McSeem

My suggestion to moderators. It looks like the "Electronics" section is mostly about DIY sustainers. Maybe it makes sense to add a special forum section? -- Guitar sustain systems?

I understand that, and I can estimate the inductance, etc. I was asking about your personal experience, if it's not a secret. Say, high strings are much harder to drive, so, seemingly, the high string coils must produce stronger magnetic field (the signal one, of course, not permanent). You can tweak it with amp gain, but you can also try to optimize the coil parameters for different strings. I have a suspicion that high string coils must be bigger, have thicker wire, while preserving about the same DC resistance. Is that correct? Once again - it's only a matter of optimization, nothing more.

BTW, Col and Psw, based on your great experience, can I ask you to roughly estimate parameters, optimized for each string, considering different frequency ranges, different string thickness, and assuming that the driver consists of 6 independent coils, driven by 6 power amps and a true hexaphonic pickup (exactly like in the Moog guitar). I'm not suggesting you to build this "nightmare", I'm just looking for a starting point for my experiments.

Yes, that's right, I namely meant nodes of the standing waves. I just read this term somewhere and it was obviously not quite correct. Thank you for the correretion. A magnetic sustainer powered by the signal from the pickup, tends to amplify a single harmonic, eventually turning the sound into just a pure sine wave. Even if the driver is not exactly at the node, but close, the corresponding harmonic won't sound, simply because another one, that is close to its antinode will win. I noticed that it's a matter of competition. You have resonances at different frequencies with very high Q-factor, and the system with natural positive feedback loop. It means that one harmonic, that has the best resonance conditions wins. And it's not necessarily the fundamental one. BTW, if you want to enforce higher harmonics, I'd suggest to try... a diode bridge at the output. I know, audiophiles would kill me :-), but it's just an experiment. Yes, 9 volts may be not enough because of significant extra voltage drop in the diodes.

First, magnetic drivers are local. Any harmonic has knots on the string. The fundamental one has knots at the bridge and the fret. The second - one extra knot exactly in the middle, the 3rd - two extra knots and so on. When you place the driver at the knot of the harmonic you can never sustain this very harmonic. And the positions of the knots depend on your left hand, so, you will inevitably have certain harmonics at certain positions that are impossible to sustain. Maybe there are some drivers on the bridge (piezo?), at least it's theoretically possible. I know sustaniac use piezo driver, passing vibrations through the neck. Similarly, you can theoretically design saddles with piezo drivers. It will solve the problem of knots. Second, phase shifting is not just a delay. Ideally, there must be frequency independent phase shift, so that, at any frequency you would have a stable 90 degree (for example) phase shift. It can only be achieved with pitch detection and digital processing. Well, you can easily have 180 degree phase shift by just inverting the signal, but not arbitrary shift.

Oh, I'm sorry for possible misunderstanding. If you want to wind a single coil driver, yes, you namely need this thick wire. But for a small, 6-coil one, there's no way to use it. You can barely achieve 0.5 ohm or so. That's what I meant.

Me too! I need a 6-pole pot, but ultra-low profile. I found ones, http://www.bourns.com/ PTD906. But the minimal order is 500 items and it takes about 4 weeks. I accidentally found Bourns 8-pole 500K ones, at Mouser Electronics.

Common, it's a shitty thick wire that you can buy in Radioshak. You need at least 36AWG, but never 30! So it goes...

Yes, the term "gauge" is confusing. In USA it makes sense to use unambiguous "AWG", which means American Wire Gauge. Here's the chart and calculator: http://www.technick.net/public/code/cp_dpa...e_awg_to_metric Oh, BTW, here's a complete list: http://www.bulkwire.com/wiregauge.asp

I wound a 5mm height coil, 38 AWG, about 26 ohms DC resistance. In fact, I made a mistake, it's supposed to be 36 AWG, but it's just experimentations, not a big deal to re-wind it. I used a drill as the lathe to make cores from humbucker pins, soda-bottle plastic for the washers. I can describe this "homecooking" technology in detail, with photos. It's not very practical and takes time, but you can make almost any coil you want for your experiments and optimizations.

The general tip is to google for a particular gauge, say, "36 AWG magnet wire" http://www.powerwerx.com/wire-cable/magnet...CFRyfnAod9U_4-Q http://www.allspectrum.com/store/index.php...4ab23ba21c3badf

No, I meant, double coils for each string, something like that: | *|* | *|* | *|* *|* | *|* | *|* | To save space you obviously have to place them in the alternate order. Your configuration looks like 6 coils connected in parallel? At least the connector looks like just a 2-pin one. It eliminates the whole idea of driving all 6 strings simultaneously. Or, am I wrong? I'm kinda confused again. What are the metal sticks on top of the coils? Does does your driver consist of two rows of coils, in a humbucker-like configuration, 12 coils all in all?

My goal is different. I anyway want to experiment with hexaphonic mag and piezo pickups, audio interfaces, digital processing and so on. I'm not so restricted with power, batteries, equipment, and yes, with my budget. My goal is to try to get new unusual sound effects, based on analog source signals (not MIDI!), digital processing and feedback. If you are so concerned about power consumption, why not to use class D amps? Analog Devices has plenty of them, for example: http://www.analog.com/static/imported-file...ets/SSM2305.pdf It's filterless (I suspect they use a bridge circuits internally), you don't even need a capacitor in the output. There's no need for super-linear-class-A zero-THD amp. The quality, noise, and other parameters are not important for sustain drivers. :-))) I'm confused. What happened last year? I've been using this nick for a long time, since I started my Antigrain Geomentry project.

Can you give me some direct links to your more detailed description? And have you ever tried a humbucker-like configuration, with two coils and the poles across the string?

Maybe there already was a discussion in the long thread, but just in case. Has anyone tried the following configuration? 1 or 2? In theory it should vibrate the string not only up-down, but also left-right (with respect to the picture). Although, it will definitely take more space.

Col, obviously, your amp has very low Z, so, the resonance is heavily damped and does not have any significant effect. In general, any sharp resonance is not good for the driver. What is good of it if you can only sustain one note? :-)

The principle is simple, the practice is not. First, all must be hexaphonic. There must be 6 single-string pickups placed as close as possible to the drivers. If the signals are cophasal the sustainer drives the string, if they are counter-phasal the signal stops vibrations. The problem is it's not that trivial to obtain strict phasing on different frequencies because of inductive and capacitive reactance. So there must be a filter that compensates the pickup/driver phase shifts. It may be non-trivial. Besides, the vibrations are polarized, in general it's an elliptical polarization that rotates in time, differently for different harmonics. So, for best result there must be a "stereo" pickup and driver, that operates in 2D, in a plane, perpendicular to the string. Besides, if so happens that your driver is placed exactly at the knot of the standing wave in the string (which is very possible), it won't affect that very harmonic at all. So, it must be localized and distributed along the entire string -- at the same time!!!

BTW, another crazy idea is to design a true stereo pickup, working in the same way the vintage vinyl pickups work. Each pickup consists of two tiny coils at 90 degree to each other and placed closely to the bridge, where the mechanical amplitude is not that high. There coils are connected to the right and left channels. The point is the polarization of string vibrations varies during the time, so that, you will have "natural dynamic panning" and some kind of natural stereo sound.

Another thing to consider is that a single, on-body mounted driver will have relative-to-string positions at the knots or near-knots of standing waves. For example, if you place the driver at exactly 12th fret, you can never drive the 2nd harmonic on an open string. And there's no particular position that will work with all frets. Double or even triple-coil driver definitely improves the things. Well, it becomes something like 3-way speakers, and who knows, maybe in 10 years true hi-fi maniacs will use tri-wiring/tri-amping to power sustainers. Just imagine 18 coils, 18 amps, and 36-wire cable for balanced outputs... )) When I get all my stuff I'm waiting for, I'll try to experiment with dynamic harmonic equalization in DSP. I can analyse the spectrum right after the attack and try to preserve it as much as possible. I'm not restricted with compact in-cavity circuits, I'm not restricted with power, and of course, it will require at least a 13-pin Roland cable, or even a separate one. I'm only afraid that the coils may start smoking :-) I also have a suspicion that rich, multi-harmonic sound may be obtained if you power your coils with very strong and narrow delta-impulses at the fundamental frequency, as it will simulate in some way your pick and/or fingers. But I'm not sure, it's just an assumption. I'll try to do that also, but I also realize, magnetic shielding may be a huge problem in this case.

why? Because of magnetic permeability. Say, neodymium magnets have low permeability, but high coercivity. In other words, if you compare the inductance of the same coil, wound on a neodymium magnet and a piece of nail of the same size, there will be a huge difference. Of course, I agree, the input and the driver must be as close as possible, but still it's interesting to experiment. With a mounted driver, I know the distance between the bridge pickups and the driver, I can compute the phases for each harmonic accurately (and yes, for each string and each fret), I can use FFT to adjust the phases in conformance with the string vibration at a particular position. The overall latency can also be accurately calculated and/or adjusted. So, it may sound crazy, but I can tweak the signal very flexibly with DSP, which you can never achieve with analog circuits. Finger mounted driver won't work properly in this case, but still can be potentially used for fancy/strange effects. Right, it's going to be more like acoustic sound, at least, with Graphtech pickups. However, I also have a hexaphonic mag pickup at the bridge, which can also be used. I'm not sure either, we will see. If the strings can be drivable at about 5mm, it can be used. At least, in my early experiments with a very weak driver (250 ohms coil) and a bridge single coil pickup, I could start 2nd, 3rd, up to 5th harmonic by just moving the coil along the string. BTW, for a true 2D sustainer you would need stereo pickups and drivers placed like this: string * / \ drivers McSeem

The problem is not only with local and strong permanent field. The core of the coil must be made of some soft magnetic alloy with high permeability, as high as possible. Ideally, it must be some alloy used for magnetic shielding. I'm trying to experiment with pins from a humbucker, doing some lathe works with a cheap drill in a vise, connected in series to a 300W lamp with a dimmer to adjust speed. Then you attach small disk magnets to the bottom in order to have magnetic bias. Still waiting for parts from Graphtech; I spend more than $500 to order 6 acoustiphonic preamps plus pickups and stuff. Then I'm going to use a low latency multichannel audio interface and process everything in software, for all 6 channels in parallel. So that, I can take a hexaphonic signal, mix and pan it as I want to stereo outputs, plus use another 6 outputs for the sustainer driver. The point is I can easily filter and distort the signal, adjusting the parameters for each string independently. Plus, it's easy to implement frequency independent phase shift, add some "burst" mode to enforce sustain quicker, and so on ...endlesss possibilities. The point of using piezo pickups is that there's no direct EM feedback with the driver, so that, the sustainer can be "portable", you can put it on your right little finger and move along the strings, starting different harmonics.