Pulling strings w/servos - check it out

Michael Johnstone · Post by **Michael Johnstone** » 9 Jul 2014 10:24 pm

Lotsa people have talked about pulling strings on a pedal steel with servos instead of mechanical linkages for years and the concept just seemed like it would be too out of touch compared to the feel of a mechanical linkage. But this guy seems to be onto something. He has four $20 off-the-shelf RC model airplane servos and some fairly simple electronics rigged up on a Telecaster with foot pedals like Phil Baugh's set up and it seems to have the tactile feel needed for this system to be quite playable and it could easily be done on a pedal steel.
Could this be the future of pedal steel???

Check it out here: http://imgur.com/a/a0ZlW

And a video of the guy playing here:

https://www.youtube.com/watch?v=fSQ9Dg6 ... e=youtu.be

Norm Fletcher · Post by **Norm Fletcher** » 10 Jul 2014 7:41 am

Very impressive stuff. Even with my Air Force training in servo systems, it seemed to me that this would be just an idea and wouldn't be very musical. This guy certainly showed not only that it could be done, but also that its time has come.
Now it's time to get a Carter Starter, strip it down and give this concept a go!

Paul Arntson · Post by **Paul Arntson** » 10 Jul 2014 9:52 pm

Quite a good sound! Nice find, Michael.
This guy is really on top of it.
Nice guitar player, too.

Great stuff.

Olli Haavisto · Post by **Olli Haavisto** » 11 Jul 2014 1:48 am

Hmm, 35 motors for my 8x6 U12...

Earnest Bovine · Post by **Earnest Bovine** » 11 Jul 2014 7:12 am

Yes, it looks as tho this system would require a separate physical stop (and maybe even a separate motor) for each pull. In my opinion this is no better than what we have now.

Wikipedia wrote: A servomotor is a rotary actuator that allows for precise control of angular position, velocity and acceleration.[1] It consists of a suitable motor coupled to a sensor for position feedback.

This is the same as our current system: you would still have to tune the pulls as your strings age.
In fact, it does not appear that he is using position feedback from the servo motors at all; it looks as tho he is just pulling to a stop that you tune with a screwdriver.

IMO a better sensor system would read tension, rather than position. Tuning would be much more reliable, and you would almost never have to fine-tune your pulls.
It would follow intermediate positions of the pedal.
And most importantly, for every combination of pedals, the temperament of all 12 strings could be tweaked in software for sweetest sounding chords.

Jim Palenscar · Post by **Jim Palenscar** » 11 Jul 2014 7:21 am

I'm thinking 24- 1 for each raise and 1 for each lower- the amount of change being controlled by an EPROM- custom settings available at the touch of a button. Long time musing of mine~~

Jerry Jones · Post by **Jerry Jones** » 11 Jul 2014 7:27 am

For pedal steel, that system would seem to be a complicated and unreliable replacement for the simple "pedal rod". For 6-string though, it might be a great way to escape the cable pulls, like the old MSA system had.

Earnest Bovine · Post by **Earnest Bovine** » 11 Jul 2014 7:47 am

Jim Palenscar wrote:I'm thinking 24- 1 for each raise and 1 for each lower-

12 is enough
No reason to distinguish between raises and lowers; the center (unpulled) pitch is like raises and lowers: just one more pitch (tension) to set.
In fact, the whole tuning, and all pedal functions, could change with one touch of a button. So you can't even say what is a raise or a lower.

Kevin Raymer · Post by **Kevin Raymer** » 11 Jul 2014 8:26 am

This is a thought tI've considered for some time.
Should'a know I didn''t have an original thought...

With the ready availability of the Arduino platform this is definitely doable mechanically.

Making it "musical" and making it feel right would most certainly be the challenge.

Along with this can come the options for:
quick and accurate
alternate tunings
automatic tuning
On the fly changes in CoPedent.

All that at the touch of a button from one song to the next.

Lots and LOTS of potential...

Lane Gray · Post by **Lane Gray** » 11 Jul 2014 8:31 am

It also seems you could map infinite numbers of compensators, possibly triggered by combinations of levers/pedals. If one wanted to go down that road.

Jim Palenscar · Post by **Jim Palenscar** » 11 Jul 2014 8:50 am

no stoppin us now

Earnest Bovine · Post by **Earnest Bovine** » 11 Jul 2014 9:22 am

Perfunctory Google search finds no suitable tension-sensing servos.
There are many devices that read & regulate tension, for example in textile manufacture, but nothing has the sort of fine accuracy & precision that we need, nor the fast response times.

Suppose we want pitch accuracy of 1 cent, which is 2^(1/1200)= 1.0005777895065548592967925757932...

Frequency varies as the square root of tension, so we would need tension accurate to a precision of about 1 part in 1000.

Kevin Raymer · Post by **Kevin Raymer** » 11 Jul 2014 12:40 pm

I think by using an arduino like that guy did that you could do a sensing of the pitch of the string via a PIEZO element to detect each string and not have to do have tentsion detecting servoes..

It's all done id "C", but the code would be basically ( no pun intended)

If Pitch <X> X
Then
Decrement servo position by Y

You would have to work on the code so to minimize oscillation during tuning but with the processing speed of some of those Arudino platforms it's very possible.

By detecting the pitch of each string, and use of the Arduino platform and it's myriad of various functional shields it would be a logical next step to offer a MIDI output so the Pedal steel could become a MIDI controller..

Kevin Raymer · Post by **Kevin Raymer** » 11 Jul 2014 12:44 pm

There is also an intermediate step here.

Hook the servos to the changer fingers instead of the string.

That way tuning would be done via the conventional methods currently used.

That way you have the flexibility of moving pedals and knee levers around to fit your comfort zones, but you don't have to worry about string detecting or tension detecting.

This would however facilitate change on the fly CoPedent changes.

This way on an all pull you'd need 20 servos per neck.

Michael Johnstone · Post by **Michael Johnstone** » 12 Jul 2014 3:25 pm

Damn - there are some smart guys on this forum after all......

I'm sure this idea has legs and many of these ideas about compensators,feedback circuits,variable copedants and more could be accomplished by the right person with the time,money and inclination. My only criterion is if I spill a beer on it - will it still work? And will it work in West Covina? What about that?

Kurt Kowalski · Post by **Kurt Kowalski** » 13 Jul 2014 4:23 am

Hi gang!
I have been working in this concept for a while.
I just passed my 5 th year being a paraplegic from a head on crash that put a stop to my beloved pedal steel playing.

I have a stripped down Sierra S-12 that I have been working on a servo retro fit with servos and sensors for hand and wrist sensor actuators for string changes.

I am now learning about Arduino and I am convinced that this will work.

If anybody is interested in helping and becoming part of this project. I welcome all and any help.
Just contact me.
I believe this concept could be revolutionary and help many out. Even more than just pedal steel.
Driving cars, moving wheel chairs, playing instruments, adjusting sound boards, lighting, HVAC controls, etc... The list goes on.

Are there any Arduino experts on this forum?
My best to all.
Kurt "killer" Kowalski

Donny Hinson · Post by **Donny Hinson** » 13 Jul 2014 12:17 pm

Yup, purty soon, what with 3D printing and servo technology that's coming, anyone will be able to play steel at the Paul/Buddy/Curly level in just a few weeks, and the value of those "old technology" instruments will plummet like film cameras and black and white TV's. Shucks, you'll prolly be able to get D10 p/p's for $250, and Franklins and Zums for $150.

"GOSH! You play a mechanical pedal steel? I thought all them things had been melted down and recycled into flower boxes and electric bicycles by now!"

Of course, the new technology is not without pitfalls:

"Hey Paul, how's come you were speeding up so much on that last song?"

"Well, from what I understand, it's called 'unintended acceleration', and it's a software problem, but the recall is supposed to fix it in a month or two!"

John Billings · Post by **John Billings** » 14 Jul 2014 9:50 am

I just don't see it! Is it gadgetry for gadgetry's sake? If something electrical breaks, you're done for.

Keith Hilton · Post by **Keith Hilton** » 14 Jul 2014 3:52 pm

In 2004 when I was in a electronics micro-controller class, the concept of hooking servo motors to pedals and knee levers occured to me. We were working with Basic Stamp, and then PIC. Programs must be written to control what the servo motors do. Instead of playing the instrument, the use I saw was for keeping a instrument in tune. It would be reasonable to write a program for each string's frequency, open, pulled and lowered. It would not be easy and would require a skilled programmer-and lots of mechanics. I do think something that would tune a guitar automatically would be helpful. We already have computer generated music. I am pretty sure I would not like computer generated pedal steel guitar.
Donny, if you see anyone wanting rid of a Emmons LeGrande for $150.00 be sure and call me.

Ian Worley · Post by **Ian Worley** » 14 Jul 2014 4:49 pm

Keith Hilton wrote:Programs must be written to control what the servo motors do

The guy with the Tele has already done it, at least a prototype, using an Arduino processor. His source code is here: https://github.com/deanm1278/servoBende ... Bender.ino

Will Cowell · Post by **Will Cowell** » 15 Jul 2014 4:57 am

Well I suppose Donny Hinson's post was meant to be a joke, but it comes too close to the assumption that all the naysayers have been coming up with these last 5 years or so: that somehow the rate of change of the note is not directly controlled by the player, and only by the player. Where do they get these crazy notions?

Servo technology is good enough these days that, if you gave somebody a controller and allowed them to see the output (the resulting movement) you could not tell whether there was a direct mechanical linkage. Obviously, we would not sense tension (pace Earnest Bovine) - pitch sensing is what we need.

The player controls the rate and extent of the note bend, using the electronics and the servo mech to effect that change. Brilliant for those handicapped from birth or war veterans, victims of accidents etc.

Dang it, we fly jets by wire these days and nobody turns a hair. Why should it be different for a musical instrument? There's still as much need for the performer's unique contribution as ever. That will never go away.

Peter den Hartogh · Post by **Peter den Hartogh** » 15 Jul 2014 7:50 am

Kurt Kowalski wrote:Are there any Arduino experts on this forum?
My best to all.
Kurt "killer" Kowalski

Arduino is an open-source physical computing platform based on a simple i/o board and a development environment that implements the Processing/Wiring language. Arduino can be used to develop stand-alone interactive objects or can be connected to software on your computer (e.g. Flash, Processing, MaxMSP). The boards can be assembled by hand or purchased preassembled; the open-source IDE can be downloaded for free.

For more information, see the website at: http://www.arduino.cc/ or the forums at: http://arduino.cc/forum/

Eric Moon · Post by **Eric Moon** » 15 Dec 2021 8:41 am

Sorry to unearth this zombie thread, but I wanted to reference the info above...

Since I moved a bit too slow, and just missed out on the PS210 for sale last week (I have a hard time bringing myself to spend money...) I am back to looking at options for experimenting with copedants.

I found these fairly inexpensive linear servos. The first selection under the 'stroke' menu is what I'm thinking of.

https://www.ebay.com/itm/313738433445

They are pretty fast (150mm/sec) which should feel pretty immediate over the throw of 10mm. Going from full lower to full raise will take about 150ms, which will seem sluggish, but the average 1 or 2 semi-tone pull should be 30-60ms, which shouldn't be horrible (though it's something I might have to consider upgrading at some point.)

But they can only push about 1.5 lbs.

My thought is to use a single changer finger, on a knife edge like the ps210, and counter-balance it with a compressible return spring. The servos would mount under the nut of the guitar, to keep noise away from the pickup. I would need a mechanism to accurately adjust the return spring to counter-balance string tension. Not sure whether getting that balance down to the 1 lb. range is going to be problematic or not. I also do not know how accurate the positioning of the servo is going to be, but if the outputs of the arduino are 1024 steps like the inputs are, and the servo can see all of it, that would give precision in the 1/2 cent range, assuming a fourth of tuning range (if I'm doing the math correctly....) Hopefully that's precise enough.

Trying to analyze string tension or pitch seems needlessly complicated. Like any other steel, you should be able to get repeatable pitch by moving the changer finger to a specific position. If you change a string gauge, or the basic tuning of a string, these values would change, obviously. However you could store all the values in memory, so you'd only have to do it once.

These actuators are 14mm wide, so you could stagger them in two rows, with short rods for the second row... I'm looking at an 11/32 string spacing, same as my supro and the ps210. Seems like that could fit.

Instructables has some basic arduino code here:
https://www.instructables.com/Control-a ... h-Arduino/

As a keyboardist foremost, I have three broken sustain pedals already to experiment with. I like the idea of the hall effect sensors the tele guy used. From what I can tell, they worked well for him.

I'm planning to buy one of these actuators, and rig up a single string test bed. I'm curious to know if anyone sees why this couldn't work, before I get started. Seems like there's some significant expertise here in this forum. Since I've been playing a modified multi-kord, I don't have familiarity with the tensions and stresses in a normal pedal steel...

Thanks in advance!
=e

Ian Worley · Post by **Ian Worley** » 15 Dec 2021 10:50 am

There have been a ton of discussions here on this topic over the years, in case you haven't read them. Lots of different perspectives and insights, ideas on different approaches with linear vs rotary servos, steppers, solenoids, different controller schemes, ...
https://www.google.com/search?q=stepper ... rforum.com
https://www.google.com/search?q=servo+s ... rforum.com.

Forumite Karlis Abolins built a prototype device last year that he demonstrated in a youtube video, there is a thread about it here: https://bb.steelguitarforum.com/viewtop ... 16#2968116.

I believe you will have an uphill battle with only the 1.5lbs of force you're proposing using linear servos, you can create a mechanical advantage via leverage but that will of of course increase the necessary throw distance. Will be interesting to see what you come up with, best of luck, have fun!

The Steel Guitar Forum

Pulling strings w/servos - check it out

Pulling strings w/servos - check it out

Servos