Pot Taper Choice: Linear/Audio/Reverse Log

[Steve Mavronis]

Dumb question: I know a certain value pot is just representing the range of motion from zero to max resistance, so the taper that you choose is more of a personal preference right? For instance if you like say a 50K level pot to be audio log and the effect pot to be 500K reverse audio log, then that works just as much as using linear or any other taper combination. So it really comes down to what control feel you like when turning the knobs?

P.S. I've read The Secret Life of Pots.

[B Tremblay]

Yes.

[Mark Hammer]

Well yes and no, actually.

The rule of thumb I apply is that if I need to know where the "middle" is, then linear is a must.  Examples include tone boost/cut controls where the "middle" is flat, or blend/mix controls where the "middle" is a 50/50 combination.

Obviously, there are many instances where there IS no "middle".  So there is no "middle" for distortion amount, or phaser speed, or tremolo depth, or flanger regeneration. Does the choice of taper become a matter of preference or taste in these circumstances?  Again, yes and no.  There are some things where log or anti-log is pretty much a requirement simply because of the way the human auditory system works, but there are other things where it is a matter of choice/taste.

For example, the human ability to make fine distinctions in speed or rate of change in time is such that log or anti-log speed pots are almost always de rigeur for LFOs or else the parameter becomes largely undialable and the control way too twitchy to be of any use ("Set it at the 7:20 position, and don't go above 7:35 or else it's too fast.").  When it comes to volume pots, human hearing dictates that they need to be log for any sense of level adjustment that is predictable and repeatable.

In other areas taste/choice is the dictating factor.  For instance, is there any theory-driven basis for using a log vs linear pot for phaser/flanger resonance?  Not really.  Some folks will like to be able to fine tune heavy resonance, while others will better appreciate the capacity to dial in shades of subtlety.  Distortion intensity controls are the same way.  Some will like a taper that lets them more easily whip through the subtle stuff and easily differentiate  between shades of intense distortion, while others will prefer the opposite.

The bottom line is that the taper needs to be selected on the basis of what you need to accomplish via that control.  Some things demand a "middle".  Some thngs demand an equivalent progression through a range of settings.  And some will be motivated by a desire to zero in on a particular range of settings.

[Steve Mavronis]

That's a very good explaination Mark! Since you mention different effects types, what about Compressor/Sustainer? I've never used one so don't know what to expect dialing things in.

[Ronsonic]

One thing that customers have taught me to do is to put some nominal "flat" "unity" "normal" most ordinary setting at 12 o'clock. In something like that compressor, the drive and output levels would result in more or less unity gain overall when straight up. That's just a very general guideline, not even a rule really, but seeing how people use stuff and asking them what they're thinking, that's what I get back.

[Mark Hammer]

That's a very good explaination Mark! Since you mention different effects types, what about Compressor/Sustainer? I've never used one so don't know what to expect dialing things in.

It will depend on how the circuit in question accomplishes its compression, and what the (pardon the pun) limits of compression are in that circuit.

So, let us say that an LDR is used to attenuate signal to ground, and the LED driving the LDR is adjusted by the amount of gain of an op-amp pushing it.  That compressor could be pushed by any of a broad range of signal levels.  At a certain point, adding more gain is unlikely to result in any additional drop in the LDR resistance, if minimum LDR resistance is reached at only a fraction of maximum tolerable input signal level.  In such an instance, if you were using the pot to adjust the op-amp gain via the feedback resistance, you may want a reverse-log pot such that the amount of resistance increase per degree rotation dropped as you turn clockwise.  If the gain of that op-amp was fixed, and you governed the birghtness of the LED (and LDR resistance) by attenuating the signal fed to that fixed-gain stage, you'd likely use a log pot to effectively "turn the volume down or up" of whatever you were feeding the fixed-gain stage with.

And so on.  "Correct" or most desirable taper would depend on the particulars of the circuit, and the circumstances you expect it to operate under.

[Steve Mavronis]

It will depend on how the circuit in question accomplishes its compression, and what the (pardon the pun) limits of compression are in that circuit.
And so on.  "Correct" or most desirable taper would depend on the particulars of the circuit, and the circumstances you expect it to operate under.

Well personally I'm in the planning stages to clone a 70's MXR Dyna Comp / Ross Compressor hybrid. Not sure what taper either of them originally used. I know there are others out there, but you know me I have to re-invent the wheel to get any satisfaction out of the process!

[Mark Hammer]

Goven that the Compression pot dictates how much current is being fed to the gain-control pin of the 3080, I'm gonna say it wants a linear pot, but I'll stand corrected by those who follow such things more closely.

[Steve Mavronis]

Yeah I was reading a R.G. Keen article about making a D-Comp/R-Comp hybrid and if I remember correctly it mentioned the Sustain control should be linear and Output should be log which I assume is audio taper. I won't be using the nice PCB layout that was included (probably should) since I'll be taking another stab at making my own compact 2.0" x 1.7" layout to etch. I'm currenty studying an MXR schematic to see where to apply the Ross changes.

[d95err]

The rule of thumb I apply is that if I need to know where the "middle" is, then linear is a must.  Examples include tone boost/cut controls where the "middle" is flat, or blend/mix controls where the "middle" is a 50/50 combination.

To be really picky, some EQ:s or balance circuits use special tapers where the center is 50/50 resistance, and then log/antilog curves towards the endpoints ("S" or "W" taper, I think). So sometimes the taper is not linear even when the middle is well defined... 

[Mark Hammer]

True.  I can't imagine any stompboxes would use such exotic tapers, though.  They would more likely be found in higher-end audio gear.

But your point is well taken; sometimes it takes more than linear to make a "middle".

[Steve Mavronis]

This is sort of a dual use topic for me mainly for pot taper uses, but since I'm researching a compressor to clone look at this layout and schematic from a Japanese site:



I notice this person who supposedly analyzed a real MXR lists the Dyna Comp Sensitivity pot as C500K which is reverse log. In our earlier replies we were thinking linear would be bettter for this application? Hopefully I'm not confused which pot controls the amount of sustain; it should be the Sensitivity right? The internal Trim pot I think is supposed to affect your string attack or something.

[Mark Hammer]

Keep in mind that the Sensitivity control does not do anything of itself.  It simply governs how responsive the unit will be to what gets fed into it.  You can set the Sensitivity/compression control to minimum and still get compression if you feed it a hot enough signal, since nothing is shunted to ground, only impeded by a series resistance.  Under such unpredictable conditions, there is really not much point, and nothing to be gained, by using anything other than a linear pot.

[zombiwoof]

True.  I can't imagine any stompboxes would use such exotic tapers, though.  They would more likely be found in higher-end audio gear.

But your point is well taken; sometimes it takes more than linear to make a "middle".

Actually, they do.  Smallbear sells a "W" taper pot that is used in some Ibanez pedal, and the common vintage wah pot was a special taper, something like an "S" taper I believe.  So it's not unheard of in effects boxes.

Al

[Govmnt_Lacky]

True.  I can't imagine any stompboxes would use such exotic tapers, though.  They would more likely be found in higher-end audio gear.

But your point is well taken; sometimes it takes more than linear to make a "middle".

Actually, they do.  Smallbear sells a "W" taper pot that is used in some Ibanez pedal, and the common vintage wah pot was a special taper, something like an "S" taper I believe.  So it's not unheard of in effects boxes.

Al

I have built Tubescreamer pedals with B20K, B25K, and W20K pots and there is a HUGE difference in the "usability" of the Tone pot when you use the recommended W taper pot. The tone is much more "shape-able" with the W taper pot.

[Projectile]

True.  I can't imagine any stompboxes would use such exotic tapers, though.  They would more likely be found in higher-end audio gear.

But your point is well taken; sometimes it takes more than linear to make a "middle".

Actually, they do.  Smallbear sells a "W" taper pot that is used in some Ibanez pedal, and the common vintage wah pot was a special taper, something like an "S" taper I believe.  So it's not unheard of in effects boxes.

Al

I have built Tubescreamer pedals with B20K, B25K, and W20K pots and there is a HUGE difference in the "usability" of the Tone pot when you use the recommended W taper pot. The tone is much more "shape-able" with the W taper pot.

Just thought I would throw in that you can use a 5k linear pot in place of the fancy W pot on the tube screamer and it works just fine. I don't know why on earth they still use that expensive W taper pot because the 5k pot works just as well or better to my ears. It gives the same tonal range as the 20k pot, and an even smoother, more natural sweep in my opinion. 

I decided to try the 5k lin pot after doing some calculations and finding that the 25k tubescreamer tone pot isn't really doing anything for most of it's rotation. It's only actually effecting the audio path at the ends of it's range, which is why they need that strange taper. A 5k pot fixes this problem. Now the pot gets used for the entire sweep. I still can't figure out why they used a 25k pot in the first place. Try a 5k pot some time. I think you'll be pleasantly surprised!

[Govmnt_Lacky]

I decided to try the 5k lin pot after doing some calculations and finding that the 25k tubescreamer tone pot isn't really doing anything for most of it's rotation. It's only actually effecting the audio path at the ends of it's range, which is why they need that strange taper. A 5k pot fixes this problem. Now the pot gets used for the entire sweep. I still can't figure out why they used a 25k pot in the first place. Try a 5k pot some time. I think you'll be pleasantly surprised!
[/quote]

Thanks for the tip! 

[Mark Hammer]

What is a W-taper?  Is it log/antilog?

[Govmnt_Lacky]

W taper is a modified Linear taper. It looks like an elongated S curve.

Check out this link to Alpha pots.
http://www.alphapotentiometers.net/html/taper_curves.html

[Mark Hammer]

Okay, so yeah it is a log/antilog pot.  Quasi-linear because at midpoint you have half the resistance, but non-linear from the midpoint outward.

That taper can actually be mimicked to some extent by addition of parallel resistors from wiper to outside lug, though I suppose having the pot come with the taper built in helps things along.

The thing with the TS9 tone control is that it produces two complementary effects which sort of cancel each other out for a portion of the range of adjustment.  The special taper gets you quickly through that part of the range where the cancellation occurs, and moves you to the portion of the range where the one effect is much stronger than the other whether it is treble boost outweighing treble cut or treble cut outweighing treble boost.  The 5k pot suggestion does pretty much the same thing, though with a different feel.

Jack Orman has a nice web-page devoted to TS tone-control variations: http://www.muzique.com/lab/tstone.htm   One of the things missing from it is any discussion of how placing aparallel fixed resistor between the wiper and one or the other outside lug can assist in improving the "dialability" for that user (some folks like nuancing the treble boost, while some prefer more dialability for treble cut).