Project Summary:
I want to build a mechanism that can be attached to a guitar to endlessly agitate and vibrate the strings rather than plucking them. This form of sound generation is used on an instrument known as a hurdy gurdy, which uses a spinning (via handcrank) wheel in a similar fashion to a violin bow. However, unlike a violin bow, which is finite in length, the wheel can continuously spin, creating a droning sound that is very unique to the hurdy gurdy. I want to create a small device which mimics this functionality for the guitar.
While the hurdy gurdy has lost its mainstream appeal for centuries, it has been occassionally used by indie bands that love obscure instruments such as the Decemberists or the Arcade Fire. Here are 2 examples of the hurdy gurdy being used in the Decemberist's song "Sons and Daughters"
and the Arcade Fire's "Keep The Car Running"
Introduction:
To create this device, certain criteria must be met.
1. First off, the device must be able to vibrate the strings of the guitar, with a minimal reduction in intensity over time (or in other words, work). The device should ideally be able to run for at least 10 minutes (possibly even longer, depending on the composition the instrument is playing) without requiring adjustment.
2. The device must not interfere with the workings of the guitar. One of the big disadvantages of a hurdy gurdy is that it is incredibly difficult to keep in tune. Ideally, the device should not knock the strings out of tune, or otherwise interfere with the guitar's ability to produce sound. While the tonal quality of the instrument will likely change due to the alternative playing method, the device should not interfere with either the electronics for an electric guitar or the sound chamber of an acoustic guitar.
Background:
There are a variety of already established methods for accomplishing something similar to this device, however, each comes with their own drawbacks.
1, First off, there is the actual Hurdy Gurdy. However, Hurdy Gurdies are both hard to find and generally, quite expensive. Meanwhile, guitars are much more widespread. Since many people already own guitars, being able to buy a relatively inexpensive device to simulate a hurdy gurdy sound would generally be considered a much more enticing option unless they have an upcoming gig at Ye Olde Renaissance Faire.
A picture of a hurdy gurdy:
A video of the hurdy gurdy in action:
2. Another option already on the market is an Ebow ('Electronic' or 'Energy' Bow), which, when held close to a guitar, uses an electromagnetic field to vibrate guitar strings. However, the sounds produced in this manner (in my humble opinion) lack any sort of interesting timbre, sounding somewhat 'thin'. I believe a mechanical, rather than electromagnetic, approach, will result in more interesting textures
A picture of an Ebow:
A video of the Ebow in action:
3. Lastly, there is a guitar built by Moog which has a technology similar to the Ebow integrated into the guitar. However, the guitar is priced well into the thousands, which makes it quite an expensive option.
A video of the Moog guitar in action:
Internet Sources:
http://en.wikipedia.org/wiki/Hurdy_gurdy
http://en.wikipedia.org/wiki/Ebow
http://www.moogmusic.com/moogguitar/
Intellectual Property:
The HurdyGurdy, built by Godfried-Willem Raes. From the site (http://www.logosfoundation.org/instrum_gwr/hurdygurdy.html [be careful if you choose to navigate the site...you are only a click or 2 away from naked people dancing around to bizzare music. I have no idea why.])
"This robot is a bowed bass instrument with two strings of equal length, covering a range of nearly four octaves. The construction of the bow mechanism is a further development of our first designs in this direction, implemented in <Flex>, our singing saw. The bowing speed, and consequently the loudness, can be controlled as well as the direction of rotation. Bow pressure can also be controlled independently for both strings. The time plot shown further below gives the details of the control possibilities. Rosin is continuously applied to the bow material through a rosin holder wherein the bottom wheel of the bow mechanism rotates. The frets are realized with strong electromagnets equipped with tangents. They are moveable, such that the instrument can be prepared to play in different tuning systems, including just intonation. The resonators for the strings are constructed from thin stainless steel pots, welded on a heart shaped sound board. The metallic and harsh sound the instrument produces when bowed was intended. Softer sounding, almost etheric string sound, including all flageolets, can also be produced using the e-drive mechanism. This machine is fully programmable and can work under midi control."
The instrument can be heard here:
http://www.logosfoundation.org/mp3/lpd019/religionszwang.mp3
As you can see (hear?) it is very droning, metallic, and German. I will do my best to create something that has the first trait, and not so much the other two.
Requirements:
Here are my project
Requirements:
1. The device must be able to attach to a variety of guitars with different body types without any kind of slipping
2. The device must be able to run successfully and continuously for a minimum of 10 minutes without need for readjustment
3. The device must be non-intrusive to the instrument it is mounted on. In other words, the instrument should work exactly the same after the product is removed as right before the product was attached.
Constraints
1. The device should be relatively inexpensive to manufacture.
2. The device must fit on the top of the body of the guitar
Evaluation Criteria
1. Does it meet the requirements and constraints stated above?
2. Is the tonal quality of the sound interesting and distinctive? (Tested using a harmonic analyzer or the opinion of a music expert)
First Prototype:
Brainstorms:
For my first prototype, I will construct a simple wheel with a nylon belt wrapped around it. The wheel is fitted to the guitar by a mount that fits snugly between the two pickups and wraps around the side of the guitar. Unlike the final product, the device will only come into contact with 1 string of the guitar. The prototype will be used to make sure that a nylon belt with rosin, when driven adequately fast, will produce sound when in contact with the guitar string. Also, this prototype uses a handcrank, while the final product will be motorized.
First Prototype: (here are my pictures, videos, schematics, drawings, brainstorm notes, parts list, and gatorade receipts)
First Test and evaluation:
Results:
3 different materials were tested for use at the contact point between the wheel and the string: A strip of nylon, a strip of rubber, and the plastic wheel itself. Of the 3, the tip of the wheel performed by far the best at exciting the string. This revealed that the best contact material for the apparatus was one that is quite rigid and non deformable.
Rubber bands were first used as a quick easy way to fix the apparatus to the guitar, however they proved to be inadequate, the device tended to move too much when cranked, causing it to push the string around, causing fluctuations in pitch, or lose contact with the string all together.
The second method of fixation, Velcro strips, proved much more effective. When correctly placed the mechanism could excite the string without slipping. However, the hand crank was still quite unwieldy and difficult to spin smoothly for multiple rotations. This meant that motorization of future prototypes would be critical. By using a motor the torque would be more evenly applied in both magnitude and direction. Also, the hand crank was quite noisy, overpowering the sound of the string.
Second Prototype:
Brainstorms:
Here are the ideas I've come up with to build and test my solutions:
The two key things I took away from the first prototype was that the way the device was held was critical to its usefulness, and that a hand crank was quite unwieldy, thus I drafted a prototype that took these two things into consideration
Second Prototype:
Second Test and evaluation:
Having pinpointed what worked and didn’t work about the first prototype, the second was assembled. This prototype used a small dc motor to drive a small steel wheel. When the prototype was first tested, it didn’t work well because the smoother steel surface of the wheel was not as good at exciting the string as the scratchier plastic from the previous prototype. However, this was quickly fixable using some violin rosin, which gave the wheel more traction.
While I experimented with a variety of ways to mount it, I found the best use musically for the apparatus was to actually hold it in the right hand. That way, it could be repositioned along different frets, while the left hand fretted strings of its own, creating a rich blend of textures. By changing the spacing between where the left hand was fretting and where the apparatus was held against the guitar, different harmonics could be excited. The device could excite 1 or 2 strings at a time, and occasionally 3 or more in specific situations (depending on the geometry of how the strings were depressed against the frets)
A video of the 2nd prototype in action:
Movie on 2010-12-09 at 12.37.mov
As you can hear, it is pretty quiet when played accousticly (the sound of the motor almost overpowers it). The apparatus really shines on the electric though, as shown in this music clip:
Results:
The results of the second prototype were very exciting. The sounds produced by the device made the instrument sound much more like a hurdy gurdy than a guitar. I would go so far to say that it actually sounded prettier than a hurdy gurdy but that is just my subjective opinion.
While my original plan had been to mount the device to the machine, I believe holding it in your hand allows for more options and flexibility while playing. The disadvantage, however, is that you have one less hand to otherwise manipulate the guitar, meaning it is difficult to pluck along with the droning, though through the guitar techniques of hammering on and pulling off this is still somewhat possible. An example of this can be heard in the above track.
Another one of my criteria that this device meets is that it is very much unobtrusive to the guitar. Someone could pick up the device for an epic drone jam, and put it down to immediately launch into the opening licks of Johnny B. Goode. The only strain that the device places on the instrument is that it can wear down the strings a bit, however this will happen pretty much no matter what method you use to play the guitar.
Ultimately the device behaves similarly to an ebow, though it produces much denser sonic textures (the ebow generally doesn’t excite a lot other than the fundamental, which creates a boring, thin tone). Also this device could probably be manufactured and sold at a fraction of the cost of an ebow.
Things to be improved upon:
When played accousticly, the instrument is very quiet. It is so quiet that the noise of the motor definitely interferes with the overall sound. Ideally, the guitar would be louder and the motor would be quieter. However, this problem all but vanishes when the device is played with an electric, which is probably how I personally would use the device in the first place, as I can then take advantage of adding different effects to further process the noise.
Also, the device tended to excite higher harmonics to the point that they became a bit piercing.