, 'cause Mr. Duck will come back and re-propose to make R11 variable..!! 

[ and I'll be left out of cider..]

[duck_arse]

...to dial down the output gain...

Don't say that word, 'cause Mr. Duck will come back and re-propose to make R11 variable..!! 

[ and I'll be left out of cider..]

fear not! antonis. also, no apple pips, pear cider for me.

I'm only slightly less confused now. the dry output would not be affected by the volume/gain mods. it's only the "dry" signal that appears at the wet output when the effect is bypassed ...... ..... R14? if that was a 500k pot, with the wiper connected to the fet, maybe?

just listen to antonis.

[Josh_bb]

...to dial down the output gain...

Don't say that word, 'cause Mr. Duck will come back and re-propose to make R11 variable..!! 

[ and I'll be left out of cider..]

fear not! antonis. also, no apple pips, pear cider for me.

I'm only slightly less confused now. the dry output would not be affected by the volume/gain mods. it's only the "dry" signal that appears at the wet output when the effect is bypassed ...... ..... R14? if that was a 500k pot, with the wiper connected to the fet, maybe?

just listen to antonis.

explodey head time.

I am going to have a fiddle this weekend, I am converting it to true bypass because it seems to suck all the fun out of the sound when in bypass at the moment

[Josh_bb]

Just break the node of C15/R21/R22 and replace R21 with 100k pot (Log if handy) wired with lug 1 to GND, lug 2 to R22 left leg & lug 3 to C15 right leg..

HI, I tried this and it sort of worked but not much happened until you turn the pot most of the way, then it got quiet quickly before going silent.

I would love to know the theory behind why replacing R21 (100k) would act to attenuate the output.

[antonis]

I would love to know the theory behind why replacing R21 (100k) would act to attenuate the output.

Let's "ignore" HPF C15/R21 (10μF is quite high value to be considered "short" at lower frequency of our interest..)

Q4 acts as output buffer - meaning it "transforms" high input impedance (510k in our case) to low output one (1/g_m of Q4 in our case - 3 mmhOs, say..)
(in case you wonder the reason for transformating an already low impedance signal to another one, it's done for power (current) benefit..)

On Q4 Source is formed a resistive voltage divider with 1/g_m (333 Ohms) for upper resistor and R20//R21 (9.1k) for lower resistor..
(ignoring R22  + with whatever impedance comes next, set in parallel with R20/R21 equivalent resistance..)

You can see we've a voltage dividing effect by the ratio of 333/9100..

Replacing R21 with an equal value pot, we form an extra voltage divider "off-setting" previously set dividing point from Q4 Source to pot's wiper - upper pot's resistance (between lugs 2 & 3) is now effectively set IN SERIES with Q4 Source..

Voltage dividing ratio on Q4 Source remains the same as previously but we now have a signal source (as seen from next effect input) of much higher output impedance..!!
(it can less effectively drive the same following impedance as before..)


P.S.
I'm not sure if I made things more clear or confused you even more..

[merlinb]

Here's what I would do. One cut track.

[Josh_bb]

Here's what I would do. One cut track.

thank you that's really interesting

[Josh_bb]

Let's "ignore" HPF C15/R21 (10μF is quite high value to be considered "short" at lower frequency of our interest..)

Q4 acts as output buffer - meaning it "transforms" high input impedance (510k in our case) to low output one (1/g_m of Q4 in our case - 3 mmhOs, say..)
(in case you wonder the reason for transformating an already low impedance signal to another one, it's done for power (current) benefit..)

On Q4 Source is formed a resistive voltage divider with 1/g_m (333 Ohms) for upper resistor and R20//R21 (9.1k) for lower resistor..
(ignoring R22  + with whatever impedance comes next, set in parallel with R20/R21 equivalent resistance..)

You can see we've a voltage dividing effect by the ratio of 333/9100..

Replacing R21 with an equal value pot, we form an extra voltage divider "off-setting" previously set dividing point from Q4 Source to pot's wiper - upper pot's resistance (between lugs 2 & 3) is now effectively set IN SERIES with Q4 Source..

Voltage dividing ratio on Q4 Source remains the same as previously but we now have a signal source (as seen from next effect input) of much higher output impedance..!!
(it can less effectively drive the same following impedance as before..)

P.S.
I'm not sure if I made things more clear or confused you even more..

Thanks I appreciate you taking the time to write this, I am on a learning curve so this is all very useful information to me. I need to think about it so it sinks in.  cheers for this

hmm I have a feeling these are dumb  questions but...   where did you get 330ohms from and R20/R21 - 10/100 = .1  I will try and work it out

[antonis]

hmm I have a feeling these are dumb  questions but...   where did you get 330ohms from and R20/R21 - 10/100 = .1  I will try and work it out

There aren't dump questions..!!
(only dump people - but they usually ask nothing  ..)

Voltage dividers are present EVERYWHERE (they live among us.. )
It's quite "odd" to get it for an JFET in the meaning of actuall resistance (much more easy for BJTs where that resistance appears as equivalent Base-Emitter forward biased diode..)

Signal on Q4 Source, looking back, "sees" a node between Source itself and source load (its eyes are splitted into 2 different paths..)
It considers Source as a Voltage source with its internal resistance(*) in series and source load as whatever Source has to feed(**)..!! 

(*) Due to JFET properties of voltage to current amplification, transconductance (g_m) is the natural parameter for gain - e.g. for a grounded Source CD amp, gain is -g_mR_D)
g_m is the inverse or resistance (in mhOs units - inversed Ω) so 1/g_m "represents" resistance..
That resistance is what signal "sees" inside Source when looking from the node of Source & source load..!!
(the upper part of voltage divider..)

(**) As source load should be considered whatever Source drives..!!
(in our case, R20 in parallel(//) with R21, because we consider OUT as an open circuit - or as a next stage with significally high input impedance..)

Estimating g_m = 3mmhOs, 1/g_m is about 33Ohms (upper part) & R20//R21 = (10k X 100K) / (10k + 100k) = 9.1k (lower part) so voltage dividing effect is 330/9.1k ..