Remote controlled potentiometer using MIDI or expresion pedal

[gtudoran]

Hello guys,
I'm working right now on a small system that will allow you to control a standard potentiometer using either MIDI or Expresion pedals. There is a lot of things to work on but i will present you the concept.
The idea behind this was simple: a electro-mechanical module that can be used virtually on any guitar amplifier to control various functions (potentiometers) and also preset storing (MIDI CC messages will be used). Also another requirement was this: anyone with a screwdriver should be able to fit it into position and give it a go in under 10 minutes.
The project is using a very small stepper motor with 1.8deg / step (standard) and it will be controlled in full step mode using a dedicated driver. The shaft of the motor will be directly connected to the shaft of the potentiometer using an elastic coupling (similar with the ones that are used in CNCs). That will allow me to have quite an ok resolution for both models of potentiometers 270deg. and 300deg. (a small switch will select the model of potentiometer).
The "zero point" setting will be made using an optical slot sensor (infrared) that will set the zero point of the motor (each motor will have a home button that will allow you to set the zero point before you will connect the potentiometer to the elastic coupling).

The system will have only 23mm wide (the motor is only 20mm wide) and can accommodate quite a large number of amplifiers. The command box will allow you to use MIDI or an expression pedal. The command unit will be able to drive a maximum number of 5 modules.
Now... because of some certain reasons i've decided that each motor will have its own drive module that will be made by a stepper motor drive unit and a microcontroller unit - MIDI signal will be split in 5 (using some inverting buffers) and each motor will be independently commanded.
The reasons are:
- modularity (maybe some users will need 2 modules instead of 5 so.... he will buy only 2)
- programs are executed in a seq. mode inside a uController so if you like to actuate the motors in the same time... well you would need parallel processing and this is what i'm doing but in a more accessible way
- fault isolation (in a live situation is mandatory)

Now some ... things i would like to clarify :
- this system is designed for existing amps / preamps / power-amps etc for a new design motorized potentiometers can be used
-  NO is not the same with the system presented at NAMM this year - totally different approach / driving methods / driving elements etc. And yes my system is faster in terms of rotating speed from point A to point B
- Yes i know the system designed for  Neil Young and no it's not the same

The above notes were just a collection of questions that i've received in various situations.

In the end the project will be posted on kick-starter (along with another quite popular project - THE GTFO) and will be posted in Open Source community free for personal use (i'm trying right now to get a small patent for it at least in EU - software and design).

Now long story short some pictures with from the building phase - as i said a long way to go especially now because i'm going to change my shop location in couple of days but i would like to finish an working prototype until the end of the month.









Relevant comments are highly appreciated and encouraged. Also as an ending note, there are other ways for zero-point detection but some are more expensive, some are not suitable for such a small stepper.

Regards,
DeX

[Gus]

Have you looked at RC servos?  PW control

[gtudoran]

Hello Gus yes that was my first thought but then a limitation comes in place. Servos (at least RC servos) have only aprox. 180 deg. of rotation. That will limit a lot the travel of a pot. You can modify them for 360deg. rotation but then you would loose the feedback control that is made with an internal trimmer.
Also the torque for the same dimensions is smaller to servos, there are some HT servos but they are much more expensive (also servos that are using multi-turn trimmers are available but they are quite rare)

Regards,
DeX

Have you looked at RC servos?  PW control

[G. Hoffman]

If your stepper looses steps (and they will, without frequent maintenance), you are going to have problems with it trying to go past the ends of the pot's travel.  You might try using double pots, and get feed back from the second pot - it's kind of tough, but if you can work out the details it would be much more effective.


Gabriel

[gtudoran]

Hello Gabriel,

Thank you for your answer. That is why i'm going to use 2 end-stops. Loosing steps is more freq. when you are using the stepper in microstep mode and in high-speed applications, so in full-step mode and low speed (max. rated speed for the model is 600rpm and i'm going to use it at 300 max).
Also using a potentiometer's shaft is kinda out of discussion because as you said ... it's kind of tough.
I'm relaying my project on the fact that: if a stepper is going more then ok in a medium dimension CNC it will perform quite OK in this environment where actually is used for very small amounts of time and under no notable torque.
But yes, i could implement a small rotary encoder with 1deg (or 2 deg.) stripes and use it as a feedback system (it should not be too hard to build it using an acetate disk and a small slot infrared sensor)

Regards,
DeX

[R.G.]

I think a system like that is a worthwhile thing to produce commercially, but I feel obligated to tell you that it may well not be patentable. You should check with your local laws and rules on patentability.

I first proposed something like that mechanical setup for mechanized pots about 20 years ago. I have posted the commentary on line several times. I am fairly sure I was not the first to come up with something like this, although I had not seen it described before when I conceived it. I thought if it as covered by the "one skilled in the art" exclusion under USA patent law. As I said, your local laws on patents may be different.

At the time, there were many 5.25" floppy disk drives available at very low prices, and they all included a small stepper motor, most with a double ended shaft, usually 0.25"/6.35mm diameter, which happens to be the diameter of many US-style potentiometers. It was quite natural to connect the stepper to the pot. In my protos, I used rubber tubing for the connection between motor and pot shaft, with 0.25" interior tubing being easily available.

It worked first time, of course. The motor turned the pot well. Using such a small motor meant that for 24mm pot bodies, the motor torque was low enough not to overstress the mechanical stops. For larger motors, a limiting resistor in the motor windings cut the available torque. Anything to limit motor stall torque will do that pretty well.

The two problems of reading position and manual adjustment remained. I solved the manual adjustment problem by the double-ended motor shaft. I installed the motor with its extra shaft sticking out of the chassis and put the knob on the motor. with the motor un-energized, the motor and pot could be turned easily, and the motor's cogging led to the feel of many small detents, which was not only OK, but pleasant in many situations. With the motor energized, the knob could not be easily turned by hand at all.

I solved the problem of tracking position by using dual pots. One pot was the actual control, the other was a position sensor, with a DC voltage applied across it and the wiper fed to an A-D channel on the controlling uC. The uC could then read the electrical position of the pot, use that to memorize the position if it had been adjusted manually, and recover the pot to that position if a recall was needed. This in combination with the immunity to over-rotation let the system recalibrate by running it to one stop, then counting to the correct position. The absolute motor position was not required, nor was a rotary encoder for position.

The equipment could be set to manual, used as normal, but then be driven by the motors. The system could memorize manual settings, or go to old settings and those be tinkered, and the new settings memorized. And of course, the status could be sent/received by MIDI or whatever digital signaling channel.

[R.G.]

Let me be clear again - I think it's a good idea and you ought to go do whatever you can with it. I'm not claiming it as my idea; I believe I did as you did, thinking it up after some other person did, long ago.

Go for it. You're welcome to use any of the embellishments I described.

[gtudoran]

R.G thank you for your more then comprehensive post. I won't patent the system itself but the binary code / source code that is used to control the whole system (don't know if it's possible to do the same for the command module design but i will see about that - i will burn that bridge when i will get there).

"At the time, there were many 5.25" floppy disk drives available at very low prices, and they all included a small stepper motor, most with a double ended shaft, usually 0.25"/6.35mm diameter, which happens to be the diameter of many US-style potentiometers. It was quite natural to connect the stepper to the pot. In my protos, I used rubber tubing for the connection between motor and pot shaft, with 0.25" interior tubing being easily available. " - i've tried to find some ... damn ... no luck they were perfect for the job, but then i've found this model which is way smaller then anything i could possibly find (20 x 20 x 25 mm).

"It worked first time, of course. The motor turned the pot well. Using such a small motor meant that for 24mm pot bodies, the motor torque was low enough not to overstress the mechanical stops. For larger motors, a limiting resistor in the motor windings cut the available torque. Anything to limit motor stall torque will do that pretty well." - i'm using a specialized driver module 8825 from TI which comes also with current limiting / fault signaling / enable disable functions and can supply up to 2A with heat-sink.

This system (at least for the moment) will not allow human interaction with potentiometers - i wanted to keep it dead simple and as smaller as it can be.

"This in combination with the immunity to over-rotation let the system recalibrate by running it to one stop, then counting to the correct position. The absolute motor position was not required, nor was a rotary encoder for position. " - this is exactly as i'm doing it - at every power on, the motor will go to "zero point" that will be made using an optical element and then will go to previous stored position - as a small caution an end-stop will be implemented to be sure that no over-travel will be achieved - on the same interrupt because i can detect quite easy what was the direction of the rotation and if the max. end point was touched then i will reverse the rotation with 2 steps and stop the motor (maybe a signaling message for fault will be available)

The idea behind it was "ease of use" so you can actually have it up and running only with a screw-driver - plug and play. Indeed and ADC would be necessary if human interaction was possible or ... maybe an optical encoder, but as long as user is not allowed to change the position by hand i think that is not necessary.

Again thank you for your intervention which is more then welcomed in any case - you have more experience to share then i could do in a lifetime

Regards,
DeX

PS: I'm sorry for the bold text

[G. Hoffman]

A couple other thoughts.  First, steppers don't actually want less load - they want a matches load.  If you put a load smaller than they are designed for, they get balky, lose steps, and burn out a lot faster than they would otherwise.

Also, as far as making this a commercial product, there is already at least one very similar product on the market - the Gig Rig just came out with something along these lines - I believe they released it at NAMM.  SO, that's your competition for price point.

[gtudoran]

Gabriel it's a totally different approach as far as i could see a DC motor was involved and also it was designed for pedals as far as i could see. And i'm not competing with anyone the design will be released in an Open Source  agreement so... you can't really compete with something that is free right (for personal use but.. never the less) if there will be clients for it a commercial version will be ready available also available as an extra option to the amps that i build.

Regarding the torque requirements i don't think that is quite true... that would mean that more then 80% of the CNCs would not work at all especially the 3D printers.

Regards,
DeX

[Hatredman]

You said you system is not for pedal use but mentioned guitar amplifiers.  My question:

Is your system going to replace the pots in the amp? Doing so, will the user lose the hability to operate the pots manually?

[gtudoran]

This system will not replace the pots on the amp Hatredman. You take the knobs down, connect the system on the pot's shaft and ... that's it. The pot in the picture is just a mock-up, that pot represents the actual pot that is build in the amp. No internal modifications will be performed.
If you like to use the amp without this system, take it down, put the knob back .. and done - 5 min max.

Regards,
DeX

You said you system is not for pedal use but mentioned guitar amplifiers.  My question:

Is your system going to replace the pots in the amp? Doing so, will the user lose the hability to operate the pots manually?

[G. Hoffman]

Regarding the torque requirements i don't think that is quite true... that would mean that more then 80% of the CNCs would not work at all especially the 3D printers.

High end commercial CNC machines almost all use servo motors.  The closed loop and higher torque are much more suitable for high precision in the VMC, HMC, and lathe market, and even in the router world they are probably a better choice.   If you want super high precision, you have to come up with a function that combines the servo's rotary encoder with a linear scale. 

As for the DIY area - well, there is a lot of down time.  The torque calculations are a pain in the ass, and a lot of people get them wrong, at least the first time.  If I had it to do over, I'd pay for a commercial  machine. 


Gabriel