robotic potentiometers

[R.G.]

What about using some of the small relays from the links below to make an attenuator? For example 10 relays per pot won't take much space. I'm using the SIL version in some of my builds and I can't hear it switching at all. Since the whole idea is to be able to memorize presets this way you don't have to worry about pot's position. The switching latency will be couple of msec.

(1) what do they cost?
(2) are you using only 10 positions per pot?

One thing that's possible is to use JFETs to select resistor dividers in a binary format so that each pot is presented a binary word of, say, eight bits to represent a stepped one-of-256 value.

The beauty of stepper-motor pots is that the pots look like and function just like they always did, except moving to new positions when you tell them to. I haven't yet found another setup that gives you both pot setting and using the pot itself as a position indicator. But yes, there are lots of ways to do this.

[PRR]

> relays .... to make an attenuator?

How many relays?

Pot is numbered 0-10, ten gross positions (rounded).

I often note half-step, 20 positions.

I sometimes note "3.75", 40 positions.

I understand that some tone-stacks (and some players) like even finer divisions.

Say 64 settings.

64 3-buck relays is almost $200 per knob. Eight knobs 512 relays you get the $2 price, but $1,000? I can hire a kid to turn my knobs!

Yeah, binary tree. (Why I picked the decimally-odd value "64".) You do need more poles and throws to get a reduction in relay-count. If you had 8-pole relays.... but they ain't $3.

For a "small" number of presets (small enough to remember which-is-which), the cost and size of relays to give reasonable "pot" resolution seems to be larger than the cost and size of a separate pot per pre-set.... pots can be quite cheap.

> 10 relays per pot won't take much space

If ten is enough (it may be), then it may make sense. Sixteen settings needs "only" fifteen SPDT relays, eight DPDT relays.

> reads its own DC pot signal.
> value of DC voltage stored ... turn its own knob to get to what it thinks #2 is for it.

You can eliminate the gang-pot (a major hassle since all stock gitar pots are single) with RC servos. Don't touch the actual amplifier/effect pot. Turn it remotely with the servo. When pleased, hit "MEM 2". The CPU(s) memorize the most recent pulse-width. Twiddle, then hit "RECALL 2". The CPU(s) apply the remembered pulse-width and the RC servo rotates the working pot to that position. (It could additionally rotate the remote setting pot but that would be a contraption-- it could rotate a dummy display knob above the smooth unmarked setting knob.)

> use JFETs to select resistor dividers

Practical for the 9V world. Easy-enuff in a +/-15V world. Needs thought and probable re-design if a tube amplifier is involved: signal at tone-control can exceed 20V peak, and should swing 60V with tube-kink not FET-kink. Master Volume in power tube grids can be higher. Modern designers use FETs for switching in tube amps but this usually means re-scaling the gain and impedance structures for "FET friendliness" rather than raw tone.

Yeah, you knew this and tube-amps are a small niche. And if you HAD to, small unprotected MOSFETs can switch hundreds of volts.

> The beauty of stepper-motor pots is that the pots look like and function just like they always did

Agree, except that feedback pot.

An alternative is an encoder ring. Open that old ball-mouse, bust the rollers out, there's a slotted disk and LED-eye (or brush). That tells relative motion. With more tracks (and eyes) you have absolute position. In principle you can print a Grey-Code pattern on thin film; if reflective, even on the back of the knob. I don't offhand know any encoders so slim that they could hide behind a standard-shaft pot knob.

[MetalGuy]

(1) what do they cost?

If you buy 50pcs the price is ~2USD (at Farnell).

How many relays?

I'm OK with 10 positions but if you need 64 positions obviously it's not going to work for you.

use JFETs to select resistor dividers

If we're talking tube amps/preamps maybe these high voltage (up to 200V!) analog switches will do the job:

http://www.maxim-ic.com/datasheet/index.mvp/id/5771

http://www.maxim-ic.com/datasheet/index.mvp/id/6127

[tokyoburns]

So I'm thinking that the control cable to my pics will be a usb. Im hoping that I can power them the motor and the voltage across the pot and save a ton of cable space as well as provide the necessary digital input to the pics. they are easy to aquire and have a compact port that wont take up space on my pedals.

[tokyoburns]

Update to the project: I got Flowcode and have been learning to program with that. It has made it a lot easier on me. I haven't been able to test my code yet but it works the way I want it to in simulation. I have got some pretty sick features and am really excited to test them out. I am about to start gathering parts. still looking for the perfect stepper motors dual shaft 4 phase unipolar. If anybody has an update as to a good place to look I would appreciate it. Thanks!

[~arph]

I'm not sure if this is posted before, but here is an idea.

Instead of using a dual shaft stepper motor, we try to find a dual shaft potentiometer. This should be a lot cheaper. Now I haven't found dual shaft pots in my quick search, but maybe it is possible to take the back end of a dual gang pot off thus creating a dual shaft single pot.

EDIT: come to think about it.. that way there is no way of recalling position.

[stephanovitch]

This solution is very funny:
http://9giga.sfr.fr/n/50-17/share/LNK51844dc80ac853a2f/

[tokyoburns]

@stephanovitch

That is a really cool idea. I would have never thought of it although I'm sure everyone else here probably recognizes it. However that does not solve my problem. I really want the user to be able to control his pots by hand as well as externally. The idea which I'm trying not to reveal too much about calls for the pots to actually turn. although I suppose I could just make the pots turning for show and have that design actually be the resistance of the pot to eliminate all that turning noise of the pot from one transition to another. Now that I think about it it would make the programming  a hell of a lot easier. haha. I might have to rethink my design. Months of research and programming thwarted by a single post. haha. I'm gonna have to meditate on this one. Hmmmmmmm...... That circuit probably does really well when recalling quick transitions between resistances but it probably causes zipper noise when cycling through. Now you got me thinking hybrid. haha. Can you point me towards small relays or something that covers that whole circuit without going digital? Should I just be searching for PCB mounted dpdt relays obviously something worthy of audio. The amount of space it would take up in a circuit seems a little big. Great reply! I appreciate it.

[anti-idiot]

Why don't you try something with LDR's, like Mesa/Boogie's Triaxis?

@stephanovitch

That is a really cool idea. I would have never thought of it although I'm sure everyone else here probably recognizes it. However that does not solve my problem. I really want the user to be able to control his pots by hand as well as externally. The idea which I'm trying not to reveal too much about calls for the pots to actually turn. although I suppose I could just make the pots turning for show and have that design actually be the resistance of the pot to eliminate all that turning noise of the pot from one transition to another. Now that I think about it it would make the programming  a hell of a lot easier. haha. I might have to rethink my design. Months of research and programming thwarted by a single post. haha. I'm gonna have to meditate on this one. Hmmmmmmm...... That circuit probably does really well when recalling quick transitions between resistances but it probably causes zipper noise when cycling through. Now you got me thinking hybrid. haha. Can you point me towards small relays or something that covers that whole circuit without going digital? Should I just be searching for PCB mounted dpdt relays obviously something worthy of audio. The amount of space it would take up in a circuit seems a little big. Great reply! I appreciate it.

[tokyoburns]

@anti-idiot
How so? Can you point me towards a circuit diagram or explain a little further?

[anti-idiot]

@anti-idiot
How so? Can you point me towards a circuit diagram or explain a little further?

http://tubefreak.com/fschema.htm

http://tubefreak.com/triaxis2.jpg

The Triaxis uses LDR/LED combos replacing the potentiometers, but don't get confused by the schematic, they are simplified: i.e. the Treble "pot" is done by using 2 LDR (27 & 28), while the Bass "Pot" is done by using just one LDR (29). As you can see, every potentiometer uses two LDR, while the rheostats (variable resistors, like bass control or the mids in a Fender Tonestack, like the Traixis' one) uses just one LDR. Each "control" uses a parallel trimpot (you the total value is close to the real counter-part).

5th Reply will tell you how to set-up this monster
http://forum.grailtone.com/viewtopic.php?t=2840

Remember that in the controls that uses 2 LDR, when one LDR goes up, the other must go down at almost the same rate*

I wanted something like this some months before (but using a Marshall Tonestack) but don't know anything 'bout programming. About the code, how to store it, how to recall it and stuff like that, I cannot help (but i will apreciate it if you help me, lol)

BTW, sorry if something is poorly written or is kinda messy. Take care

[tokyoburns]

I will look in to it. I guess I don't quite understand the advantage. I cant replace the physical potentiometer in any way. I wouldn't want a pedal with no knobs or knobs that were just there for decoration. adding LEDs and LDRs sounds very space consuming. but like I said maybe I just dont understand. I will look in to it.

I have been teaching myself to program PICs using a program called FLOWCODE. It was made for students to learn on so it was very helpful. A lot of it was drag and drop using a flow chart. It also has a virtual simulation for mechanical parts like stepper motors and potentiometers so it worked out pretty well for me. I could help with some basic stuff but only if it pertained to Flowcode otherwise I am just as clueless. It still took me months to figure out what I was trying to do with a big hiatus in between but I got there. I understood very well how I wanted my pedal to work mechanically so preparation was a big key for someone like me who has limited knowledge of this kind of stuff. I go to school for electrical engineering and I own a small DMX lighting business at www.lightatmospheres.net so I have a pretty good base knowledge of computers and electronics but in no way am I a programing expert or even close to being an engineer yet.

I am really hopeful about this pedal design. It is far more complex and awesome than i have given away and am really excited to share it when it is finished. Thanks for all the replys!

[anti-idiot]

I will look in to it. I guess I don't quite understand the advantage. I cant replace the physical potentiometer in any way. I wouldn't want a pedal with no knobs or knobs that were just there for decoration. adding LEDs and LDRs sounds very space consuming. but like I said maybe I just dont understand. I will look in to it.

Well, you'll end up using the "outside pots" to save control the internal pots. You can use, let's say, five 10k (Gain, Bass, Mids, Trebele, Volume), store a setting (the "internal pots" will use the real values (1M, 250k...) in the uC, have a few settings and recall 'em while ignoring the position of the "outside pots".

Marshall's JMP-1 (rackmount preamp) uses buttons for the selection (bass, mids, treble, gain, etc), a rotary encoder for changin' the value and a LED-Display so you know the setting of the "pot" you have selected.

[tokyoburns]

yeah that would be the idea if I did some sort of hybrid thing but I would need a way to switch from the "outside pots" (as you call them) and the "internal". Wouldn't be hard to figure out. The trick would be finding the most compact and practical solution that resembles one of the ways we are discussing for the internal and I think I would incorporate the idea but so far space on a pcb board seems to be the deal breaker for all the ideas so far. but I am sure there is a solution that is not digital somewhere.

[Boobslappy]

What about running an Arduino to read the pots' resistance, store the value after you press a button (get pots with a switch) then recall this value via a motor shield and small servos to move the pots.  I assume you could even interface with an LCD to cycle through different save presets

[stephanovitch]

@stephanovitch

That is a really cool idea. I would have never thought of it although I'm sure everyone else here probably recognizes it. However that does not solve my problem. I really want the user to be able to control his pots by hand as well as externally. The idea which I'm trying not to reveal too much about calls for the pots to actually turn. although I suppose I could just make the pots turning for show and have that design actually be the resistance of the pot to eliminate all that turning noise of the pot from one transition to another. Now that I think about it it would make the programming  a hell of a lot easier. haha. I might have to rethink my design. Months of research and programming thwarted by a single post. haha. I'm gonna have to meditate on this one. Hmmmmmmm...... That circuit probably does really well when recalling quick transitions between resistances but it probably causes zipper noise when cycling through. Now you got me thinking hybrid. haha. Can you point me towards small relays or something that covers that whole circuit without going digital? Should I just be searching for PCB mounted dpdt relays obviously something worthy of audio. The amount of space it would take up in a circuit seems a little big. Great reply! I appreciate it.

Switching pulse to ground(mosfet on volume pot) during relays driving will cancelling noise

[rip]

Any progress with this?

I was working on the same thing, and found this post in searching.

I must say while this is a great idea, after pricing things out it gets very expensive.

Thanks
Rip

[R.G.]

I must say while this is a great idea, after pricing things out it gets very expensive.

I finally tumbled to something a while back. If something is good and useful, but no one does it, it can only be for a few reasons. They run something like:
- it's too expensive
- something else is better
- no one really knows how

That last one is where inventors live. If it's obvious how to do it, then it's nearly always too expensive to be practical for most people.

Occam's Razor is a B$#C#.

[teemuk]

I have to admit that if I had to make a system similar to this thread's idea, then I'd rather choose rotary encoders and those "electronic volume ICs".

[R.G.]

For function, yep.

That still leaves visible feedback to the guy turning the pot as a problem.