Reversing pot taper?

[ashcatlt]

So I'm building a volume pedal into a crybaby shell.  Using a 500K pot from a Les Paul, which I've modified using conductive paint so that it will actually go from 0 to 10.  This has reduced the overall resistance to somewhere around 380K, and I'm thinking of buffering on either side of the pot itself.  

Problem now is that the log taper pot turns the wrong way.  Starting from heel down it gets louder very fast up to about halfway and then the change is very subtle.  It really needs a reverse log pot, but I'm wondering if there's a relatively easy way to make it act that way.  I'm open to active solutions, as long as it doesn't get too complicated.  

Thanks guys.

[JohnH]

I've also had a few cases where I really wanted a reverse log pot. I havent come up with an easy electronic solution to convert a normal log or lin pot. What you probably really want is a Type C taper, which is reverse log, as against A = normal log and B = linear.

Other ideas:

1. Could you get a linear pot of about 30-40% of the value you need, and scrape along the edges of the track over one half to increase its resistance?
2. Could you get a linear pot of about 2x what you need, and along the edges of the track on one half, put a series of non-continuous dashes of conductive paint - - - - to reduce resistance on that half? Dashes on one side of the track would be offset from those on the other, ie space dash space on the outside side, and dash, space, dash on the inside.

In each of 1 and 2, you can have a multimeter giving you a stroke by stroke read-out of what you are changing overall resistance to. The centre of the track would probably be best left unaffected for smooth operation. Maybe a combination of 1 and 2 where you start with a linear pot of perhaps 50% to 100% of the right value and do some of each idea.

3. Is there a mechanical solution for the pedal, where you mount the pot in the other direction or change the linkage in some way so it works with a normal pot?

cheers

John

[sbgodofmetal]

not sure of a way to do this but why not try opening the case and just flipping the internal component on the shaft wouldn't that work

[JohnH]

Further thought 4: I assume that heel down= minimum volume is whats needed? Because if you could use toe down = minimum, you could just reverse the pot outer lugs and the current log taper would be all correct for smooth swells.

5. It sounds like as you start to press down the toe end from fully heel down, then the pot is initially tracking up the steep part of its taper, resulting in a too sudden swell? I was thinking about what kind of active circuit could use that characteristic and give an output that is more gradual. I think there is a JFET circuit that will do that, and also possibly an opamp version. I can picture the JFET one, but I know opamps are more your thing?

John

[ashcatlt]

Well, you know, I was thinking about just propping up the heel end, but you've got me thinking that I might just be able to work the thing backwards. Would be a bit counterintuitive and take some getting used to, but it's the easiest option.

I do understand opamp circuits a bit better, and have a bunch. I can follow a schematic, though. I was looking at an inverting opamp gain stage with the pot as part of Rin, and some very small resistor in series and have Rf the same as the fixed one. 5spices linear pot sweep seemed to give something close to a log sweep, so maybe if I wired this log pot correctly I could at least get close to linear? Problem there is that it doesn't get to silence. Even with 47Ω for the fixed R's it only gets less than 80db attenuation. This is probably well down in the noise floor, but I don't think it's going to be enough, especially if there's any significant compression/overdrive/distortion downstream.

[JohnH]

ash - just been fiddling with 5spice, ill show you the JFEt one in a few minutes....

[JohnH]

OK, thinking of something like this:


Your 380k pot is in there, doing two jobs. At min volume, it is grounding the input for silent output. Towards max volume, it is bypassing the source resistor R1, for more gain.

Here's a db pot:



The plot shows the pot changing from 10% up to 100%, ie from about half way up to max, and theres some significant gain happening in this upper part of the travel. Not shown, the range from 0 to 10% goes from huge negative db up to -6 db.

Ahead of this circuit would be a simple buffer - another JFET in a standard configuation. After it, these values are fine to drive most guitar amp inputs (output Z = 10k)

I've put a 10V battery - i find this gives closer results to a real 9V using this JFET on 5spice.

cheers

John
ps C2 is a bit mean! it should be about 220nF or more

pps: maybe your pot is actually more like at 80 to 90% at the mid position? in which case these values might be better, following further fiddling:
R4 = 270k, R6=1.5k. These seem to give, zero output at 0%, -10db at 10%, +3db at 90% and +10db at 100%

[ashcatlt]

I have to admit that I'm having some trouble figuring out how the pot taper will effect this. It looks to me like this just going to exacerbate the problem. Those bottom jumps are huge with this linear pot, and the resistance of my pot increases even more quickly down there. Or maybe I'm missing something?

Then, can this be done with a 2n5458? I have a few of those around. And can we keep it from going past 0db? I guess it could be attenuated after, though I'm unsure how this would affect out-Z.

[JohnH]

2n5458 will work, with a tweak or too. Just see my pps note above - if you have some, its worth a breadboard test.

If you try it, change values of R2 to get about 5.5v at the drain - not critical but thats about optimum with a 9V bat.

[ashcatlt]

I'm sorry John but I'm still not seeing how this helps. It looks like the first 90% of this graph is a reverse log response. Since I've already got reverse log response, won't this just make it worse?

I'm starting to think that it would be much easier to get a linear pot and apply an appropriate tapering resistor to get what I need here. If it's buffered on either side, then it doesn't much matter what the effective pot value is. I need to test the other excess pot I've got out of that LP and see what it's taper is.

[JohnH]

These are plots with the revised values as above:



OK, at 0%, no output, not shown
At a small press of the pedal, say 1 on a knob and about 20% of resistance, you get -5db
At halfway, say between 85% and 90% resistance, you have about +3db
At 100%, full toes down, you have +10db

So for two roughly equal part turns of the pot, from 1 to 5 and 5 to 10, you get roughly similar increase in db, being 8db then 7db.

Not a perfect taper, but maybe useful? At least there is significant extra for each increment.

I’ll shut up now.

The linear pot with taper resistor is a good idea. Maybe its a 500k pot with about an 82k resistor?

John

[ashcatlt]

Thanks for looking into this with me John. After all that, I guess I'm just going to go the passive route. Turns out my LP studio has 300K linear Volumes and 500K audio Tones. I had chosen one of the Tones for this because the volume portion was originally going to be purely passive, and I wanted the 500K. Now that I've decided to buffer before the Volume pot, I can really use anything. Since I've converted the guitar to Master V and T, I've got a spare of the linears as well.

I didn't really want to have to paint another pot, but I've done it, and while I haven't wired it in yet, all tests are positive. After painting it's down to about 222K altogether. The Secret Life of Pots says that somewhere around 1/4 to 1/5 the pot value makes a good tapering resistor. I think I've got some 68Ks around somewhere which should be plenty close enough. Heck, I may even just try it straight linear.

I'm looking for nice smooth volume swells from 0 to full on. As you may be aware most of my guitars have no pots at all. Also, I've only got so many hands. It just seems more natural to do volume swells with my feet.

I've talked about this pedal before. I'm using the switch to put a cap (the tone cap out of the same LP) from the hot input to ground, so it's like turning a tone pot down to 0. Then there's a DP5T switch which selects the input impedance of the basic opamp buffer stage. I had the passive volume in between there, but now I guess I'll stick it after the buffer and put in a second opamp. Luckily I left plenty room on the perfboard.

[JFrankParnell]

i imagine this set up to be rack and pinion... what if the pinion was cam shaped? With a linear pot. Theoretically, yes...practically...eh.

[sumgai]

ash,

First, if you short the feedback loop (R_f) of an op-amp's inverting input, the output should be 0*. Choose R_i to equal your pot's max resistance value, and your op-amp will have a gain factor of 1, when the pedal is all the way forward/down. Presto, instant volume control. Of course, this is just a fancy, power-wasting way of doing the same thing as a passive device.....

The point being, the response curve at the output will work exactly opposite of a variable and non-linear R_f, all things considered.

Second, while I appreciate all this Nuttiness, it seems to me that the KISS principle should be in play here. Try imitating the very guitar you pulled the control from - reverse the wiper and hot terminals. Now reverse the hot and ground terminals, and reposition the shaft within the gear for best effect. That should yield something approximating the desired results (albeit without grounding the amp's input at zero output - but LP's have been doing that for years, and nobody has really complained yet...).

HTH



sumgai


* For the curious....

The gain equation for an inverting op-amp is usually shortened to:

V_out = R_f/R_i (feedback resistor divided by input resistor)

If R_f is equal to 0Ω, then dividing that by any R_i is gonna yield the same answer: 0. From the equation, that is the number to be assigned to V_out. Kinda equals a pretty quiet output, eh?

For those that still have doubts, try looking up the terms "inverting op-amp virtual ground", and see what turns up. Interesting, but no more than that (unless you like that sort of stuff.....).

[ashcatlt]

Hey, it works!

I used a 47K resistor between ground and wiper. It seems it may have gone a bit too far in the "logarithmic" direction. Or maybe I just need to learn how to use the thing... Maybe I'll try just the straight linear pot and see what happens.

Next I have to re-figure the variable impedance thing. It works and all, but this was my first active circuit. I just kind of chose the values based on some poor understanding of the consequences. If I remember correctly the lowest value is 500K. That would be good for a typical amp input, but remember that my guitar doesn't have any pots. With a standard strat, the Volume and Tone work to bring that effective impedance down to 125K before you even factor in the amp. So, yeah, some things there to consider.

Apr 2, 2011 23:26:47 GMT -5 sumgai said:

...
The point being, the response curve at the output will work exactly opposite of a variable and non-linear R_f, all things considered.
...
The gain equation for an inverting op-amp is usually shortened to:

V_out = R_f/R_i (feedback resistor divided by input resistor)

Gain increases proportionally to Rf. If resistance is increasing more slowly as we reach toe down then gain also will, no?

It seems, also, as though if I was to put ground at the "toe-down" lug I'd get operation opposite of what I wanted no matter what I do with the wiper and other lug. If I could live with putting down my heel to get louder, this whole thread would be moot.

[newey]

Glad it works, Ash!

If I could live with putting down my heel to get louder, this whole thread would be moot.

Ever since the first automobile was fitted with a foot pedal accelerator, we have been conditioned, Pavlov style, to want to "floor it".

[JFrankParnell]

spin the whole thing around 180*. Shim the 'new' toe side up a bit.

[ashcatlt]

That was an idea, also. It would kind of end up a bit high. Also, the fulcrum is off center toward the heel. There's a springy thing and some resistance from the pot itself, but it feels more secure at the toe down. I'd hate to have the thing fall to silence in the middle of a song.