I need to make a thin wire vibrate
March 10, 2025 2:49 PM Subscribe
I'm trying to reproducing Alvin Lucier's "Long Thin Wire" setup (related):
I made a prototype work! With an amp putting out 40 watts, but the signal picked up by the contact mics was so thin that I had to turn their gain up a problematic amount; the signal to noise ratio was poor. So:
1) would a more powerful amp (e.g. 100 watts) make the vibrational excursions bigger/more perceptible?
2) Would short-circuit protection baked into the amp prevent use of just a wire like this?
3) would I need to breadboard something to emulate speaker ohm ratings (e.g. 4, 8 ohms), between the amp and the piano wire?
4) am I missing some other, obvious path to making wire vibrate vigorously?
- sine wave at line level into power amplifier, with speaker terminal A connected to end A of a ferreous wire of, say, 2 meters, terminal B to wire end B.
- That wire should vibrate, reproducing the sine wave that is being played into the amp.
- It will run between poles of a strong magnet, creating sonically interesting artifacts
- Practical questions below the fold.
I made a prototype work! With an amp putting out 40 watts, but the signal picked up by the contact mics was so thin that I had to turn their gain up a problematic amount; the signal to noise ratio was poor. So:
1) would a more powerful amp (e.g. 100 watts) make the vibrational excursions bigger/more perceptible?
2) Would short-circuit protection baked into the amp prevent use of just a wire like this?
3) would I need to breadboard something to emulate speaker ohm ratings (e.g. 4, 8 ohms), between the amp and the piano wire?
4) am I missing some other, obvious path to making wire vibrate vigorously?
With an amp putting out 40 watts,
Some calculations around this:
- those 40W are specified into a nominal speaker load, say 4 Ohm. As power is current squared times resistance, P=I²R, this means the current through the speaker at that output level is sqrt(10)A, a bit over 3A.
- the force on the wire (in this setup) is linear with the current and the magnet's field strength.
- the amplifier will likely limit its output current to what the power stage can handle. Let's take this as 50% over nominal, so 4.5A.
- the resistance of even a long ferrous wire is way below 4 Ohm, but this is irrelevant to the force on the wire from the magnet as that's linear with the current. With a low resistance like here you don't need many volts to get at the maximum current the amp can deliver. Basically, how many watts the amp can put out is not the determining factor, the maximum current is. One that is spec'ed as able to deliver 40W into 2 Ohms (or even less) would do better here. Whether more watts will help depends on the resistance it's spec'ed against: 80W into 8 Ohm will drive just as many amps through the wire as 40W into 4 Ohm (as per the formula above)
- your amp may object to a load resistance as low as that wire is, but putting a real 4 Ohm resistor in series with it will only minimally alter the current through the wire.
Basically, if you need the wire to vibrate more strongly you need a higher magnetic field strength (stronger magnet and/or a smaller gap between the poles where the wire runs), and/or a higher current through the wire.
posted by Stoneshop at 3:47 PM on March 10 [2 favorites]
Some calculations around this:
- those 40W are specified into a nominal speaker load, say 4 Ohm. As power is current squared times resistance, P=I²R, this means the current through the speaker at that output level is sqrt(10)A, a bit over 3A.
- the force on the wire (in this setup) is linear with the current and the magnet's field strength.
- the amplifier will likely limit its output current to what the power stage can handle. Let's take this as 50% over nominal, so 4.5A.
- the resistance of even a long ferrous wire is way below 4 Ohm, but this is irrelevant to the force on the wire from the magnet as that's linear with the current. With a low resistance like here you don't need many volts to get at the maximum current the amp can deliver. Basically, how many watts the amp can put out is not the determining factor, the maximum current is. One that is spec'ed as able to deliver 40W into 2 Ohms (or even less) would do better here. Whether more watts will help depends on the resistance it's spec'ed against: 80W into 8 Ohm will drive just as many amps through the wire as 40W into 4 Ohm (as per the formula above)
- your amp may object to a load resistance as low as that wire is, but putting a real 4 Ohm resistor in series with it will only minimally alter the current through the wire.
Basically, if you need the wire to vibrate more strongly you need a higher magnetic field strength (stronger magnet and/or a smaller gap between the poles where the wire runs), and/or a higher current through the wire.
posted by Stoneshop at 3:47 PM on March 10 [2 favorites]
would a more powerful amp (e.g. 100 watts) make the vibrational excursions bigger/more perceptible?
Yes. Will it be enough to make a significant difference in the S/N ratio of the end result? Dunno.
2) Would short-circuit protection baked into the amp prevent use of just a wire like this?
It shouldn't. Speakers are basically a length of long very thin wire arranged in a coil. You're not actually short-circuiting anything with this setup.
Plus the usual short circuit protection for amplifiers is it blows a fuse or trips a breaker or does some more complex stuff but the end result regardless is almost always no sound at all rather than weak sound.
BUT see below re: wire gauge.
would I need to breadboard something to emulate speaker ohm ratings (e.g. 4, 8 ohms), between the amp and the piano wire?
Maaaaybe? Although it seems like a bit overkill? First thing I would do is look at something like this wire resistance chart and this chart and calculator, see if you can approximate the gauge of the wire you're using, figure out the resistance of the length you're using.
As you can see from the chart, (and as Stoneshop points out), your resistance is likely very very low, in the range of milli-ohms if you're using like 10 foot of 16 gauge wire, which might not make your amp happy. A high power 4 or 8 ohm resistor in line with one end on the wire should be all you would need to eliminate that potential problem.
signal picked up by the contact mics was so thin that I had to turn their gain up a problematic amount; the signal to noise ratio was poor
Are you running your mics through an audio mixer or external preamp? People often don't grasp that mics put out very very very low voltage, and you need an initial gain stage circuit to bump that up so the rest of the audio system can further amplify the signal. If you're just, like, plugging the contact mics into a guitar amplifier or something you might have an input impedance and gain mismatch, so your output set up is more of your problem, rather than the input from the vibrating wire to the contact microphones.
posted by soundguy99 at 4:08 PM on March 10 [2 favorites]
Yes. Will it be enough to make a significant difference in the S/N ratio of the end result? Dunno.
2) Would short-circuit protection baked into the amp prevent use of just a wire like this?
It shouldn't. Speakers are basically a length of long very thin wire arranged in a coil. You're not actually short-circuiting anything with this setup.
Plus the usual short circuit protection for amplifiers is it blows a fuse or trips a breaker or does some more complex stuff but the end result regardless is almost always no sound at all rather than weak sound.
BUT see below re: wire gauge.
would I need to breadboard something to emulate speaker ohm ratings (e.g. 4, 8 ohms), between the amp and the piano wire?
Maaaaybe? Although it seems like a bit overkill? First thing I would do is look at something like this wire resistance chart and this chart and calculator, see if you can approximate the gauge of the wire you're using, figure out the resistance of the length you're using.
As you can see from the chart, (and as Stoneshop points out), your resistance is likely very very low, in the range of milli-ohms if you're using like 10 foot of 16 gauge wire, which might not make your amp happy. A high power 4 or 8 ohm resistor in line with one end on the wire should be all you would need to eliminate that potential problem.
signal picked up by the contact mics was so thin that I had to turn their gain up a problematic amount; the signal to noise ratio was poor
Are you running your mics through an audio mixer or external preamp? People often don't grasp that mics put out very very very low voltage, and you need an initial gain stage circuit to bump that up so the rest of the audio system can further amplify the signal. If you're just, like, plugging the contact mics into a guitar amplifier or something you might have an input impedance and gain mismatch, so your output set up is more of your problem, rather than the input from the vibrating wire to the contact microphones.
posted by soundguy99 at 4:08 PM on March 10 [2 favorites]
Response by poster: One contact mic is active (48V), one passive, both going into a MixPre 6--that was where I was boosting gain for each.
posted by working_objects at 4:13 PM on March 10 [1 favorite]
posted by working_objects at 4:13 PM on March 10 [1 favorite]
Can you tell us a bit more about your magnet setup?
I might throw some bigger magnets at your problem. Though, do note, big big magnets are dangerous.
posted by gregr at 5:35 PM on March 10
I might throw some bigger magnets at your problem. Though, do note, big big magnets are dangerous.
posted by gregr at 5:35 PM on March 10
Response by poster: gregr: I went with this one. I now have 2 matched magnets from inside a spinning hard drive, which I'll try attached to that lab frame.
It's a matter of some suspense how close I can bring those two before they violently reunite; they're only adhering to the variable gap screws by the factory-attached neodymium magnets in the screw mechanisms.
So far, it seems like things to tweak before I try a higher-wattage amp are:
1) Size of magnet gap around the wire
2) Strength of the magnets
3) Fixture of mics to "pickups"--each end is a brass screw as a bridge, screwed into a small, hollow wooden box.
4) Tautness of wire; I don't have a means of measuring this, unfortunately, but I think there's room to make it both tighter and slightly looser.
5) Mayyyybe how fast the pickups are to the surface they sit on? In version 1, they were just sat on road cases held down by the wire tension.
In any case, thanks, all, for these responses. I think I have enough for attempt #2.
posted by working_objects at 6:00 PM on March 10 [2 favorites]
It's a matter of some suspense how close I can bring those two before they violently reunite; they're only adhering to the variable gap screws by the factory-attached neodymium magnets in the screw mechanisms.
So far, it seems like things to tweak before I try a higher-wattage amp are:
1) Size of magnet gap around the wire
2) Strength of the magnets
3) Fixture of mics to "pickups"--each end is a brass screw as a bridge, screwed into a small, hollow wooden box.
4) Tautness of wire; I don't have a means of measuring this, unfortunately, but I think there's room to make it both tighter and slightly looser.
5) Mayyyybe how fast the pickups are to the surface they sit on? In version 1, they were just sat on road cases held down by the wire tension.
In any case, thanks, all, for these responses. I think I have enough for attempt #2.
posted by working_objects at 6:00 PM on March 10 [2 favorites]
Another way to drive more current through the wire without actually getting a higher-powered amp is to put a down-transformer between the amp and the wire.
As an example, a mains transformer with a 10:1 winding ratio will take your 110V AC mains and turn it into 11V AC. Drawing 1A from the output will take 0.1A at the input: the same power but at more amps and less volts. Which is what you want for your wire.
If you can get hold of a transformer that can handle 40W or more, and has an input to output voltage ratio of between 5:1 and 20:1 you might want to try if that improves things. Amp output connected to the transformer input, wire connected between the transformer output terminals. If the transformer isn't that beefy (even a doorbell transformer would do for a test), just keep the amp at a lower volume and check that the transformer is not getting excessively hot.
posted by Stoneshop at 12:54 AM on March 11 [1 favorite]
As an example, a mains transformer with a 10:1 winding ratio will take your 110V AC mains and turn it into 11V AC. Drawing 1A from the output will take 0.1A at the input: the same power but at more amps and less volts. Which is what you want for your wire.
If you can get hold of a transformer that can handle 40W or more, and has an input to output voltage ratio of between 5:1 and 20:1 you might want to try if that improves things. Amp output connected to the transformer input, wire connected between the transformer output terminals. If the transformer isn't that beefy (even a doorbell transformer would do for a test), just keep the amp at a lower volume and check that the transformer is not getting excessively hot.
posted by Stoneshop at 12:54 AM on March 11 [1 favorite]
As you can see from the chart, (and as Stoneshop points out), your resistance is likely very very low, in the range of milli-ohms if you're using like 10 foot of 16 gauge wire, which might not make your amp happy. A high power 4 or 8 ohm resistor in line with one end on the wire should be all you would need to eliminate that potential problem.
Putting a resistor in series with your wire is going to result in almost all the amplifier's output power being dissipated in the resistor, which is wasteful. You want it dissipated in the wire.
As Stoneshop says, the way to get this done is with a transformer. The impedance seen at a transformer's primary terminals is the impedance connected across the secondary terminals multiplied by the square of the turns ratio.
Exactly what the impedance of your long wire is will depend on what metal it's made of, how thick it is and how long it is. There are formulae available to work this stuff out, but the easiest way will just be to measure your wire with a multimeter set on its lowest resistance range. This will get you its DC resistance, which for two metres of steel wire at musical frequencies I would expect to be the dominant component of its impedance. If you don't have a multimeter, I think 0.1Ω would probably be a reasonable guess.
So to match that load to an amplifier output that's built to drive the 4Ω - 16Ω impedances that are typical for loudspeakers, you'd want a transformer with a turns ratio of √(8Ω / 0.1Ω) = about 9.
The cheapest off-the-shelf transformer capable of handling power of the order of 40W will be a mains power supply type, and the easiest way to work out the turns ratio for these is as the ratio between the transformer's designed supply voltage and its output voltage. So a transformer built to step 110V down to 12V would be about right. As for transformer power rating: assuming that the power rating you want is about equal to the power you intend to deliver to your wire is probably a safe bet. So for a transformer with a 12V output, you'd want one rated about 4A. Something like this should work. NB: you do not want something built to deliver 12V DC because what you'll get is a switching power supply, not a simple transformer.
Mains power transformers are no great shakes as far as audio quality goes because they're designed to work way down in the bass region at 60Hz and consequently have no attention paid at all to reducing the kinds of parasitic internal capacitance whose effects start to become audible at a hundred times that frequency. But for a musical instrument experiment like this, all that their "deficiencies" are going to do is add potentially interesting timbre.
posted by flabdablet at 9:41 PM on March 12
Putting a resistor in series with your wire is going to result in almost all the amplifier's output power being dissipated in the resistor, which is wasteful. You want it dissipated in the wire.
As Stoneshop says, the way to get this done is with a transformer. The impedance seen at a transformer's primary terminals is the impedance connected across the secondary terminals multiplied by the square of the turns ratio.
Exactly what the impedance of your long wire is will depend on what metal it's made of, how thick it is and how long it is. There are formulae available to work this stuff out, but the easiest way will just be to measure your wire with a multimeter set on its lowest resistance range. This will get you its DC resistance, which for two metres of steel wire at musical frequencies I would expect to be the dominant component of its impedance. If you don't have a multimeter, I think 0.1Ω would probably be a reasonable guess.
So to match that load to an amplifier output that's built to drive the 4Ω - 16Ω impedances that are typical for loudspeakers, you'd want a transformer with a turns ratio of √(8Ω / 0.1Ω) = about 9.
The cheapest off-the-shelf transformer capable of handling power of the order of 40W will be a mains power supply type, and the easiest way to work out the turns ratio for these is as the ratio between the transformer's designed supply voltage and its output voltage. So a transformer built to step 110V down to 12V would be about right. As for transformer power rating: assuming that the power rating you want is about equal to the power you intend to deliver to your wire is probably a safe bet. So for a transformer with a 12V output, you'd want one rated about 4A. Something like this should work. NB: you do not want something built to deliver 12V DC because what you'll get is a switching power supply, not a simple transformer.
Mains power transformers are no great shakes as far as audio quality goes because they're designed to work way down in the bass region at 60Hz and consequently have no attention paid at all to reducing the kinds of parasitic internal capacitance whose effects start to become audible at a hundred times that frequency. But for a musical instrument experiment like this, all that their "deficiencies" are going to do is add potentially interesting timbre.
posted by flabdablet at 9:41 PM on March 12
Actually, just thinking about that some more: I've probably not been conservative enough with the power rating calculations there. Given that your load resistance is as absurdly low as it is, the DC resistance of any smallish transformer's secondary winding is probably going to exceed it, which means that you might well end up with more of the amplifier's output power dissipated inside the transformer than by your wire. It still won't be anywhere near as inefficient as a series resistor, but do keep a finger and a nose on the transformer temperature if the amp needs to be turned up loud to get you the output you want.
posted by flabdablet at 9:46 PM on March 12
posted by flabdablet at 9:46 PM on March 12
Another thought: if your wire needs more mechanical drive, you could just add more magnets along its length so that it's getting "plucked" in multiple places at once. The way you space those will affect the harmonics you get out of it too.
posted by flabdablet at 9:51 PM on March 12
posted by flabdablet at 9:51 PM on March 12
It's a matter of some suspense how close I can bring those two before they violently reunite
Should be able to deal with that by just letting them attract each other as hard as they want while keeping just the ends of them separated with something non-magnetic - multiple layers of thin cardboard maybe - so as to make a slot for your wire to vibrate inside.
Or you could remove the entire voice coil drive assembly from the hard disk drive, keep the magnets in their original mounts, and string your wire between them where the voice coil used to go. Those mounts are designed to create a nice consistent field in between the pole pieces for the voice coil to work against.
posted by flabdablet at 9:59 PM on March 12
Should be able to deal with that by just letting them attract each other as hard as they want while keeping just the ends of them separated with something non-magnetic - multiple layers of thin cardboard maybe - so as to make a slot for your wire to vibrate inside.
Or you could remove the entire voice coil drive assembly from the hard disk drive, keep the magnets in their original mounts, and string your wire between them where the voice coil used to go. Those mounts are designed to create a nice consistent field in between the pole pieces for the voice coil to work against.
posted by flabdablet at 9:59 PM on March 12
flabdablet: Putting a resistor in series with your wire is going to result in almost all the amplifier's output power being dissipated in the resistor,
The property you care about is the current through the wire, not the power dissipated in it. Having a suitable (4..8 Ohm) resistor in series will not appreciably change the current through the setup, just keep the amp in its proper operating region.
posted by Stoneshop at 1:25 AM on March 13
The property you care about is the current through the wire, not the power dissipated in it. Having a suitable (4..8 Ohm) resistor in series will not appreciably change the current through the setup, just keep the amp in its proper operating region.
posted by Stoneshop at 1:25 AM on March 13
Quite so, but if you've got an amplifier dumping its rated output power into a series string consisting of a speaker-like resistance plus two metres of piano wire, there will be a lot less current flowing through the piano wire than you'd get by wiring it across the secondary of a suitable transformer instead.
Let's assume that the amp is good for 40W into an 8Ω load, that the piano wire is 0.1Ω and you've got that in series with an 8Ω 40W power resistor (which is a pretty chonky resistor - you'd probably need to make it up from four 33Ω 10W ceramic power resistors in parallel). The maximum current the amp can deliver to this setup, running flat out, is going to be √(40W/8.1Ω) = 2.2A.
But the power actually delivered to the wire with this setup is just (2.2A)2 × 0.1Ω = 0.5W, nowhere even close to the 40W coming out of the amp. Connecting the same piano wire across the secondary of a 9:1 step-down transformer instead, so as to multiply the impedance the amp is driving by the square of the turns ratio, would allow the amplifier to deliver that same 2.2A current into the 0.1Ω wire with just 0.5W into the primary (ignoring transformer losses, which for the relatively small 60W transformer I linked above is probably not something one ought to do). Take the amplifier output power all the way to 40W and you'd see √(40W/0.1Ω) = 20A in the piano wire.
Which makes it perfectly clear that the power rating I guessed at for the transformer I recommended above is totally inadequate. Power dissipation inside a transformer's secondary winding is I2R loss from the winding's DC resistance, and if that resistance is such that the transformer runs warm but safe when providing 12V at 5A, it's going to burn horribly when asked to supply four times that current because it will now be dissipating 16 times its rated power. You'd want a mains transformer rated for at least 20A output, which is going to be a lot bigger and chunkier as well as more expensive than a little pool-light jobbie.
Path of least cost would probably be to acquire the power transformer from a junked microwave oven. Those things are built to deliver a kilowatt to the magnetron, so the core will have no problem dealing with anything that ever comes out of an audio power amplifier. Destroy and discard the existing high-voltage secondary winding, and wind on your own secondary made from much thicker copper wire.
Once you've got rid of the existing secondary, you should have a good enough view of the primary to be able to work out how many turns per layer it has, and how many layers. Multiply those to get the total number of primary turns, divide that by 9 and that's the number of turns to wind onto your new secondary. Use the thickest wire that will fit that many turns through the available holes in the core. Don't wind the new secondary straight onto the bare core - wrap on a few layers of thin cardboard first to stop its sharp corners cutting into your wire's insulation.
posted by flabdablet at 2:56 AM on March 13
Let's assume that the amp is good for 40W into an 8Ω load, that the piano wire is 0.1Ω and you've got that in series with an 8Ω 40W power resistor (which is a pretty chonky resistor - you'd probably need to make it up from four 33Ω 10W ceramic power resistors in parallel). The maximum current the amp can deliver to this setup, running flat out, is going to be √(40W/8.1Ω) = 2.2A.
But the power actually delivered to the wire with this setup is just (2.2A)2 × 0.1Ω = 0.5W, nowhere even close to the 40W coming out of the amp. Connecting the same piano wire across the secondary of a 9:1 step-down transformer instead, so as to multiply the impedance the amp is driving by the square of the turns ratio, would allow the amplifier to deliver that same 2.2A current into the 0.1Ω wire with just 0.5W into the primary (ignoring transformer losses, which for the relatively small 60W transformer I linked above is probably not something one ought to do). Take the amplifier output power all the way to 40W and you'd see √(40W/0.1Ω) = 20A in the piano wire.
Which makes it perfectly clear that the power rating I guessed at for the transformer I recommended above is totally inadequate. Power dissipation inside a transformer's secondary winding is I2R loss from the winding's DC resistance, and if that resistance is such that the transformer runs warm but safe when providing 12V at 5A, it's going to burn horribly when asked to supply four times that current because it will now be dissipating 16 times its rated power. You'd want a mains transformer rated for at least 20A output, which is going to be a lot bigger and chunkier as well as more expensive than a little pool-light jobbie.
Path of least cost would probably be to acquire the power transformer from a junked microwave oven. Those things are built to deliver a kilowatt to the magnetron, so the core will have no problem dealing with anything that ever comes out of an audio power amplifier. Destroy and discard the existing high-voltage secondary winding, and wind on your own secondary made from much thicker copper wire.
Once you've got rid of the existing secondary, you should have a good enough view of the primary to be able to work out how many turns per layer it has, and how many layers. Multiply those to get the total number of primary turns, divide that by 9 and that's the number of turns to wind onto your new secondary. Use the thickest wire that will fit that many turns through the available holes in the core. Don't wind the new secondary straight onto the bare core - wrap on a few layers of thin cardboard first to stop its sharp corners cutting into your wire's insulation.
posted by flabdablet at 2:56 AM on March 13
Music On A Long Thin Wire is available on Bandcamp for others whose interest has been piqued by this question.
posted by flabdablet at 6:20 AM on March 13
posted by flabdablet at 6:20 AM on March 13
So, I've actually built this, and fielded it at several festivals. You've got a lot of good theoretical advice, but hopefully I can give some practical advice.
I built mine almost entirely out of cheap-and-cheerful parts from Amazon. I used a programmable function generator to generate the sine wave, and two cheap Lepai amplifiers - one to drive the sine wave, and one for audio. I strung the wire over two cello bridges set on plywood plinths. I think the resonance helped a bit. I stuck contact piezo mics to the cello bridges. As others have pointed out (and it sounds like you know), the contact mics need pre-amps. I used two $10 9v powered preamps designed for cigar box guitars and ran those into the second Lepai and out to a couple of cheap outdoor speakers. I feel like I could use more gain in the system, but this still worked pretty well. I use 0.037" diameter high carbon steel wire with a #2B smooth finish. I don't quite remember how I picked the diameter, but it seems to work ok. I usually set up the two plinths around 20 feet from each other.
I found that in order to work well, the wire and the function generator need to be tuned pretty close to the same frequency (or some even multiple). I mounted the wires on tuning machine heads for a bass guitar, but in practice I mostly just used the tuners to given the wire a good amount of tension, and then tuned the system by changing the frequency of the sine wave I was outputting. You can dial it in quickly by using an app that tells you the exact frequency of a noise and plucking the wire, but it also works to just scan through frequencies on the function generator until a strong vibration sets up. With that long a wire, the tuning tends to drift as the temperature changes (especially because I set mine up outdoors), and I had to adjust the function generator fairly often. I want to say tuning around 1000hz (two octaves above middle C) was about the sweet spot, but you'll want a lot of empirical experimentation for this.
I originally used a stack of neodymium magnets on either side of the wire, but as you saw, I had to have the magnets very close to the wire, and the wire would often collide with one of the magnets and get stuck to it. I ended up searching eBay for "vintage horseshoe magnet" and buying a giant gray industrial magnet from Raytheon that is very strong, looks cool, and works well. If you do an image search for "Music on a Long Thin Wire", you'll see that's the sort of thing the original installation used. They're expensive, though, and your mileage may vary.
The amp driving the wire got quite hot - I'm sure it was doing it level best to drive a lot of current through the very low resistance, and I was probably overdriving the front end with the signal generator. It did work though, and never failed. I experimented with transformers, as a few other posters have suggested, but I had issues getting it to work, and the no-transformer version worked well enough that I didn't try very hard.
The wire does exert a lot of tension on the system when tuned high enough to work. My two plinths are staked to the ground and have guy lines running fore-and-aft to keep them from collapsing inward. (I'm actually currently mulling over how to adapt it for indoor installation where I can't just run stakes into the ground.)
The running system definitely has its quirks and needs some tending, but it is very cool when it works, and was pretty popular with folks at the events I brought it to.
If you have any other questions, happy to answer them to best of my abilities. Good luck!
posted by aaronstj at 11:14 PM on March 15 [4 favorites]
I built mine almost entirely out of cheap-and-cheerful parts from Amazon. I used a programmable function generator to generate the sine wave, and two cheap Lepai amplifiers - one to drive the sine wave, and one for audio. I strung the wire over two cello bridges set on plywood plinths. I think the resonance helped a bit. I stuck contact piezo mics to the cello bridges. As others have pointed out (and it sounds like you know), the contact mics need pre-amps. I used two $10 9v powered preamps designed for cigar box guitars and ran those into the second Lepai and out to a couple of cheap outdoor speakers. I feel like I could use more gain in the system, but this still worked pretty well. I use 0.037" diameter high carbon steel wire with a #2B smooth finish. I don't quite remember how I picked the diameter, but it seems to work ok. I usually set up the two plinths around 20 feet from each other.
I found that in order to work well, the wire and the function generator need to be tuned pretty close to the same frequency (or some even multiple). I mounted the wires on tuning machine heads for a bass guitar, but in practice I mostly just used the tuners to given the wire a good amount of tension, and then tuned the system by changing the frequency of the sine wave I was outputting. You can dial it in quickly by using an app that tells you the exact frequency of a noise and plucking the wire, but it also works to just scan through frequencies on the function generator until a strong vibration sets up. With that long a wire, the tuning tends to drift as the temperature changes (especially because I set mine up outdoors), and I had to adjust the function generator fairly often. I want to say tuning around 1000hz (two octaves above middle C) was about the sweet spot, but you'll want a lot of empirical experimentation for this.
I originally used a stack of neodymium magnets on either side of the wire, but as you saw, I had to have the magnets very close to the wire, and the wire would often collide with one of the magnets and get stuck to it. I ended up searching eBay for "vintage horseshoe magnet" and buying a giant gray industrial magnet from Raytheon that is very strong, looks cool, and works well. If you do an image search for "Music on a Long Thin Wire", you'll see that's the sort of thing the original installation used. They're expensive, though, and your mileage may vary.
The amp driving the wire got quite hot - I'm sure it was doing it level best to drive a lot of current through the very low resistance, and I was probably overdriving the front end with the signal generator. It did work though, and never failed. I experimented with transformers, as a few other posters have suggested, but I had issues getting it to work, and the no-transformer version worked well enough that I didn't try very hard.
The wire does exert a lot of tension on the system when tuned high enough to work. My two plinths are staked to the ground and have guy lines running fore-and-aft to keep them from collapsing inward. (I'm actually currently mulling over how to adapt it for indoor installation where I can't just run stakes into the ground.)
The running system definitely has its quirks and needs some tending, but it is very cool when it works, and was pretty popular with folks at the events I brought it to.
If you have any other questions, happy to answer them to best of my abilities. Good luck!
posted by aaronstj at 11:14 PM on March 15 [4 favorites]
You are not logged in, either login or create an account to post comments
I plug my guitar into my amp and fiddle with all the knobs to get a cool sound and play.
This seems great.
1) More watts are always better?
2) Given the number of times my cheaper amps have shocked me? Maybe
3) If you can do that! Might as well, (I'm not sure what ohm ratings mean)
4)Hope someone with knowledge can help you!
Super cool
posted by Windopaene at 2:58 PM on March 10 [1 favorite]