psw Posted May 1, 2004 Report Posted May 1, 2004 Sustainer Ideas I've started this topic to continue some discussions from the topic on Rare Earth Magnets and to share ideas, experiences or links on the subject. Sustainer systems have always got to me! They appear to offer not just a means of feedback but a whole new dimension to the instrument. If only someone could get them to work effectively! Heres a your chance to contribute Many years ago I wound a crude coil and ran the guitar output to both my amp and a trashed hi fi amp. The output of this went directly to the coil placed beneath the strings. It worked!!!!!!!!!!!! However some notes better than others, but still it worked!!!!!!!!!!! Only one drawback. Within minutes the coil heated up like a bar radiator and proceded to melt the electrical tape holding it together. The strings really heated up and the whole thing threatend to catch on fire! Still it worked!!!!!! I've got some recordings somewhere to prove it. Would make a great stage act! Any one got experience and info or reference to other stuff on this site or any other, please post here!
STAHLER Posted May 1, 2004 Report Posted May 1, 2004 read this If the link has worked you should read about a sustainer mod that ANSIL came up with. I have not tryed it yet but plan to and if it works well will be fitting it to all my guitars. Ansil is a electronic wizard, i think the guy knows everything.
lovekraft Posted May 2, 2004 Report Posted May 2, 2004 Acoustic coupling seems problematic to me, at least if you're talking about separate drivers for each string. You might be able to make a piezo bridge that had separate bridge pieces, but I think isolation would be difficult. Magnetic drivers wouldn't need isolating since the field could be focused to operate on a single string. Just a couple of thoughts. For reference, here's the patents on the Ebow: D248,122 4,075,921 Notice that the driver coil is wound with much heavier wire than the pickup coil (like Ansil's relay coils) - it's gonna take some power to move the strings, so that probably explains your heating problems. Keep us up to date on your progress.
psw Posted May 2, 2004 Author Report Posted May 2, 2004 Thanks guys for the interest. keep it up Stahler. Ansil's sustainer mod, as he is at pains to explain does not drive the strings. Ive made one but have yet to fit it to the guitar. The idea is to feed the signal back into the pickup, not provide infinite sustain. I'll explain what it sounds like once I get around to loosening all the strings, etc. Lovekraft. The ebow is a classic simple design. I don't quite get why you don't get wierd noises or phase cancellations given the simple circuit. I've moved on from using the brute force of my original experiments. I've done extensive patent searches on the subject. See Floyd Roses sustainer patent for instance to get an idea of the phase problems and interfence assosciated with it. These sustainers are fitted to the neck PU but can only operate when using the bridge pickup because of this radiation. (eg sustaniac/ fernandes) Why the ebow does not appear to suffer from this problem is a mystery to me. Perhaps it is because you can manually move it to different places along the string and that, as it is only sensing and driving one string, it requires less power. The heavier driving coil is because of the heat generated will inevitably burn through normal winding wire. You need quite a bit of current to vibrate the string. Heat is also energy lost. The idea of using piezo's to move magnets beneath the string is that you don't have these problems, nor the effects of an inductor under AC conditions. The piezo could not in itself drive the string (too much power required) but a permanent magnet moved at the resonant frequency by a piezo would create a similar effect as a magnetic pulse from a coil without the complex circuitry and power requirements. Here's an interesting experiment. My last bright idea was to make a small coil through which the string is threaded. The signal is amplified and sent through the string. The string is alternately attracted or repelled by the pickups field or a set of magnets placed just after the bridge. It worked under crude test conditions but not very effectively as the string is not the best conductor once it has left the coil and the metal parts of the bridge absorb the magnetism somewhat. Quite a bit of power was required but the coil was very compact (5mm dia x 25mm). The pickup also picked up the magnetic signal. Keep the feedback coming!
lovekraft Posted May 2, 2004 Report Posted May 2, 2004 <edited for perverse and profuse spelling errors> The ebow is a classic simple design. I don't quite get why you don't get wierd noises or phase cancellations given the simple circuit. Probably for the same reason you don't have problems with a loud amp coupling the signal acoustically - the whole system is a massive lowpass filter, so if the fundamental frequency from the driver is (even close to) in phase with the string, phase shift at higher frequencies becomes moot, and the two are driven into phase "lock". Keep in mind, you don't want a hifi signal to drive the string, in fact a hard-clipped hot signal should be more efficient, since it'll be lowpassed by both the inductance of the driver coil and the mechanical inertia of the string. These sustainers are fitted to the neck PU but can only operate when using the bridge pickup because of this radiation. (eg sustaniac/ fernandes) The neck pickup slot is used because it takes less power to drive the string (closer to the midpoint of the string) - the Sustainiac can be used with a mini-hum in the neck. It disables the neck pickup in sustainer mode to avoid magnetic feedback. Why the ebow does not appear to suffer from this problem is a mystery to me. Perhaps it is because you can manually move it to different places along the string and that, as it is only sensing and driving one string, it requires less power The Ebow senses the string directly through its own pickup, so it's isolated from the guitar pickups. It also drives the pickup as well as the string (thus needing less power), but since its only input is string vibration, there's no chance for magnetic feedback. The idea of using piezo's to move magnets beneath the string is that you don't have these problems, nor the effects of an inductor under AC conditions. The piezo could not in itself drive the string (too much power required) but a permanent magnet moved at the resonant frequency by a piezo would create a similar effect as a magnetic pulse from a coil without the complex circuitry and power requirements.While that sounds like a good idea, in practice it's going to cause all manner of harmonic mayhem. The static magnetic pull on the string sets up an artificial node, causing all kinds of non-harmonic content, like a Str*t with the pickups set too close to the strings. And again, the effect of the inductance of the coil is probably serendipitous, since we want to emphasize the fundamental and not the harmonics. If you want the octave jump, add a full wave rectifier (a la the Octavia). Remember, no-one is going to hear the driver signal (if you design the system right), so optimize it for drive and forget about fidelity. Here's an interesting experiment. My last bright idea was to make a small coil through which the string is threaded. The signal is amplified and sent through the string. The string is alternately attracted or repelled by the pickups field or a set of magnets placed just after the bridge. It worked under crude test conditions but not very effectively as the string is not the best conductor once it has left the coil and the metal parts of the bridge absorb the magnetism somewhat. Quite a bit of power was required but the coil was very compact (5mm dia x 25mm). The pickup also picked up the magnetic signal. I spend too much time trying to keep any voltage away from my strings. You know, Kevin Godley and Lol Creme (ex-10CC) developed a mechanical sustainer using a rotating nylon wheel with tiny picks (plectra?) on it called the Gizmotron, but never could get past production problems to actually sell any. A variation of that, or the rosin-impregnated leather disc in a hurdy-gurdy, using post millenial technology might also yield results. Me, I think I'll stick to magnetics - it just sounds like a guitar should.
psw Posted May 3, 2004 Author Report Posted May 3, 2004 Sorry about the spelling!!! type in a hurry at 4am before work!!! Actually I'm old enough to remember the gizmotron and actually played one! I couldn't afford, nor did I want one, but I stole my flatmates electric toothbrush, cut a slit in it to insert a pick and...instant Dick Dale!!! of course due to the motors hum you got a constant G note through the pickups!!!! you have some good points but... Inductors under AC conditions are out of phase with the actual movement of the string. While with enough power at the fundumental frequency it can be forced into locking into phase, initially the electromagnet is working against the vibration. This dampening effect will reduce the signal so that the necessary power to lock on is lost and the note dies away. The phase variance accounts for the feedback being sensitive to various positions in a room or angle to a loud amp for different notes to work. Your right about it being easier to drive a string from its centre but the EMI is a problem. I would have thought that the ebow's PU coil was just as likely to pick up EMI and RF signals as any other. It's closeness to it's own driving coil should surely cause problems? Is it because any interferance to the pickup is amplified but does not impede on the strings vibrations as it is not related to it's frequency? Why doesn't the radiation of the driving coil not get picked up by the guitars output pickup and get sent to the amp as an audio signal anyway? It is right over the PU coil afterall! How do you mean it drives the pickup and the string requiring less power? The original ebow patent showed a built in version. Effectively six ebows! I noticed that the drivers had to be isolated by an iron block. A more recent sustainer patent shows a driver with another coil with an "error signal" to cancel out the radiation effects. I have a DIY effect called a twang-o-matic (sounds a bit like a drobo) that is basically a half wave amplifier. I thought I could use that as a drive signal as a grab and release type of effect. The driver soon becomes magnetised though to one polarity as it hasn't had time to relax before the next positive wave. square waves are generally best. I test these ideas using a tone generator on the computer to a one string guitar tuned to the frequency so I can try any wave shape or combinations. I don't know that I agree about the harmonic mayhem. The RE magnets I use a powerfull but I only use tiny ones! I dont see it as being anymore powerfull than a conventional pickups mags. Infact the sustainers mags could attenuate the guitar pickup's own natural magnetic field to heighten the effect without adding magnetisim. My early experiments where brute force was used to drive the string always played sharp. Static or electromagnet, its much the same! Newer generation string drivers use coils mounted sideways to cut back on radiation cutting the guitar PU coils and to reduce pull on the string. I would propose a similar thing with the piezo vibrated magnets mounted a right angles to an adjacent pickup. This would cause a distortion in the existing field without extra pull! As for the hurdy-gurdy. Ever played the guitar with a bow a-la Jimmy Page. The rosin don't half make a mess on your axe Thanks for the feedback, it's making me think some more, or maybe I'm thinking too much! P.S. Anyone know about phase-locked loop chips? I heard Boss used one in their distortion/feedbacker mini pedal. Maybe theres an application here
lovekraft Posted May 3, 2004 Report Posted May 3, 2004 Inductors under AC conditions are out of phase with the actual movement of the string. While with enough power at the fundumental frequency it can be forced into locking into phase, initially the electromagnet is working against the vibration. This dampening effect will reduce the signal so that the necessary power to lock on is lost and the note dies away. The phase variance accounts for the feedback being sensitive to various positions in a room or angle to a loud amp for different notes to work.Maybe I'm thick, but I have no idea what that means (in real terms, applied to a sustainer). Wouldn't simply reversing the leads get your signal in phase? I do know that speakers, pickups and vari-tones are all inductors running AC signals, and no insurmountable problems exist. Perhaps I'm missing something. Since an inductor's impedance rises as frequency rises, doesn't that make it a low-pass filter, and ideal for this application? Why doesn't the radiation of the driving coil not get picked up by the guitars output pickup and get sent to the amp as an audio signal anyway? It is right over the PU coil afterall!...How do you mean it drives the pickup and the string requiring less power? If you re-read that, I think you answered your own question. In use, my Ebow gets louder the closer it is to the polepiece on the pickup. I don't know that I agree about the harmonic mayhem. The RE magnets I use a powerfull but I only use tiny ones! I dont see it as being anymore powerfull than a conventional pickups mags. It's a problem with conventional pickups as well. From BluesHawk: "Another problem with powerful magnets is that they can change the pitch ("tuning") of a string - these effects known as "wolf-tones" are common on the lower strings of Strats if the pickup is raised too close to the strings. BluesHawks are subject to "wolf-tones" if the neck pick-up is raised too high." From Stew-Mac: "Adjusting single coil pickups any closer can cause false notes or "wolf tones." If you've ever played a Strat® that sounded like you were getting two notes from one string (usually the wound strings) in the upper registers, you have witnessed this anomaly. It is most often referred to as "Strat-itis" and can cause a lot of head scratching unless you know what is happening." Static or electromagnet, its much the same! Well, only if you never use the guitar with the sustainer turned off. I don't know any way to "turn off" a permanent magnet. P.S. Anyone know about phase-locked loop chips? I heard Boss used one in their distortion/feedbacker mini pedal. Maybe theres an application here A PLL is basically a voltage controlled oscillator whose frequency is controlled by a phase comparator that compares the VCO to an input signal and generates an error voltage if there is a difference (in effect, "locking phase" with the input). It's the electronic analog of the physical system that is a sustainer. The "feedback" signal is generated in the pedal - it's a cool little box, especially when you use it with something other than guitar (like saxophone), but it doesn't really generate conventional feedback.
daveq Posted May 3, 2004 Report Posted May 3, 2004 Has anyone considered doing something like the sustainiac model C acoustic sustainer. I got one a couple of weeks ago and don't want to play without it. It attaches to the headstock with a simple clamp mechanism. On the headstock is basically (from what I can tell) the guts of a speaker minus the woofer (a coil and magnet I think). The only problem with this is that the extra wires have to be dealt with. You can of course build them into the neck for a new project but for existing guitars, it's just something that has to be dealt with. I wish I had some experience with magnetic sustainers so I could compare. The model C has the smarts to deal with the out-of-phase issue when you happen to hit a note that is out of phase with the signal being sent to the transducer. You can literally sustain a note as long as you want with this thing. Does anyone know if this is true with the magnetic type? Anyway - just a thought.
psw Posted May 3, 2004 Author Report Posted May 3, 2004 Ok so this is what I mean.... "Maybe I'm thick, but I have no idea what that means (in real terms, applied to a sustainer). Wouldn't simply reversing the leads get your signal in phase? I do know that speakers, pickups and vari-tones are all inductors running AC signals, and no insurmountable problems exist. Perhaps I'm missing something". Comparing the phase difference between the current and voltage through and across each component, notice the following: For a resistor, there is no phase difference; for a capacitor, there is a +90 degree phase diference and for an inductor there is a -90 degree phase diference.from Practical Electronics for Inventors by Paul Scherz Basically, it takes time to create the magnetic field and then to reverse it, it needs to fight the polarity and then re establish the new field. By the time this occurs, the physical vibration of the string has already changed. The field is working against the movement of the string. A 180 phase reversal would cancel it out (ie stop the string (at least at the fundumental). As you described earlier, with enough power from a loud amp, the string can be forced into phase (locked at 0 degrees phase) by physically altering the movement of the string but this takes a bit of work, meanwhile the signal is dying away (there is often a compressor in the sustainer circuitry to keep this up) A compressor or a fuzz's natural compression is large factor in getting natural feedback, not only the square wave. The phase difference accounts for some notes to feedback more easily than others and to change as you move about as the time it takes to get from the speaker to the string compensates for the phase. I hope this helps. I thought that perhaps a PPL could be used to compensate phase diffence in the driving circuit. oh yeah, and the phase difference is related to frequency. The higher the frequency, the less time between cycles to relax the field and reverse it. The Floyd Rose patent swiched in various capacitors via a frequency analyser. Inductors -90 capacitors +90. In tone circuits the inductors are sending signals to ground so phase does not matter!
psw Posted May 3, 2004 Author Report Posted May 3, 2004 daveq Never tried one but there was a great flexidisc promotion in guitar player many years ago. This model is the only sustainer system that allows you to patch effects into the drive signal which is cool. As it does not rely on batteries it has a lot of power and a little smarts. As it is on the end of the neck it is far enough away from the pickups to avoid problems and, it uses the leaverage of the neck to vibrate the whole instrument! I hear that it feels like the gyuitar is alive in your hands!!!!
Primal Posted May 3, 2004 Report Posted May 3, 2004 Why not build something that, instead of trying to keep a constant signal going, have it "ping" the string?
lovekraft Posted May 4, 2004 Report Posted May 4, 2004 psw I apologize for giving the impression that I didn't know how inductors work - obviously I failed to make myself clear. The phase relationship between voltage and current is -90 degrees, but that really doesn't have anything to do with what we're talking about. If the string moving in a magnetic field induces a current in the pickup coil and we use that current (amplified, and clipped/compressed) to induce movement in the string, where is the phase shift? It's a moot point, there is obviously a solution, since the Heet/Ebow patent clearly shows it being done. At any rate, I've enjoyed this discussion, but before this becomes more about who's right than how thing work, let me exit gracefully. When I have time, I'll mock something up to try this stuff out, and let you know my results. Good luck with your experiments.
psw Posted May 4, 2004 Author Report Posted May 4, 2004 Lovekraft Thanks for your input, I really appreciate your time and effort. I guess I've read too many patent docs. Perhaps I am over concerned about the phase relationships. That's exactly why I am so intriged by the ebow: its so simple yet it works! I think I'll mock one up or get one to study how it performs. Rock on!
psw Posted May 4, 2004 Author Report Posted May 4, 2004 Hey primal good suggestion! I'm not sure what you mean't but I built a DIY effect called a twang-o-matic that is basically a half wave amplifier. It sounds a little like a resonator guitar. Anyway, this only amplifies the positive side of the wave so as to create a kind of "ping" signal if you like. The problem with it is that the driving coil charges and relaxes slower than you want but there is potential there I think. or did you have in mind actually hitting the string? Thanks for joining in, more ideas please
Primal Posted May 4, 2004 Report Posted May 4, 2004 Okay, I didn't word that very well. Here it goes again: Instead of creating a device that will sustain a string at a certain pitch (i.e. a field that oscillates at 440 Hz to maintain the note A), create one that will emit "pulses" strong enough to get the string vibrating (or to maintain the vibration of the string). This would keep the device from fighting the vibration of the string when notes are changed. Does this make more sense?
psw Posted May 5, 2004 Author Report Posted May 5, 2004 Primal I know where your coming from, sort of But, if the thing pulses it must do so at some kind of frequency, yeah??? My idea of using vibrating motors to shake the bridge was like that. I was hoping that the strings would move from that alone. I found some in Hong Kong from a manufacturer who is prepared to send some samples for free but the freight is about US$50!!!!! LOvekraft had some valid concerns about standing waves, etc. so perhaps it wouldn't work
psw Posted May 12, 2004 Author Report Posted May 12, 2004 I have made a working prototype practical sustainer!!!!!!!!!!!!!!! Yesterday I suscessfully put together a magnetic sustainer. The device uses a compact coil (5mm round x 4mm deep) with a ferrite coil and a tiny Rare Earth Magnet. It is powered by a 9v battery and a tiny and very simple amplification circuit. It was tested on my one string test guitar. This is a crap guitar neck bolted to a piece of pine with a very low quality single coil pickup taped to it. This test guitar is used because it has low natural sustain and high suseptability to hum etc. It is my worst case senario instrument as I figure if it works on that it will work on anything! Electromagnetic radiation is present but only when the coil is close to the guitar's sensing pickup. It will successfully drive the string at about the centre strat pickup position without noise. This was one of my aims as the eventual design idea is to slot it into this position. Reversing the polarity of the magnet (or the coil) suppresses the fundumental and drives the string at higher order harmonics (eg a fifth above, etc). The sound. Think the intro to U2's With or Without You. Infinite clean sustain! Oh, Happy Day
Biohazard Posted May 12, 2004 Report Posted May 12, 2004 WOW, thats amazing. Let me congratulate you, well done. What are your plans now for your sustainer? How long have you been working to get to this stage of having a working prototype?
lovekraft Posted May 12, 2004 Report Posted May 12, 2004 Glad to hear those inductors weren't a problem Seriously, congrats! How about some details, like amp circuit, wire gauge, # of turns, slug size, maybe a wiring diagram? Sounds like fun, record some clips when you can.
psw Posted May 13, 2004 Author Report Posted May 13, 2004 LK Well I still get electromagnetic radiation which makes it squeel violently if it gets to close to the driving pickup, so I still don't get how the ebow does it! Biohazard - I've been trying various things for a couple of months this time around. Its something I just keep coming back to. It looks as if all those patent guys were being too clever by half with the phase problems etc. Maybe you were right LK. or maybe I'll come across them as the prototypes get more sophisticated. Basically I used the 386 sustainer mod circuit of Ansil's (couldn't get the sustainer mod to do much by the way) to drive a coil which I held close to the string. I used a ferrite bead 4mm x 5mm high as the core. I have been experimenting with ferrite as, while it is easy to magnetise it does not hold it's magnetism like iron does. My theory is that alot of problems with the phases involve risidual magnetism in the core and the reactive speed to rapidly changing AC signals. A 5x1.5mm rare earth magnet was attached to the outside (bottom) of the coil. A smaller mag could be inserted within the ferrite for the same effect and this would be a better thing. The coil was, I think, about 300 turns of 0.2mm enamelled wire. I used a silicone gasket RTV as I was winding as a kind of potting. I didn't want vibrations in the coil causing problems. The final coil measures 12mm round and comes in at 4 ohms exactly. It will produce quite strong continuous sustain even at half strength so seems to require little power to run. The next stage is to make a multi-string driver to fit in a strat cover for the centre position. I might find that a larger coil, or multiple coils cause even more radiation so I have come up with a kind of humbucking configuration and various focusing plates and shielding ideas if I run into problems. The signal came from a bridge position cheapo single coil pickup which also ran to the amp. I tried driving it from a small coil but it lacked power I think. I tried other configurations to make an independent ebow copy but could not get it to stop squeeling! Can anyone think of a way of using the same coil to both pickup up and drive the string or am I expecting too much.
lovekraft Posted May 13, 2004 Report Posted May 13, 2004 LK Well I still get electromagnetic radiation which makes it squeel violently if it gets to close to the driving pickup, so I still don't get how the ebow does it! That's good ol' garden variety feedback - the Ebow avoids it by not using the guitar pickup in the loop - it has it's own pickup coil. Can anyone think of a way of using the same coil to both pickup up and drive the string or am I expecting too much. Ideally, you'll need 6 sensors and 6 drivers - one for each string. Trying to use one coil for both sensor and driver gives you an immediate feedback loop (squeal!!). Using either a single pickup or a single driver will cause the system to settle towards a single frequency closest to the resonance of the circuit/strings system (my Fernandes doesn't sustain chords well - one note quickly becomes predominate, even with double stops). The only other solution that's occurred to me is to apply a bandpass filter to the input of each driver, but that seems both complicated and problematic to tune, so individual sustainers for each string seems the most elegant idea. So, a hex pickup, 6 current amps and 6 driver coils - or one Ebow per string, in effect. Sounds like you're on the right track, keep us updated. Great work!
otgordin Posted May 13, 2004 Report Posted May 13, 2004 Hey guys, I am not sure if it was here or on MIMF, but i remember a thread where a bass player made little "dot" pickups for each string. I would assume that that same approach would work here. I tried to find it in both places. Hope this helps. -Vadim
Primal Posted May 13, 2004 Report Posted May 13, 2004 That was on MIMF. I think it was called the dragonfly bass.
psw Posted May 13, 2004 Author Report Posted May 13, 2004 Thats where I'm going with the hex idea but I'd like to keep the sensors and drivers in the one package like the Ebow. I don't want to ditch other pickups to just for this gizmo. Actually, I described the squeel wrong. When I tried a pickup, separate from the drive chain, thats when I got the characteristic feedback squeel. I tried various configurations, side by side (ala ebow), stacked and sharing the same magnet, non-magnetic piezo drivers, etc. LK: That's why I still don't know why the ebow doesn't do the same to itself and/or the signal pickup! The squeel I refered to when using the guitar's coil is controlled. Moving the driver coil around produces an amazing theramin type sound. More magnets raises the frequency, different angles and locations slide the tone. The sound is unrelated to the string but is induced in the coils. How about filter caps on the input. I'm not using any. Would they help cut back on the feedback squeel between driver and sensor? What are they for (I know they are commonly used). The driver/sensor idea I had was to Have the positive cycle of the wave sensed by the coil; A sample and hold chip would store it: The wave passing through zero would trigger it: An amplifier would increase and invert it back through the coil through the negative cycle. The idea being to sense the positive cycles and drive the negative. Avoiding feedback squeel as the sensor and driver would not actually be on at the same time. thanks to everyone for their contributions and encouragement, progress is being made me thinks! psw
lovekraft Posted May 13, 2004 Report Posted May 13, 2004 LK: That's why I still don't know why the ebow doesn't do the same to itself and/or the signal pickup! The squeel I refered to when using the guitar's coil is controlled. Moving the driver coil around produces an amazing theramin type sound. More magnets raises the frequency, different angles and locations slide the tone. The sound is unrelated to the string but is induced in the coils. That's the feedback I'm talking about - the direct magnetic coupling of the driver to the pickup - moving the coil around varies the frequency just like moving around changes acoustic feedback. The ebow avoids the problem because it is not connected to the pickup - it has it's own sensor which is magnetically isolated from the driver coil but still physically very close to it. If you add a second pickup to your system and don't connect it to your driver amplifier, but connect it to a guitar amp, I think you'll find that you can put the driver coil directly over it with no whistle at all. Remember, an oscillator is just an amplifier with positive feedback and a resonant tuning circuit. I hope that's a bit less obscure. The driver/sensor idea I had was to Have the positive cycle of the wave sensed by the coil; A sample and hold chip would store it: The wave passing through zero would trigger it: An amplifier would increase and invert it back through the coil through the negative cycle. The idea being to sense the positive cycles and drive the negative. Avoiding feedback squeel as the sensor and driver would not actually be on at the same time. I'm afraid that's beyond my capabilities - I understood that a S&H circuit stored a voltage level, not an AC signal. Sounds complicated, don't know how you'd implement it.
Recommended Posts