Guitar Strings and Inductor Coils

[dnb1007]

I need a little help with a project I'm working on. Here's the gist of it; I am taking two coils, hooking one into the input of an amplifier circuit and the other to the output. Both coils will be over an electric guitar string. The guitar string when vibrated will have a changing induction field which results in an induced voltage in the input coil, which is then amplified and goes through the output coil which creates a stronger induction field increasing the vibrations of the guitar string. The result is a feedback loop causing the string to make a volume swell. The circuit will also pickup background induction fields from the pickups of the guitar so the string doesn't have to be initially vibrating.

Anyways, I've developed the amplifier portion of the circuit using the data sheet for the LM386, but I don't know what inductance I should have for the coils. I'm guessing I'll need some way to measure the changing flux of the guitar strings when they vibrate, but I'm at a loss there as well.

I don't have my circuit diagram with me at the moment, but if it is helpful I can attach it later.

 

[davenn]

hi there
welcome

I'm really trying to understand the purpose/point of this?
Feedback is generally an unwanted thing and when it is there can sometimes be difficult to control

Dave

 

[sophiecentaur]

hi there
welcome

I'm really trying to understand the purpose/point of this?
Feedback is generally an unwanted thing and when it is there can sometimes be difficult to control

Dave

Howl round (controlled) is frequently used to produce a very long sustain, which can change the sound as it progresses.
He's trying to introduce some electro-mechanical feedback / howl round. People usually do that by sticking a loudspeaker near the guitar, don't they? But it's a fun idea and I can see the attraction. Where were you thinking of having this feedback coil? You want it fairly close coupled to the strings so it would go near one of the pickup coils? It would need to be near the middle of the string (not near the bridge) or the coupling at the fundamental frequencies would be small, I think (many many Watts of amplifier power needed).
Quality would possibly not be needed so you could suppress HF in the feedback to keep things more tamed.
I wonder whether magnetic linking via the steel strings to the pickups could be a problem. Also, would the LM386 produce enough power?

 

[thankz]

http://www.fernandesguitars.com/sustainer-kits.html

I wouldn't attempt to build this from scratch, best of luck.

 

[dnb1007]

Right, the purpose of this project is to create a gadget that will essentially "pluck" the guitar string through feedback. Obviously, this will have a very different effect from plucking because the amplitude of the string's vibrations will swell. I'm not worried about the feedback getting out of hand because the op amp will eventually become saturated.

There is something already operating like this called an Ebow, so you're welcome to google it and see why I'm interested in pursuing this. I'm adding a few extra elements to my project to make it a bit more interesting than the ebow however.

I was planning on having the input coil over the pickups, and the output coil closer to the middle of the string (as long as frets can still be used). But they wouldn't be very far apart from each other. I figured the LM386 was a good starting place in terms of power, if the ebow can operate off of batteries and accomplish this.

Sadly, I don't really have much experience with inductors and I'm not sure how to go about finding the right inductance for each coil. From Faraday's law, I'd expect I would have to know something about the change in flux for each coil to keep the induced voltage within a good range for the LM386, but from here I'm stumped.

And I'm very interested in pursuing this project rather than a kit. It'd be more versatile than the Fernandes Sustainer and give me a little more pride in accomplishment.

 

[davenn]

we line and learn :)Dave

 

[Averagesupernova]

I would model the output coil after a speaker since the 386 is designed for that. I would also do some experimenting to find out how different coils sound.

 

[sophiecentaur]

I would model the output coil after a speaker since the 386 is designed for that. I would also do some experimenting to find out how different coils sound.

The coupling to the string will be very inefficient, with the coil core situated a practicable distance away from the string. You have, effectively, a very inefficient loudspeaker design here. That means you will almost certainly need lots of Watts of drive. From what I can see, the 386 will give around 0.5W so I wonder how well it suits the situation. Fine for testing, perhaps but you may have to find something more beefy.

To find the power needed, I guess you would need to work out the displacement needed on the string (probably sub-mm would be enough) and the B field needed to cause it, given the tension and the amount of steel string in the field. You then need to find the number of Amp Turns on a solenoid with a core to produce that field at around 10mm from the string. You can have many more turns than a loudspeaker voice coil because the coil doesn't need to be light enough to move. There are dozens of Google links to information about the Field from an electromagnet but I can't find one to tell you what force you can expect on a piece of steel. It strikes me that you are talking Mechanical Actuators rather than Loudspeakers - 12V and many Amps, even for a purpose built solenoid actuator.

Imo, this calls for a set of experiments, involving coils on an iron core and a DC source with Ammeter. See what actual current you need, to deflect a guitar string. See if you can actually 'pluck' the string magnetically with a DC pulse. The guitar pickup would let you hear the result. I still have reservations that you are very likely to get straight magnetic coupling from your coil to the pickup, providing a feedback path which is independent of string movement. This would not affect the plucking experiment but would be relevant for the final system. That's just an intuitive reaction, though; you'd have to measure it; an easy exercise.

 

[dnb1007]

I still have reservations that you are very likely to get straight magnetic coupling from your coil to the pickup, providing a feedback path which is independent of string movement.

Thinking about it, if either coil is right above the pickups this might be an issue. All I know is that the Ebow doesn't seem to have this problem and that keeps giving me hope.

You gave me an idea with the DC pulse testing. If I take a coil with known values for N, length, R, and μ then hook it up to a power supply and increase the voltage with each DC pulse until the string deflection is strong enough, this would give me an approximate B field because $$B=μ\frac{NV}{lR}$$
Then I can set V to be something like 8.2 Volts and mess with the other parameters to make my own coil that provides the same B field with the voltage I need.
The only thing that is unsettling is this isn't really setting the inductance, because cross sectional area is arbitrary in this method. And it would seem that this plays a large role in setting the flux which is apart of Faraday's Law. Faraday's Law being what this whole gadget is centered on.

 

[sophiecentaur]

Glad the DC test appeals to you. It would be very easy to do and would give you an idea of what ball park you are in. If it weren't for the fact that you have quoted a piece of kit that does it, I would be very skeptical - but what do I know? Intuition can lead us astray. When you think of the very weak coupling between the strings and the pick up heads, you are dealing with a similar relationship between source power and power out.
The Inductance of your coil is a 'dependent variable'. It will depend upon all the other primary requirements; you can't start usefully with a value out of your head. Clearly, it would be possible to make the strings move, with enough Amp Turns. This could, however, call for a massive inductance which could only be driven from a high voltage or loads of current (high Power, however you do it), in order to get the required deflection. There is no way that the amplifier power can all get to the strings.
I wonder whether Ebow doesn't actually have the problem of magnetic coupling. It may not matter.
There could be a massive frequency / power slope in this system, I think. Both the Inductance and the dynamics of the strings have the same response slope, I think.

 

[dnb1007]

Thank you for the help so far. If I am eventually able to construct a driver coil that can deflect the string with a reasonable amount of power, should my input coil be the same as the driver coil?

 

[sophiecentaur]

The idea for your final driver coil is to produce a high enough B Field to move the string at up to several kHz. It would need a low enough inductance to allow a suitable voltage amplifier (say 12V) to drive the required AC (this time) current. The amplifier will be putting most of its output power into heating the coil just to produce enough B field where you want it - it will be a necessarily inefficient system. unlike a normal loudspeaker.
I suggest you do this one step at a time. First find out the Amp Turns needed and then decide how to get that at your highest AF.

 

[meBigGuy]

I just see a linear coil along the string driven by its own pickup. Seems like one of the many magnetic experts here should be able to determine how much magnetic force is required to vibrate a guitar string enough for the pickups to hear it.

The ebow pickup would need to be somehow isolated from the drive coil. The phase of the vibrations in the string relative to the the magnetic field from the drive coil could probably be compensated for (kind of like echo-cancellation in a phone line).

As for direct coupling to the guitar pickups, that would just add to the effects. I don't see it as a negative. The phase of the vibration down the string and any magnetic field at the pickups (from the driver) would create interesting phasing effects as you moved.

You might need an AGC (automatic gain control) to limit the feedback such that it remained linear, or maybe a square wave drive would produce richer sound? (no idea)

They also have a mode that reverses the drive causing (they say) cancellation of the fundamental but allows rich harmonics. I don't quite understand how that would work.

 

[dnb1007]

The ebow pickup would need to be somehow isolated from the drive coil. The phase of the vibrations in the string relative to the the magnetic field from the drive coil could probably be compensated for (kind of like echo-cancellation in a phone line).

Since they are going to be about an inch away from each other, I don't think the pickup coil will interfere with the drive coil, unless there is some effect from the harmonics of the string causing them to be out of phase.

As for direct coupling to the guitar pickups, that would just add to the effects. I don't see it as a negative. The phase of the vibration down the string and any magnetic field at the pickups (from the driver) would create interesting phasing effects as you moved.

If direct coupling happens with the guitar pickups, I think I'll just have to move the input coils slightly off center from the guitar pickups. When the gadget is finished though, it will just be stationary due to other things I'm trying to accomplish with this project.

You might need an AGC (automatic gain control) to limit the feedback such that it remained linear, or maybe a square wave drive would produce richer sound? (no idea)

Thank you, if I'm not happy with the shape of the swell, I'll definitely keep this in mind (Assuming I get there).

They also have a mode that reverses the drive causing (they say) cancellation of the fundamental but allows rich harmonics. I don't quite understand how that would work.

I think this would work by deflecting the string under the driver coil out of phase with the input coil making the first fundamental much weaker but allowing the higher harmonics with shorter wavelengths to be more prevalent, but that's just a guess. It'd be interesting to try and incorporate that too.

 

[meBigGuy]

Since they are going to be about an inch away from each other, I don't think the pickup coil will interfere with the drive coil, unless there is some effect from the harmonics of the string causing them to be out of phase.

a.) The waves propagates slowly on the string (driver to bow pickup), and are instantaneous from the bow driver to the bow pickup. 1 inch on a guitar string is a lot of phase. That phase shift is also part of your feedback loop as is the instantaneous crosstalk.
b,) The bow pickups are sensitive and the field fom its driver will likely affect them, even at 1 inch. Remember, you want the feedeback to be through the strings. It's hard to imagine enough magnetic drive to create audible string vibration that is not strong enough to affect the magnetic pickups.

If direct coupling happens with the guitar pickups, I think I'll just have to move the input coils slightly off center from the guitar pickups. When the gadget is finished though, it will just be stationary due to other things I'm trying to accomplish with this project.

The coupling to the guitar pickups will just be another phasing effect.

Thank you, if I'm not happy with the shape of the swell, I'll definitely keep this in mind (Assuming I get there).

Not the shape of the swell, the shape of the drive signal. Sine vs square wave. Creating your waveform by clipping is limiting. Maybe a sine or triangle drive will sound better. I don't know at this point.

You are going to learn a lot doing this. I expect some serious mechanical and electrical engineering has occurred in that ebow. Even the shape and look are good designs.

Should the pickup signals be turned into pulse streams? Maybe PWM provides the richest sound. Certainly the best switching efficiency. Maybe the most efficient is to feed the audio into the reference of a switching converter where the driver is the coil. Lots of places to go with this.

 

[sophiecentaur]

@meBigGuy
Some very good points there.
PWM could be expected to give a linear response and could involve a more efficient drive. I would say that would be for a future stage of development, though. What is needed here is, imp, a good old fashioned power amp to drive the feedback coil.
My initial opinion was to agree with your point b but I have now realized that all the feedback needs to provide is enough power to compensate for the energy lost per cycle, in order to achieve infinite sustain time. The Q of a string on a guitar must be at least 100 (much more in many cases) so the drive power may not need to be all that excessive.

The phase shifts are difficult to evaluate. The magnetic coupling between drive and pickup could be quite high (a significant component) via the steel string and that will involve a 90 degree phase shift between drive current and pickup volts (as in a transformer, I think). The mechanical phase shift will involve a further 90 degrees, I think - plus anything due to propagation along the string. It would be fairly simple to look at the magnetic coupling by immobilising the string. This implies more experimentation, of course, which is another reason to get down and start on some practical measurements. I don't think that there is any worthwhile simulation to be done here as there are so many unknowns to plug in.

 

[dnb1007]

Thanks again for all the help! Sounds like I just need to find a DC power supply for determining the coil properties then go for it and troubleshoot as I go. You guys have given me a lot to think about if I encounter problems with the current design or if I want a more refined design in future.

 

[sophiecentaur]

One step at a time. You will learn such a lot as you need to.

 

[MalcolmB]

FYI: - http://www.ebow.com/home.php

Cheers

Malc

 

[sophiecentaur]

One step at a time. You will learn such a lot as you need to.

I meant "as the need arises". That post could be misconstrued! Sorry.

 

[meBigGuy]

I'd buy one and take it apart. A friend of mine has one, but I don't think he will let me take it apart.

 

[dnb1007]

Well, if I had the spare money to buy one I probably would, but I'm also just interested in finally doing something from start to finish with more intimate knowledge of what I'm doing. Opening one up to see what's inside seems like it would take away from that.