col Posted March 3, 2008 Report Share Posted March 3, 2008 Yes, I'm building dual core driver, so that both coils have separated inputs. That way I can test them in series and parallel, in phase,and out of phase. Maybe I wind 16 ohms coil around those cores also. That means six wires... and lot of testing possibilities. I have no clue about theory, and how that works with sustainer, but for me there is only one way to find out... If you want to keep the coils separate in that way, you will need two separate driver circuits (might work well, but will certainly be much more difficult to fit the whole think into your guitar, and there will still be a fair amount of magnetic coupling between them anyway). BTW, I took stick of wood (about 35 cm), and wound 10 rounds of coil wire to it. Then measured resistance, and did some simple calculations. Certain amount of rounds gives about 8 ohms, twice that gives about 16 ohms etc, so that I can cut 4, 8 or 16 ohm amount of wire for future drivers. Don't know if that method is already in use, but no more guessing when winding. Well, that is not hugely useful. There are programs available that can tell you number of turns for a certain guage and core size that even take into account that as the coil grows, one turn uses more wire, but they still are only partially useful for us. You have to learn the difference between resistance and impedance. They are both measured in ohms, but resistance does not take inot account the effects of capacitors and inductors on an alternating current (like our drive signal). It is possible to avoid most of the effects of impedance. If you make sure that the inductance of the driver is low enough, most of the impedance comes from the resistance, and you can ignore the 'reactance' - this is what Pete ended up with through trial and error - the highest inductance you can use and still have DC resistance as the significant part of the impedance. The downside of this is that a lower inductance gives you a weaker magnetic field for the same current. Using a higher inductance driver should be more efficient, but it will be important to take into consideration the reactance at different frequencies to make sure that: #1 you don't kill your amp by trying to draw too much current at low frequencies #2 the reactance at the upper frequencies that you want the driver to drive is low enough so that it still drives... heres an example: A driver with 4ohm resistance that has an inductance of 4mH (2x the magnetic power of 1mH(I think?)) will have: roughly 4 ohm impedance @ 80Hz (no good for bass guitar, but cool for electric) roughly 8 ohm impedance @ 300 Hz - best performance exactly where we need it near open 1st string roughly 16 ohm @ 600 Hz about 30 ohm @ 1200 - this is getting to the stage where drive will be heavily reduced, the question is does the fact that playing up the neck brings the string much closer to the driver counter act the negative effect ? Another similar set of figures exists for a DC resistance of 6ohm with a 3mH coil - you get lower power over all, but the performance at 80Hz and at 1200 Hz are both better than with the 4ohm 4mH driver... these higher inductance drivers have slightly more phase shift at low frequencies, but not an unmanageable amount. It is important is that we look carefully at the current that is being sucked from the battery, with these higher inductances, we have to be even more vigilant. If you take a look at the last three graphs on page 4 of the LM386 datasheet, you can see that at 4 ohms, the chip will waste a lot more energy as heat, so getting the balance right is not just about drive vs current through the coil, but also about 'dissipation' (heat loss) vs impedance... to get maximum drive through 4ohms at 80Hz is going to kill the battery stone dead pretty quickly... so as usual, compromises are necessary... or switching to class-d maybe there are amp chips that are better at handling a variety of loads ? cheers Col Link to comment Share on other sites More sharing options...
psw Posted March 3, 2008 Author Report Share Posted March 3, 2008 If you take a look at the last three graphs on page 4 of the LM386 datasheet, you can see that at 4 ohms, the chip will waste a lot more energy as heat, so getting the balance right is not just about drive vs current through the coil, but also about 'dissipation' (heat loss) vs impedance... to get maximum drive through 4ohms at 80Hz is going to kill the battery stone dead pretty quickly... so as usual, compromises are necessary... or switching to class-d smile.gif maybe there are amp chips that are better at handling a variety of loads ? I think both FF and UI are still using more powerful amps with current drawn from powerpacks...2.5 watts as I recall. UI did mention he was getting a lot of heat (something I get no discernible amount from, I actually considered this as an attribute of the ultra thin coil that it's very low mass and flat design would keep cool...if that was going to be a problem). It is possible to avoid most of the effects of impedance. If you make sure that the inductance of the driver is low enough, most of the impedance comes from the resistance, and you can ignore the 'reactance' - this is what Pete ended up with through trial and error - the highest inductance you can use and still have DC resistance as the significant part of the impedance. I think you may be right, and this is of course exactly what I have been aiming to do, find that window in which a simple circuit can drive a simple driver effectively over the whole range of the guitar. There is still some mystery as to why the harmonic mode does not work so well and yet high harmonics can be sounded...perhaps at these higher frequencies the phase is off too much with this particular driver. I have a kind of feeling that the thicker "thin driver" (3mm thick) was more "efficient" and seemed to do the harmonic thing really well...but then maybe it had more phase drift or something. On the other hand, it would seem that the newer drivers require less amplifier gain to run it...quite a bit less it would seem. If you want to keep the coils separate in that way, you will need two separate driver circuits (might work well, but will certainly be much more difficult to fit the whole think into your guitar, and there will still be a fair amount of magnetic coupling between them anyway). I had always aspired to multi-coil drivers...so convinced of this approach that I spent a year on the HEX driver designs (six individual drivers, one per string). I am not quite so convinced anymore, but the polyphonic sustain (so that I can sustain evenly small coils and devise music with sophisticated voice leading "pads" for instance) was and now that I am playing with the thing again, remains an important attribute. One of the problems with these ideas is that each driver, whether 2 or 6, right next to one another will influence the other. In a setup devised to reduce or contain EMI as with the bi-lateral thing, that is an attribute...with something running different signals or filtered signals or dissimilar coils and such...perhaps crosstalk would make a difference. You have to consider that no matter the number of driver coils, you are running from a mono source. My goal with the hex designs was to limit cross talk effects a bit (the reason for the 12 magnets in there) and aim towards tailoring each for string gauges and frequencies later to get a more even response once the design looked to be practical. I also made them with the aim of eventually running each separately, perhaps from a hex piezo system...ghost of variax type saddle Pups for instance. As the hex things effectively worked as one from the mono source, I never heard the benefit of these aims (if there is any) and in making separate tiny drivers, I did not notice a significant difference in sound over the more conventional designs. The big improvement in the hex designs was compact size (looked like a hex synth pickup kind of thing) and the containment of the EMI. It was working with these that directly lead to the idea of the thin coil designs. Oddly enough, the newer drivers appear to do the polyphonic thing a little better...lower notes will often predominate, but picking strength also has a big effect. The amount of signal that a chord puts into the circuit and driver compared to a single note is significant and does have some effect on clarity or something. This is an area of performance where each guitar, setup and style of play is going to impact a lot. I think I went too far with the hex things...at least to far two quickly. I think that it is easy to push 'u-beaut' designs a little too far. Sometimes you can get a little too clever and the devices get harder and harder to make. Making a dual driver would appear to be twice as hard as a single coil (two coils, cores plus you have to wire them together and stick them all together and mount them into the guitar somehow) for instance...so you'd want some kind of payoff for all this work. I get the feeling, although col's work is an exception, that the most has not been made by some of the basic design. Making a more complicated design does not seem to be working towards solving the problems some are dealing with (construction, circuitry, practicality, etc). Also, col suggests that these bulkier coils will have more impedance and thus phase/frequency effects which will need to be addressed...col has the skills to do this, I don't nor do I think that phase correction is an area that needs to be pursued very strongly if the driver design is such that it is not really required...as mine appears not to. So, we are traveling on parallel but different roads with this at this time. I do see some benefits of various dual coil driver designs, and fully intend to make some HB versions. It may even partly address my switching issues with equal and opposite coils producing less buildup in the secondary coils of the underlying pickup. I may even make a pseudo bilateral with half a coil on one Hb coil and the other half on the opposite coil...kind of what I have proposed before...but only after I have the basic driver and circuit designs working so I can transfer the technology and not transfer over the problems not solved. At the moment, these "problems" with my setup is the whole switching and integration things...unfortunately time to think, but none to do on this lately. The performance is fairly impressive and sure tweaking is possible to improve it or improve the range of effects it can produce (harmonics) or control. Circuit size is tiny, even DIY 5-6 times smaller than the commercial units...it will be hard to get these kind of attributes with a driver that requires electronic compensation for impedance effects and I am not sure that efficiency gains will be such to influence the power requirements either. But, mainly I don't pursue such things anymore while outstanding problems exist with things like switching and such...at least for me pete Link to comment Share on other sites More sharing options...
Fresh Fizz Posted March 4, 2008 Report Share Posted March 4, 2008 I think I went too far with the hex things...at least to far two quickly. I think that it is easy to push 'u-beaut' designs a little too far. Sometimes you can get a little too clever and the devices get harder and harder to make. Making a dual driver would appear to be twice as hard as a single coil (two coils, cores plus you have to wire them together and stick them all together and mount them into the guitar somehow) for instance...so you'd want some kind of payoff for all this work. The danger is that once you're start copying a part of sustainiac's device you might need to go all the way and you'll finally end up copying the complete sustainiac. That's not exactly the spirit of DIY. I'm only interested in that tiny part in the sustainiac that reduces the fizz. If the bilateral driver is the solution than I will build one, if I could only find some decent magnets. So I'm awaiting the results of UI's, Col's and Curtisa's projects. In the meantime I'm reading a bit of the boring patents hoping to find something fizz-related. Good luck to you all Fresh Fizz Link to comment Share on other sites More sharing options...
joshvegas Posted March 4, 2008 Report Share Posted March 4, 2008 I have a question that may be stupid. Why do you need to use a magnet at all? If you rap a piece of iron in wire and pass a current through it you already have a magnet so surely using another magnet is jsut complicating things or is it the reaction between the two magnets that creates the sustain in the string? Link to comment Share on other sites More sharing options...
psw Posted March 4, 2008 Author Report Share Posted March 4, 2008 The danger is that once you're start copying a part of sustainiac's device you might need to go all the way Absolutely...and the sustainiac is very good at what it does and may be cheaper in the long run even if we had the exact plans, specifications and access to the materials to build one. I have always wanted to take the technology further and in the past have gone to great lengths to try to do this technically, especially in relation to the driver. My HEX drivers were an attempt to throw every EMI reducing strategy at it..if a paired coil driver is good, six matched drivers is better, if side coils sound like a reasonable proposition, lets try cross drivers, if opposing coils, magnets sound like a good idea, then lets go for 12!!! But then these things were very difficult to make and in the end I believe the greatest contribution was their necessary small size (each driver had to fit between the strings), and this is what informed the thin driver. I would really like to hear this "fizz" because I am not sure we are all taking about the same thing. With a good battery my device runs very quietly with little to no distortion on low settings and still able to achieve infinite sustain. This is with a buffer only and a LM386 at lowest possible amplification (x20). That sounds reasonably efficient to me. What I do get is a fizzy distortion when the battery struggles. Tonight I have been playing a lot lately and the battery is giving out and so it is "fuzzy", turning up the gain only makes it worse. I suspect that a constant and high current remote power would potentially clear things up a lot, as would a new battery. Now, am I taking about the same fizz as you guys? Can you hear the fizz in those recordings I put up? These were made with a half dead battery perhaps, so I can hear it, especially on the lower strings and on long notes...or at least what I call "fizz", certainly some kind of distortion. There are lots of different flavours of distortion that can creep into the system and a lot of them sound very similar. The classic EMI fizz is generally not so subtle....this is the sound you get if you move the driver closer to the bridge pickup, just before it squeals. There also seems to be some signal distortion...where the drive signal somehow gets, (possibly also through EMI), into the guitars signal (a danger of the clipping preamp idea). But the idea of multiple coils is attractive and I encourage exploring that avenue. I suspect though that if the problems that exist are not directly related to the driver presently (circuitry or build quality) then those problems will be carried over into the new design and it will be very hard to tell if the new driver is at fault or you are simply not addressing problems that already exist. As I recall both curtisa and col built scratch single coil drivers, as did avalon before embarking down the dual coil driver road...with some success. They are trying to make a good thing better, I doubt a dual coil driver will solve some of the problems I suspect may be going on here...but I may be wrong, without hearing and seeing them it is pretty hard to tell. Still, for me the switching integration aspects are probably of more importance right now. I have been doing a lot of playing the last few days and I really am missing the variety of sounds available from different pickups, sometimes a hot bridge HB is just too much unless all you do is rock! The sounds I am seeking are pretty clean in general and the sustainer is a boon for this and a unique sound unobtainable in any other way. You get the sustain of a loud amp and the articulation, tone, dynamics and detail of a clean sound...power and grace!!! Also, if you don't hold onto the note, you wouldn't know the sustainer was there, it only becomes apparent when you realize that the note wont decay or is actually increasing in volume. Also, techniques like tapping and sliding notes are easier with the sustainer helping to drive these notes making for new techniques and making some of these "trick" techniques a lot easier to perform. Some of these techniques you simply don't seem to hear except with lashings of distortion. My temporary solder joint broke before I had a chance to fully test it, but it would seem that connecting the middle pickup and simply reselecting it may work which would be good. The neck pickup then is likely to be the main if only culprit as to switch noise. Perhaps the parallel driver and pickup combo will aid in addressing this as it would seem that all pickups and the driver will have a permanent connection to ground with this new configuration...quite a big change really...but we will see! I built the power box lead, but have not got around to wiring the guitar for it. I was hoping for a scheme that would allow for both power and battery use would come along...it is cool to be able to have the thing run on batteries if recording the thing for instance and not using an amp. Speaking of which, my newish BR600 arrived today so once I have got a good sense of all its functions (the manual is bigger than the machine!) and got a bigger compact flash for it, I should be able to apply the device to some actual music and use the practical application of it to inform future directions in performance. As I say, I think that there is more to this technology than infinite sustain and my aim is not only to make a different sustainer but one that takes the concept of electromagnetic string driving to a different place. That is in part why I resist going down the sustainiac road despite it's admirable design features as I think it could go a bit further on and achieve it by simpler means and in so doing, remove the sterility of the sound of the commercial units which have a kind of "compressed" sound to them and kind of characterless...but then, perhaps I just needed more time with it or a different installation than the one I saw and played. Still, a pretty rare beast to find down here. Anyway...carry on. pete Link to comment Share on other sites More sharing options...
col Posted March 4, 2008 Report Share Posted March 4, 2008 I have a question that may be stupid. Why do you need to use a magnet at all? If you rap a piece of iron in wire and pass a current through it you already have a magnet so surely using another magnet is jsut complicating things or is it the reaction between the two magnets that creates the sustain in the string? It's not a stupid question at all - I think most of the imaginative folks who have come to this project without an existing understanding of the physics of magnets have asked exactly the same question. The answer is (at least part of it - I don't remember/understand the physics beyond this point): Magnetic domains! (look it up on wikipedia or something) For a piece of iron to be attracted to a magnet, the magnetic domains in the iron must be aligned - the stronger the magnet, the more of the domains become aligned, and the stronger the attraction is. For an electromagnet to provide enough flux to do this by itself, you need a hefty coil and a hefty current. The trick with the sustainer is that a strong permanent magnet creates a field that can align the magnetic domains in the guitar strings, after which, a relatively weak electromagnet can 'influence' the intensity of the field and so cause the force on the string to be more and then less.... Basically, the permanent magnet turns a section of the guitar string into a magnet which allows the very weak electromagnet to have an influence on the string when it wouldn't otherwise be strong enough. To prove this to yourself, build a driver, then try it with and without a magnet Col Link to comment Share on other sites More sharing options...
joshvegas Posted March 4, 2008 Report Share Posted March 4, 2008 or alternatively hook up the sustainer to the mains 240v really would make an electric guitar! That maketh perfect sense I was having a drummer moment. thanks Link to comment Share on other sites More sharing options...
Fresh Fizz Posted March 4, 2008 Report Share Posted March 4, 2008 To prove this to yourself, build a driver, then try it with and without a magnet That was my very first sustainer project. I thought I could get sustain without a magnet. I used a single power darlington pinched off in class B, the driver coil as some sort of emitter resistance. It was a total failure. Lack of power, no decent sustain possible. Instead I got cricket-like sounds. Changing the bias of the darlington in class A didn't improve the performance. Except that my transistor and driver started to get really hot. FF Link to comment Share on other sites More sharing options...
psw Posted March 4, 2008 Author Report Share Posted March 4, 2008 Why do you need to use a magnet at all? It's not a stupid question at all - I think most of the imaginative folks who have come to this project without an existing understanding of the physics of magnets have asked exactly the same question. Quite right...but you may as well ask the same of passive pickups too, why do they have magnets. I kind of look at it as a kind of bias. There is a slight pull on the strings. The strings are vibrating within the magnetic field of the driver (or indeed pickup) and so this links them together. As the strings vibrate over a pickup's magnetic field it disrupts this field and causes currents to flow analogous to the strings vibration in the pickup coil. Now with a driver, the same applies in reverse. The strings are in the sway of an existing field, the driver coil is manipulating it, making it stronger and weaker and so moving the string. Without the magnet, the coil would have to establish such a field which would require a lot of energy. For the magnetic field to change states completely at the speed at which a string vibrates is a big ask. However, if you look at the drivers field as a kind of spring you can see that the driver is pulling on the field (making an existing field a touch stronger) and then letting it go (making it a touch weaker)...the string is already vibrating within a field and so all the driver needs do is manipulate it, without such a field it would need to create it (requiring a sizable amount of energy) and to be able to release it (remember how a nail attached to a magnet, or used as a core of an electromagnet will retain this magnetisim) which can also be quite difficult to loose at these speeds. When I discuss "speed" in a driver, I often mean the ability for the driver to change states (in reality a stronger or weaker attraction) efficiently. One of the reasons I believe that my thin coil designs have been successful is not only the low impedance variances that col surmised recently, but the inherent lack of permanence of such a thin core compared to a sizable piece of iron say. This is as much a reason for laminated or powdered cores in such designs and that I am not overly concerned that my pickup poles have less of a core (mainly they are air!) than previous blade designs. I have previously had a lot of success with ferrite cores and nearly all of the HEX designs had ferrite and epoxy iron composite cores and outer shields as well...they took inspiration from radio inductor coils and the like. Anyway...I hope this might aid further in the explanation. This is one of the most common questions as col points out, but understanding the fundamentals and even repeating them for ourselves is always a worthwhile exercise. This device not only makes one consider these things, but shows how we take for granted such things in pickup designs. All the ideas here in terms of drivers have implications to pickup designs also. For instance, when we think of a "hot" pickup, we often think of increased windings...more windings, more power. However, stronger magnets (within reason) have a similar effect. More core material and the type of material also has an influence. It is interesting to see that some more modern pickup designs are experimenting with these attributes, many of the new generation of stacked pickups have extra poles within the lower coil and less windings for instance. Makers such as Steven Kersting (SKG) (who got me into femm stuff and has some cool pickup models in FEMM somewhere on his site) SKguitar femm pickups...illustrated for instance is the way baseplates effect the "shape" of a magnetic field. While a pickup and a driver are similar in principle, they are quite different in function and one that is good for one may not be good for another (despite my continuing efforts to combine the two). I was quite stunned at how good the sustainiac driver (which is very different from an average pickup) for instance sounded as an active pickup. So while things we consider here are principles equally applicable to pickups...there are other other "stupid" questions yet to be asked or strategies to be tried. Don't make the mistake in thinking that more power is a better strategy, or that more powerful magnets will create a better driver (or pickup for that matter)...the biggest thing I can advise is that this is an exercise in balancing forces, many of which we don't understand. Exotic materials are also not "required" and while I mention in this post ferrite cores, this material is almost impossible to work with...others have been very attracted to the idea of neodyminium magnets...but generally these things are far too powerful unless very small, and then these little things are highly attracted to themselves so don't have the throw of even a fridge magnet...but get to close and the string will bend out of tune or even stick to them. I have used them, very, very little ones (2mm round/deep) for specific designs but they fail as a substitute for ceramics and alnico magnets conventionally used in pickup designs... Anyway, perhaps even more than you needed to know, but as I say...sometime it is good to remind oneself before you make the same "stupid" mistakes. At first instead of asking this question, I took it for granted...now how stupid was that, I didn't really consider why even a pickup had a magnet or why different magnet materials and shapes influence the sound of these things!!! pete Link to comment Share on other sites More sharing options...
utopian isotope Posted March 4, 2008 Report Share Posted March 4, 2008 That got me thinking - what about some sort of U section plastic strip of some kind - you could use it to make something with reasonable structural rigidity, it would be slightly thicker due to some overlap, but I think that will only be important if your trying to build a piggy-back driver. A refinement if this kind of idea might do the trick ? cheers Col Link to comment Share on other sites More sharing options...
psw Posted March 5, 2008 Author Report Share Posted March 5, 2008 That got me thinking - what about some sort of U section plastic strip of some kind - you could use it to make something with reasonable structural rigidity, it would be slightly thicker due to some overlap, but I think that will only be important if your trying to build a piggy-back driver. A refinement if this kind of idea might do the trick ? cheers Col That came to my mind ,when I finished wounding dual cores. They look quite awful after all. I need some practising with those, so maybe I'll take your idea even further, and make template(s) out of nylon. As nylon resists powerglue and epoxy, it makes easier to snap finished coil(s) out of it. Anyway, lot of work needs to be done with the mill. This last idea is similar to how I make my bobbinless coils. One Warning...do not use superglue..it is dangerous when inhaled and is definitely not the right stuff. It dries too quickly and too hard and brittle sticking to everything. The result is that you wont be able to wind with it and although it penetrates reasonably in a surface, it will ste before it saturates a core. Since you are generating heat in your designs and any loose windings will vibrate (another source of heat and fizzy distortion by the way) you know that your copper coils will expand and contract...the result will be potential and eventual stripping and shorting of the enamel coating protecting the wire itself. 5 minute epoxy might work, but also has it's draw backs. You don't really have 5 minutes, as soon as the parts are mixed they start to gel and so you perhaps have 2 minutes and then once set, the coil will not be able to "settle down". Winding tightly onto an elongated coil former will cause some stretching (and relaxation of the sides) that creates a considerable internal pressure. Even winding on my machine with a counter and a proven method, this stuff will not work for me. Instead, I use a slow acting (perhaps 8 hours plus) more liquid type of epoxy. Remember that without a bobbin or a core the coil will need considerable strength it it is not going to turn into a copper banana if not completely self destruct...do you recall this photo... These are just some of the many failures I endured with the bobbinless driver coil alone with my present design. However, I am fascinated as to how you "mill" this kind of thing. I found it very difficult to make mine and with it's slow build time I only really have one that works to specifications. Is this something you are skilled in, can you give me some hints on how you would build such a shape? What kind of tools would you use? People may wish to look at the method I made the original sustain-o-caster driver pickup which I was (during happier times) able to do a step by step photo guide as I made it. I used a thin plastic and slit and folded the centre section and used the small folded down edge to mount the top plate onto the core with PVC tape. I strengthened this while winding and the setting of the PVA potting on this coil, and while it settled down. If I were to make such a design I may be tempted to follow some of these methods but more likely use a powdered pure iron core poured into the coil...but then as a one off, I would be less likely to do the bobbinless coil as it is a lot of work, perhaps then I would use some kind of molded core and wind around this. A fault that did occur with my present design is that if there is any chance of movement, the coil and or core will try and vibrate. This caused rubbing between the core and inner windings eventually shorting it out and resulting in the symptom of sparking and squeal if a pole was touched. I have since needed to guard against this by sticking it down with tape. Anything that has this potential (the ability to be removed) will require considerable strength to counteract the internal forces of the coil and close tolerances to ensure close contact with the core without movement. That came to my mind ,when I finished wounding dual cores. They look quite awful after all. I need some practising with those, This is my point about build quality. It may well be that your present design ideas are being let down by the initial build quality rather than the design itself. That is not to say that the thing will necessarily need to be pretty, but some things are important...no loose windings, total saturation with as little air as possible. Anything less will result in inefficiencies and possibly even create signals. When I say loose windings by the way, any coil, especially hand wound has the potential to move under these kinds of forces and any air will create inefficiencies. My new coils designs are possibly as good as they are simply because of these build factors, they are absolutely rock hard and every bit of excess glue is squeezed out with any air gaps within under considerable pressure and time is given, and lubrication of the slow epoxy used, to allow the coil to settle and squeeze into it's final shape. According to col's posted coil calculator, there is no way that my coils should fit into the space they do, one reason that they do is that even some of the rounded shape of the wires themselves are deformed to fit with each other and the epoxy is able to heal any coating rifts on the wire itself before it sets. This kind of thing can only be achieved with total saturation and chemical bonding. Tp achieve this result took week and weeks and many many dollars and time with different epoxy types before I found one that would work. Making the former that builds them went through a lot of evolution too. You do not need to go to such extremes to get the device to work, hand wound coils are fine, so is PVA and some filled air gaps between windings is likely but not that bad. Fancy laminated coils will not have that big an impact I suspect and certainly wont overcome badly wound coils for instance. May I suggest that you try a basic design with a basic core and try to do it as simply and accurately as possible, if only for practice. Such a coil is in fact likely to be "pretty" as this is a sign of care and attention being put in to it and producing a good result. You could then plug it into your present system and see if it isn't a lot better. From there perhaps refine the circuitry, perhaps switch to a more conventional lower powered LM386 type circuit...easy and cheap, you may even find a kit for the kind of thing you are after. For instance, I have always wanted to try this which looks promising, perhaps with a few modes for a DIY sustainer LM386 canakit with preamp assembled even! however they wont sell to Australia these things...if fact, someone sell me one to try!!! After you have this working is the time to get fancy with the designs I would suggest. It is all a matter of finding the best balance between efficiency and power and a build quality sufficient to give it the best chance of success. Anyway, I hope these post help in your goals. pete Link to comment Share on other sites More sharing options...
col Posted March 5, 2008 Report Share Posted March 5, 2008 Here's a what if brainstorm kinda idea for you all to ponder: We've talked about about distorted fizz getting to the pickup from the driver via flux and/or eddy currents in the strings. Also assumed that if this a problem, then it will be tricky to fix. We've looked at the stainiac literature. One of the things that seemed odd to me is that little 'tab' thing that they provide for removing fizz. They suggest that for some installations, that little tab can be placed on the driver and if adjusted correctly can help remove fizz. Doesn't it look a bit small to be any use as a shield or as a 'flux conduit' ? Heres the silly brainstorm: What if that little tab is a little EMI transmitter actually broadcasting its own interference signal... and if its moved to the 'correct' position, can broadcast EMI that cancels that from the strings ? cheers Col Link to comment Share on other sites More sharing options...
psw Posted March 5, 2008 Author Report Share Posted March 5, 2008 Doesn't it look a bit small to be any use as a shield or as a 'flux conduit' ? As far as I could tell from stuff I had seen and in particular the patents...these things seem to be aimed at narrowing the energy field on higher strings so as not to pass too many nodes on these faster and shorter waved string vibrations. It was something discussed a little way back when were debating whether a narrow core would be superior to a thick one such as the sustainiac appears to have. I tended towards the narrower side of things settling on about a 3mm blade style by then. I used some of this in support of this. There was a thought then that perhaps the core should be asymmetrical, being wider at the bass strings (and there may be some merit in that) I tended to stick to the single width thin blade arguing that if it were less efficient on the lower strings that this may be a good thing as these tend to be easier to drive and so may create a natural balance and be a lot easier to make and to wind. I also felt and still feel that bulky coils will be less fast and require compensation for lag as described in the patents for such devices. Although my cores are now based on conventional poles, within the coil itself I have mostly air and so the principle of low bulk cores still applies and it is working well for me. Doesn't it look a bit small to be any use as a shield or as a 'flux conduit' ? When I have added such things, I have had little discernible effect, although it should add to impedance like an internal core I suspect. The mid-driver had two internal blades and two external based on the same idea. Here is a pic of another one with a similar addition... A Kramer as I recall...notice the thin metal on the lower portion of this twin blade rail sized driver. Words like fizz and grunge are not that specific and I still wonder whether we are talking about the same thing and if there is one simple cause. I do think though that the strings are the most obvious conduit for such distortions to reenter the signal chain and generally falls under the heading of EMI. Some of these sounds on my system are alleviated by more power, so I am questioning whether the assumptions we are making are entirely correct...hmmm However... Heres the silly brainstorm: What if that little tab is a little EMI transmitter actually broadcasting its own interference signal... and if its moved to the 'correct' position, can broadcast EMI that cancels that from the strings ? If the brainstorm is, is this what sustainiac are doing...it would appear not as the patents clearly show that this is a simple metal plate designed to bring the field more into focus on the high e and b strings. If fizz is a generalized condition, bringing this about only on these two strings would be ineffective, my feeling is that they are there to aid with high string response, not noise reduction. If they were effective in that way, then I would expect more along the lines of the above kramer model and protect the whole driver from stray EMI, not just a couple of optional tags on the high strings. It was interesting that the sustainaic I saw did not have these things fitted and worked pretty well. It was also fitted to a 24fret Jem copy, so pickups were closer together than some. As a side note, on my test guitar, I fitted a middle pickup although presently not connected. I was interested to see how this might affect fizz and the like. I figured if a little shield or tags help...imagine the effect of addint a whole array of magnets centred between the driver and bridge pickup...hard to say if it really had much effect, but you would perhaps expect it to have a little more affect than a little metal shield. Maybe it was because it was working already that little was gained, maybe the additional coil in there negated some effect...very hard to tell. So...I don't think you will find that this is quite what these things are, and I know you have already experimented with and proposed some shielding strategies. I had kind of thought you had removed and abandoned these for now...I know I tried a few things but now have put them on the back burner. The last one I tried was very much in line with this idea. I had the bulk of the wind going one way around the core, then for the last 30% I wound the opposite direction. So effectively a series dual reverse coil or an "active shield". The idea is exactly as you propose, an opposing EMI transmitter surrounding the whole coil. It is hard to tell the results of this, the coil was built fine and I still have it. It seemed to click less in a direct comparison with my conventional coil, but it lacked the power and high string response of the normal coil. The idea is not a failure though, but in my ultra thin coils and with this kind of setup it is perhaps not an appropriate design to take full advantage of it. It could be seen to be a scheme that wastes power as the shield presumably does not add to drive as a dual coil/core driver does, but it didn't seem to be detracting any more than the 30% of coil dedicated to it (in other words it wasn't 60% less efficient, it did actually work). I think on one of the many fernandes related patents such a scheme was proposed but I have never seen it, or a lot of other proposals, featured in the commercial products. Another interesting idea were the internally slotted blade cores of some proposals...giving more core to different strings, again, more of a string balance high string response thing than EMI reduction. However, if such things aided in the efficiency of the device, it would reduce fizz by being able to run the thing at a lower gain. If this is how it is described as reducing distortion, I suspect this is why. If narrowing the core gap on the higher strings improves high string response, then naturally, less gain can be used for the same effect resulting in less EMI output. Again, it comes down to balance and efficiency. That is why I get a little worried/critical when I see people going for 2.5 watt amps and such...Me and col appear to be running at perhaps 0.5 watts or less and with mine, no preamp gain...this appears to be plenty of power (with an HB as the pickup anyway). The only reason I contemplate more power would be to maximise clean headroom, but now I am thinking that a lot of this has to do with the power supply potential...as my battery runs low, the amp distorts more and I get more fizz type effects, that is the only way to describe it. I have run the battery so low that the LED will actually dim to nothing when sustaining strongly, such is the power drain...and yet it continues to sustain, albeit with increased distortion. Evidently the effect can be had with very low power, but a clean effect seems to require a clean drive signal. The other reason that perhaps you might want to increase gain is to produce harmonic effects. It would seem that fresh fizz's is doing this with clipping and gain...mine does it too at higher gains (but I can't produce that much) and col has suggested some saturation effects that may explain them in another way. These things would be nice to explore perhaps. Anyway...that is some of my ramblings on the proposal col...correct me if I have gone off on a tangent. evening all... pete Link to comment Share on other sites More sharing options...
utopian isotope Posted March 5, 2008 Report Share Posted March 5, 2008 (edited) The last one I tried was very much in line with this idea. I had the bulk of the wind going one way around the core, then for the last 30% I wound the opposite direction. So effectively a series dual reverse coil or an "active shield". The idea is exactly as you propose, an opposing EMI transmitter surrounding the whole coil. Is it possible that certain amount of opposite winding has the effect similar to bifilar coil? Cancelling some part of the signal? -------------------------- However, I am fascinated as to how you "mill" this kind of thing. I found it very difficult to make mine and with it's slow build time I only really have one that works to specifications. Is this something you are skilled in, can you give me some hints on how you would build such a shape? What kind of tools would you use? I have a small "3 in 1". Means lathe, mill and drill in the same machine. One disadvantage of using mill is, that cutter(s) has to be razor sharp, when machining nylon(aka Polyamide/Ertalon). Most challenging part is the centerpiece. It can be made without mill, if you have any kind of grinding/sanding machine available. Even with humble power drill and sanding drum, it can be made. But the nylon blank needs to be attached firmly to... -- shame on me as I don't know how to explain it in english. Have to make some animated gif's to clarify the idea. Edited March 5, 2008 by utopian isotope Link to comment Share on other sites More sharing options...
psw Posted March 5, 2008 Author Report Share Posted March 5, 2008 That's ok IU...I tried making a jig for a router so I could make multiples and so make more than one coil a day. You do need a sharp bit and it needs to be cold as my stuff will melt. Also, I had to be real accurate. I found a thing called emachineshop which is a supercool 3d softawre for making parts in just about anything, then you can get a quote online for it being made by various means (CNC, Drill press, sheetmetal folding, etc) and quantities. For one of these kind of things it costs about $100, for 30 about $15 each. A fascinating bit of software, very easy to use though...very much like google sketchup. Making my own of various designs sometimes took all day, I made about a dozen various designs before I got one quite right. I still have to spray it with silicone so that the coil doesn't stick a little too much and I still would not be happy winding with epoxy by hand. You will have to apply the epoxy during the winding process as there is no way to get the saturation required to hold it all together otherwise. You can apply it inside the bobbin and then wind into it and the glue squeezes out through the windings as you go, but otherwise, it wont have the strength. For one off kinds of things, bobbins are still the way to go. You could even consider getting a hold of commercial pickup bobbins and adapting that. I do know a bit about epoxy casting and can make simple molds but I have never been able to workout how to do a bobbin shape suitable, always I get some kind of air bubble or something. Is it possible that certain amount of opposite winding has the effect similar to bifilar coil? Cancelling some part of the signal? Well, this was the "risk" but unless you try. I had always wanted to do this but way back when I had tried similar things before I had not built them very well (sound familiar) and so could not really tell how it really performed. My first version of my coreless winding bobbins was a little like what you have drawn, the idea was that I could make coils of variable depths. I was unable to make something suitable that worked...they need to be very accurate...the idea being I could experiment with various depths and make various coils, perhaps wound to spec for members here. That didn't quite work out with the present design...however I did feel that before I went to far, I should try a version of this with my present design. I got the guitar working with a very thin coil and then wound an identical spec coil but with a reverse winding after about 70% of the winds. I did it by counts, not reading (epoxy is a little messy and you have to work pretty fast) and took a final reading of the total coil. I then swapped out the conventional coil for this one so I could get a direct comparison. It did have a reduced effect, similar perhaps to the coil if it had been wound to 70% I suspect. It did work, but lacked the power of the conventional single direct coil. However, it did not seem to be canceling the drive as far as I could tell. That is, it did not seem to be subtracting from that 70% of inner winds around the actual core...it didn't seem to be working at 40% or anything. The idea was that these outer winds would be cancelling in EMI but further from the core and so not operating on the strings but forming a focusing shield that pulsated exactly as the inner coil and core. My coils though are only 1mm thick... The design is so different now from the sustainaic or conventional pickup coils that it is pretty difficult to compare them except with similar type designs. The kinds of things you guys are contemplating may well benefit from this kind of thing...I kind of thought that this was perhaps the kind of thing col was suggesting with his brainstorm... What if that little tab is a little EMI transmitter actually broadcasting its own interference signal... and if its moved to the 'correct' position, can broadcast EMI that cancels that from the strings ? Something else that I had often considered but never tried...Litz wires. These are multi strand, usually twisted wires in the coil...eddy currents effect efficiencies not only in the mass of the core but in the coil itself. One technique is to have multistrand core wires. While such things would be very expensive to try, you could DIY this kind of thing by winding with multiple strands or winding off a few suitable lengths and twisting them together before winding. I have spent so much time doing these kinds of things in the past, that I really have to stop and concentrate on making a good thing better. You don't need to go to such exotic lengths to make something that works! A hand wound PVA potted coil on a bobbin with a magnet below does work quite well. How much of a payoff will you get from a laminated core or litz wiring I am not really sure. As for my own direction, it would seem that by reducing the original "thin driver" design (3mm deep) by two thirds and improved building techniques (epoxy and pressure, plus machine winding) to this "ultra thin" design (1mm deep) has made quite a bit of an improvement. I have various theories for this and others have offered some explanations, the most recent being col's suggestion that I have got the thing to work largely at resistance with this design and avoided a lot of the impedance issues. This may be so. One thing that drove the original and this design was the idea of concentrating the driver coil energy as much as possible directly under and as close to the strings as possible and in doing so, maximize the action on the string and limit the spread of EMI in the process. A thinner coil will have a smaller core mass (although perhaps debatable as the thing is mounted onto a "deep" pickup core") and so avoid some of the eddy current issues of bulky cores and such and work "faster". I also felt that working on the very tip of the magnetic core, this was having more of an effect than had the driver been wrapped around the entire core...the very edge has interesting qualities, it is where the field leaves the core and so curves back around to it's opposite pole, dissecting the strings as it does. Also this point is the furthest part...the most north or south, of the core and so all the energy is applied at this point. All of these "theories" seemed to be worthy of consideration and informed the design. The idea of mounting it onto a pickup came later...a happy coincidence. Such a scheme supplies the magnets, core and mounting and it appears to have very little if any effect on the normal working of the pickup itself...very cool! It is not really a DIY thing though as the bucket of failures show...it took a lot to be able to get them to this size. These coils work just as well with a core and a magnet applied to it like a stand alone driver. There are some secondary coil effects that were an unexpected consequence, but there is always something in any great plan!!! As it is, this design is very close to what I set out to do, it does work surprisingly well. Now, the question is, how much better would it be with the kinds of variations suggested of late, multicoil drivers and such. I suspect not that much, though I could be wrong. Would it be so much more dramatic an improvement as to run on significantly less power (I doubt it) or so EMI reducing as to allow far more gain and so a more powerful effect (also debatable)...would it enable the location of the driver between the neck and bridge pickups allowing the use of either/and in sustainer drive (something dizzy achieved it wuld seem with a bi-lateral driver)...I have not been able to get this design to work closer than the thin coil version. So, what is the payoff, probably a potential slight refinement or improvement. Meanwhile, the other problems remain...switching and integration into a conventional multi-pickup guitar. These things would remain regardless of the efficiency of the driver and so I am loath to pursue this aspect until these issues are addressed. Similarly, there may be gains in tweaking my circuit or designing another...but while this one is working, there seems little point in tweaking this while there are installation issues still pending. People with single pickup guitars need not worry however, the thing switches silently in such a scheme and is pretty easy to do. Harmonic mode does seem to have gone out the window replaced by harmonic blooms with increased gain. On this last point, col might find this interesting. I get a fairly consistent harmonic drive, even at low gains on the high e string above the 17th fret. Playing here, the string is pushed by the action very close to the driver (and the pickup) yet the same is true with all strings. Col had an interesting idea about saturation and I was wondering whether this may relate with the high e being so thin and close to the device. There are of course a lot of other factors. You can get these notes to sustain normally but you need a light touch and low gain...but held long enough it will almost always end up a piercing harmonic. Anyway...enough from me...have an nice day/night wherever you guys may be. Oh...speaking of which, where are you guys, I know col is in the UK, I am down in Melbourne Australia, Curtisa is across the water even further south. I do feel guilty about my rambling posts for people for which english is not native to them, even english speakers have trouble with my convoluted conversations...good work getting anything out of my posts!!! I am happy to try and simplify anything that people may be struggling with, no matter the language if that would be of use. pete Link to comment Share on other sites More sharing options...
col Posted March 6, 2008 Report Share Posted March 6, 2008 ...it would appear not as the patents clearly show that this is a simple metal plate designed to bring the field more into focus on the high e and b strings. If fizz is a generalized condition, bringing this about only on these two strings would be ineffective, my feeling is that they are there to aid with high string response, not noise reduction. From the Sustainiac installation guide: "DETAIL OF MAGNETIC TAB INSTALLATION: This tab is not needed unless grunge and/or squealing is present when you turn on the sustainer....Set the instrument amplifier to a clean setting, and turn up the treble so you can easily hear distortion. Move the tab is to the position where noise and squeal is minimum." So...I don't think you will find that this is quite what these things are, and I know you have already experimented with and proposed some shielding strategies. I had kind of thought you had removed and abandoned these for now...I know I tried a few things but now have put them on the back burner. I have experimented with 'shielding strategies' although I don't like the term 'shield' as it is a red herring. My first attempt at enclosing the driver in a metal 'cage' failed, but I worked out why, and it was a silly mistake (I've explained this a few times now). More recently, I have improved a single coil driver by adding a 'cage'. It had no loss in drive power, but significantly reduced squeal and noise. If you use this approach on a humbucker style driver, it does remove much of the flux from around the strings. this is probably what the Sustainiac tab is really for (if it was bigger, it would kill the sustain). As my femm recent animations show, there must be quite a large gap between the two cores on a bi-lateral driver for an even field across all strings. I guess that the small tab 'sucks' just enough flux to reduce the peak that will exist in the sustainiac driver with its smaller gap between cores. The last one I tried was very much in line with this idea. I had the bulk of the wind going one way around the core, then for the last 30% I wound the opposite direction. So effectively a series dual reverse coil or an "active shield". The idea is exactly as you propose, an opposing EMI transmitter surrounding the whole coil. Adding reverse windings to the coil is not the same idea at all. The reverse windings will cancel some of the noise, and some of the signal, and they do it by creating an opposing field. The tab most likely works (in the sustainiac case) by re-routing the flux slightly away from the strings - it doesn't in any way reduce the power of the magnet. With my 'cage' idea it works (and it does work, though only on a single core driver) by focusing the field on the strings. Going by your 70% - 30% windings your driver would have something like 40% (70 - 30) of the magnetic force of one with 100% of windings going the same way! No wonder it lacked the power and high string response of the normal coil. You do say that it was not 60% less efficient, but its hard to believe that it was only 30% down (thats not even possible IMO). I reckon if you were to set up an experiment and MEASURE the magnetic field above it, you would be surprised at how much you were losing.... It would be a waste of time though - basically, it doesn't work, so why use more time up playing with it. Another interesting idea were the internally slotted blade cores of some proposals...giving more core to different strings, again, more of a string balance high string response thing than EMI reduction. I think this is an area that has some promise, and is one of the few places where being DIY gives us an advantage. For a commercial product, the cost difference between bog standard laminated core and fancy-pants custom stepped width core will be significant. For us, depending on how we go about it, it needn't be any extra work or cost. However, if such things aided in the efficiency of the device, it would reduce fizz by being able to run the thing at a lower gain. If this is how it is described as reducing distortion, I suspect this is why. If narrowing the core gap on the higher strings improves high string response, then naturally, less gain can be used for the same effect resulting in less EMI output. There are possible downsides that maybe the reason why its never gone into production. The strings you need the most power on are the higher ones. Addin more core (or less gap?) to those areas may increase the effective inductance for those strings, in turn reducing the frequency response - just where you need it to be good!.... It's another area that needs R & D and is probably 50:50 whether its a complete waste of time or not. Again, it comes down to balance and efficiency. That is why I get a little worried/critical when I see people going for 2.5 watt amps and such...Me and col appear to be running at perhaps 0.5 watts or less and with mine, no preamp gain...this appears to be plenty of power (with an HB as the pickup anyway). Totally agree about balance and efficiency - as far as 0.5 watts, I seriously doubt that. The only way to know for sure is to measure the current at the battery, I'd be surprised if you were getting good drive with an LM386 just below clipping without using at least 1.5W probably more! (thats output + dissipation losses) I'm actually concerned about the safety of using PP3 batteries, as most of them are not designed to safely provide anywhere near as much current as we are expecting from them. I would suggest that it is a good idea to source 'alkaline ultra+' extra super duper stylee batteries that are specially designed for high drain applications - don't want any nasty leaks(or worse) in our axes after all! I wouldn't be surprised if Sustainiac went for class-D partly because of safety limitations. cheers Col Link to comment Share on other sites More sharing options...
col Posted March 6, 2008 Report Share Posted March 6, 2008 Hmm, the Sustainiac driver is looking better and better - here's another factor that I hadn't considered: orientation of magnetic circuits. It is well known that to avoid interaction between transformers, they should be placed with the fields perpendicular to each other. I believe this is the reason that fernandez attempted to put the driver coil on it's side in one of their earlier systems. Looking at the sustainiac system, although the driver as a whole, and the windings are in the same orientation as the bridge pickup, the magnetic circuit is not. If the axis are: x => along the neck y => out from the body (the way a pickup pole piece faces) z => across the body Then the pickups magnetic circuit is in the xy plane, while the drivers circuit is in the yz plane. Does this mean that there is less interaction between the permanent fields of the pickup and driver ? If so, then the same should be true of their electromagnetic fields. I want to make a bi-lateral even more now Col Link to comment Share on other sites More sharing options...
ghendrickson04 Posted March 6, 2008 Report Share Posted March 6, 2008 (edited) Col, as an alternative to what your suggesting as far as increasing core density to alter inductance around specific strings, would it be efficacious to split the top bobbin, probably between the pole pieces for the D and G strings (but maybe G and B ), and then wind two coils with different resistances adding to ~8ohms? Edited March 6, 2008 by ghendrickson04 Link to comment Share on other sites More sharing options...
psw Posted March 6, 2008 Author Report Share Posted March 6, 2008 (edited) Going by your 70% - 30% windings your driver would have something like 40% (70 - 30) of the magnetic force of one with 100% of windings going the same way! No wonder it lacked the power and high string response of the normal coil. You do say that it was not 60% less efficient, but its hard to believe that it was only 30% down (thats not even possible IMO). I reckon if you were to set up an experiment and MEASURE the magnetic field above it, you would be surprised at how much you were losing.... It would be a waste of time though - basically, it doesn't work, so why use more time up playing with it. Well...I can't quantify it with measurements this is true, but I did use the same guitar, core, magnets, circuit and pickups and simply changed the coil one for the other and back again...this is a pretty good test in the circumstances. I built it not as a waste of time but on an impulse, I did come up with this and have mentioned it a couple of times over the years but not really been able to do it, I now had an opportunity to make something using the exact same processes and so, very little time (other than the day to cure it) was wasted in trying it. It is also suggested in one of the patents, confirming my thoughts that this was not such an unreasonable strategy to try. I agree that it may seem that a massive loss of power would result, but all I can say is what it appeared to me to have been the results of this test. After testing, the coil was removed and has never been used again. The reasoning that a massive loss of power may not have resulted was that the outer coil has effectively no core. From my observation then, I suggest that this seems to be in part the case, but this was a simple test, done and gone. I have given myself quite restrictive constraints to work with, at this point a single coil size and magnetic field and retaining passive pickup capability. As this is my aim with this particular model, this idea was one of the few that I could try easily that varied my design as it stands. I appreciate that others don't have such restraints as this or my wish to have multiple pickups in a guitar out of sustainer mode, but I am only relating what I have tried and what I have gleaned from the many patents on the subject...this apparently being yet another one. Hmm, the Sustainiac driver is looking better and better - here's another factor that I hadn't considered: orientation of magnetic circuits. It is well known that to avoid interaction between transformers, they should be placed with the fields perpendicular to each other. I believe this is the reason that fernandez attempted to put the driver coil on it's side in one of their earlier systems. Absolutely...it would seem from patent descriptions that this is the primary reason for the design. It was a major factor in the HEX designs and definitely an elegant approach to this. The Hex ideas used six opposing tiny driver fields using 12 magnets to aid in string to string isolation and many manipulated them primarily on the Z axis. These too will work very effectively...there are so many options depending on how far you may wish to go with it. So...I am not at all knocking the sustainiac approach. You and everyone who is inclined should build one immediately. I really don't know why I haven't simply done the same in the past, but if I were not committed to my own design restrainants and motivations, that is exactly what I would do! Dizzy for instance built an almost exact copy of it and the compensation circuitry of the sustainiac and had it working in the middle position of a strat as can be heard in the sounds post...very, very effective. Unfortunately I didn't think to copy and save a photo of the thing at the time and it would seem to have disappeared into the cyber-ether, but it was very similar and very well built and thought out. Circuitry remained secret but quite a bit of effort was put into the phase compensation circuitry as I understand. I built my own model of a three blade side by side device but interestingly it worked better on it's side in more of a stack. It was actually very effective in this format, unfortunately due to handling, a critical wire broke and now is just a lump of metal. Z axis drivers are a completely different beast with some very unusual characteristics especially related to alignment and physical problems of the core presented to the strings being completely covered by the windings with conventional coils. QUOTE(psw @ Mar 5 2008, 01:31 PM) * ...it would appear not as the patents clearly show that this is a simple metal plate designed to bring the field more into focus on the high e and b strings. If fizz is a generalized condition, bringing this about only on these two strings would be ineffective, my feeling is that they are there to aid with high string response, not noise reduction. From the Sustainiac installation guide: "DETAIL OF MAGNETIC TAB INSTALLATION: This tab is not needed unless grunge and/or squealing is present when you turn on the sustainer....Set the instrument amplifier to a clean setting, and turn up the treble so you can easily hear distortion. Move the tab is to the position where noise and squeal is minimum." I know what the sustainiac instructions say, and what their patents illustrate, but my only point really is that if these small single string tabs have such an impact you would expect then to want or even require it along that whole pickup side as illustrated by this kramer driver... But hey, if it works, then do it I say. My present approach is to work a good design (the thin driver) to get as much out of it as I can and maximize it's attributes and performance (ultra thin driver). I am not advocating it as an ultimate driver or even suggesting anyone should build such a device themselves, in order to get it this thin is physically difficult, but I have learned a lot that could be applied to any design in it's construction that I am willing to share. Totally agree about balance and efficiency - as far as 0.5 watts, I seriously doubt that. The only way to know for sure is to measure the current at the battery, I'd be surprised if you were getting good drive with an LM386 just below clipping without using at least 1.5W probably more! (thats output + dissipation losses) I'm actually concerned about the safety of using PP3 batteries, as most of them are not designed to safely provide anywhere near as much current as we are expecting from them. I would suggest that it is a good idea to source 'alkaline ultra+' extra super duper stylee batteries that are specially designed for high drain applications - don't want any nasty leaks(or worse) in our axes after all! I wouldn't be surprised if Sustainiac went for class-D partly because of safety limitations. Well...I guess I don't know the draw. My impression was that a LM386 with an 8 ohm load and a 9 volt supply had a typical output of 700mW at 9 volts into an 8 ohm load. Given that you and I have both reported that we can run the sustainer with minimum gain (pins 1 and 8 open), then I could only imagine that it may be less, and give the restraints of a battery supply, likely to be on the low side...perhaps 0.5 W. LM386 data sheet Correct me if I am wrong. I was comparing this to IU's amp rated at 2.5 watts and I think fresh fizz is running something similar, both using power supplies with huge current potentials compared to our battery operated versions. Mine has run on ordinary batteries but they wont last that long (a day or so of intermittent use) while I have tried a few and generally do use alkiline and even the ultra+ lately too. I even dragged my very old rechargeable out and that is working fine, but runs down pretty quick. At over $5 a battery, this can get expensive, and so the attraction of outboard power. Power does effect performance in response but mainly in increased fizz/distortion as it runs flat. As I have reported, I have run my present setup till the LED goes out in operation and still get infinite sustain, but plenty of grunge. Hence I thought that performance and fizz may be impacted if a more "reliable" power source was available. Also, the cost in batteries is significant and with my set up, the battery is by far the biggest component of the entire thing. I finally got a little time last night to redo the switch ideas. I was too slack to get the computer out and so redrew it partly from memory, partly "reinventing it" on paper. The new scheme that resulted is simple and elegant using a dpdt switch and the two sides of the selector to separate the bridge and other pickups. It was only possible because of the new ideas of parallel pickup/driver combinations...again, thanks fizz...but only building it will prove it...maybe later today. I used the -ve switching again as this saves a whole switch, I connected this to the negative side of the driver so that when the driver is combined with the pickup, it simultaneously applies power. Hopefully, the constant ground will help in the switch clicks...otherwise, these problems will still exist. One problem with negative switching though is that the idea of using a common ground power supply through the stereo lead would not work as the circuit would see power regardless with such a scheme and be always on. I imagine the only way around this would be to add an additional switch (3pdt) to positively switch the power separately. Or, perhaps to make a different power distribution option. This mornings feeling is that perhaps this is a better option. Keep the battery and have a 4 core lead with a separate plug or multi pin DIN or the like so that the signal and power remain separate. Can someone tell me though, what would be the effect of running the device from an outside source with an internal battery onboard and connected also.? Would it damage the battery, run it down anyway??? Certainly you would not want a leaking battery in a guitar, but I have never had any hint of this over the years...with this scheme, I really dont know! It is simply amazing how we can be discussing so passionately and for so long the concepts of this project. Obviously there is a lot we don't know and so many variations to consider and try. I hope that we see a bi-lateral driver type very soon, maybe several as this is a major deficiency in the research and development of the DIY sustainer and this thread in general. I wonder if anyone out there has a fernandes and can take a peak at their driver...theire newer models seem to be a little bit of a mystery...hmmmm pete EDIT I did a quick look at the sustainer site of fernandes, seems to have improved since I last looked at this stuff. Here is a pic of their latest product as fitted to their guitars... and here is a link to the site, this is the sound page where you can hear the thing...fernandes site, demo page It occured to me that one of the easiest ways of making a good looking bi-lateral design would be to get a pickup bobbin with separate poles and cut it in half with a sharp knife. Reverse the 3 poles on one side and then wrap your driver (perhaps blocking it up to make a thin driver and then sticking the two halves/coils back together with superglue...put a cover on it and you would never know. Of course this would not be a pickup anymore, I am not even sure if you could put a pickup next to it, wouldn't the differing magnetic fields adversely effect it? However, it would be a neat and easy way of doing it...just a thought! Edited March 6, 2008 by psw Link to comment Share on other sites More sharing options...
utopian isotope Posted March 6, 2008 Report Share Posted March 6, 2008 Totally OT again... One hint when rounding ends of the centerpiece(s) and bobbins/templates with sanding drum: Just a tip of an iceberg... If there is enough interest, I can make more animated stuff about it, as there is also some other things that can be done.. It really isn't beginner's project as it requires almost an extreme accuracy. Link to comment Share on other sites More sharing options...
psw Posted March 6, 2008 Author Report Share Posted March 6, 2008 Ah...a topic I have a vested interest in...that is great...has given me a few ideas. What I had tried was using a router on a jig. You have to watch heat as a lot of these plastics will melt with tooling...hehehe. What I hoped I could do, when the problems are sorted and the design concept and product is complete, is to have a number made with a CNC so they can be made on mass. On the above post, I added an edit about fernandes. Very few clips of either the fernandes and sustainiac are about that are clean and so offer a way to evaluate the "fizz" grunge aspect. Here is fernandes clean demo... http://www.fernandesguitars.com/sustainer/...o/COUNTRY_S.WAV It may be my speakers (the computer runs to a mini-hifi so it is better than typical computer speaker setups) but to me, I hear fizz exactly like mine...a kind of fuzzy distortion after the note is held for a little while. Back to UI...I may send you some details of the kind of jig I use now, perhaps you could offer some tips on how to approach making something like it or it may give you some ideas. My router jig made an outside version 9mm wider than the 5mm core. I then used a ball bearing bit to route out around this leaving a raised core. One of the problems is that any imperfections in the larger outer shape are magnified in the inner shape and while the one I use is passable, it is very difficult for me to make and I know it could be better. As it is, it took a couple of days to make, so I am reluctant to spend a lot more time on it without an alternate and easier plan. Better to get the design right and then get them made $$$ if I am to do any kind of mass production of such coils I suspect. Similarly with the circuit. While a very good one off DIY version is possible to do relatively easily and successfully, even if most will need to have a few goes at it, making these kinds of specialized things requires more skill to make the machines and tooling than the coil itself and so not really worth the effort. Anyway...more than enough from me...yes I like to see animations (who doesn't) and some practical work like this (that is very clever, I know I will use it for something eventually)...not at all OT...keep it up! pete Link to comment Share on other sites More sharing options...
col Posted March 6, 2008 Report Share Posted March 6, 2008 Well...I guess I don't know the draw. My impression was that a LM386 with an 8 ohm load and a 9 volt supply had a typical output of 700mW at 9 volts into an 8 ohm load. Given that you and I have both reported that we can run the sustainer with minimum gain (pins 1 and 8 open), then I could only imagine that it may be less, and give the restraints of a battery supply, likely to be on the low side...perhaps 0.5 W. Gain doesn't mean much in this respect, a low input with a gain of 50x can use less power than a higher level input with the gain set at 20x.... Look at the dissipation vs output graphs in the datasheet. class AB is not particularly efficient - around 70% at best... LM386 will be no where near that for much of what we expect it to do, so although the output power may be something like 0.5 or 0.6w, the overall power used including dissipation etc. may be over 1w. It could be considerably more if we are trying to work it too hard - over time this will damage the chip. Might be a good idea to fit a heat sink... or try some other class of amplification... class B or some such... Col Link to comment Share on other sites More sharing options...
psw Posted March 6, 2008 Author Report Share Posted March 6, 2008 Look at the dissipation vs output graphs in the datasheet. Ok...it seems to be saying that between an output of 300-700mW it puts out about 500mW of dissipation. By dissipation, I gather that this is heating of the chip in producing this amplification? So, combined, yes, possibly over 1 watt...however only half of that appear to be sent to the driver as actual driver power. So, maybe talking about two different things here. I was suggesting that we are only putting in perhaps about 0.5W into the driver, not the efficiency of producing that power. Of course better amplification and such (I am concerned about that 10% THD as much as anything) would make better use of our power resources (a battery) and class D an excellent modern choice if it were easily available. However, in regards to the amount of power used to actually drive the strings it would appear to be about 0.5-0.7W as I understand it. My circuit has no preamp gain and works at minimum (x20) amplification with this chip. This is significantly less and cleaner than previous designs I have worked with...the sustain-o-caster for instance had enormous preamp gain. Perhaps this accounts for the better harmonic drive performance and is something I wish to look into once the switching is sorted. So, either I am missing something, or as far as the fizz/EMI issues, we are only putting out half a watt of power, a lot less than UI and Fizz are doing anyway, and with the restrictions on the available power we have by using a battery and the inefficiencies of the amplifier drawing another half watt in wasted energy, I don't see that we could be generating much more. As for that THD, I suspect that this has a major impact on the fizz or distortion that I experience and a reason to explore alternatives or to give enough power to run the LM386 cleaner by giving it plenty of juice. Of course, may very well be misunderstanding or be wrong in my readings of the symptoms and remedies that appear to be presenting themselves with my particular approach. pete Link to comment Share on other sites More sharing options...
utopian isotope Posted March 6, 2008 Report Share Posted March 6, 2008 (edited) Thanks for reply! @psw: Machining nylon is tricky thing to do. Does your router have a speed adjustment? Too high rpm simply melts nylon, as you have experienced. No matter how sharp bit is in use. Higher rpm requires higher feed, so that nylon doesn't have time to melt. Under 5000 rpm is more suitable. But then again, lower the rpm, the more prone the working piece is to kick on your fingers. For example, on the mill I've used under 1000 rpm when machining nylon. Sanding/grinding is more forgiving, since it doesn't have only two or four teeth as routing/milling bits have. So that it allows you to do things more gently, little by little. Anyway, feel free to post to me, I'll be glad to help on those machining things. ---------------- Yes, I'll make some animations. Edited March 6, 2008 by utopian isotope Link to comment Share on other sites More sharing options...
curtisa Posted March 7, 2008 Report Share Posted March 7, 2008 It occured to me that one of the easiest ways of making a good looking bi-lateral design would be to get a pickup bobbin with separate poles and cut it in half with a sharp knife. Reverse the 3 poles on one side and then wrap your driver (perhaps blocking it up to make a thin driver and then sticking the two halves/coils back together with superglue...put a cover on it and you would never know. Of course this would not be a pickup anymore, I am not even sure if you could put a pickup next to it, wouldn't the differing magnetic fields adversely effect it? However, it would be a neat and easy way of doing it...just a thought! So if I understand you correctly you're slicing the bobbin along its length, from bass to treble, and the two halves you end up with are the top plate plus half the bobbin, and the bottom plate plus half the bobbin? The two coils in that instance would then be wound in top of each other. Or maybe I'm misunderstanding here... I tried to get my hands on some plastic extrusions the last couple of days but couldn't get anywhere. I was really after a rectangular tube with a 5mm x 20mm window with the thinnest walls (around 1mm) I could find. The best prefab option I could find was a rectangular tube with a 5mm x 8mm "window" at 1mm wall thickness from a hobby shop made from styrene. At a plastics manufacturing shop they suggested they could build something out of 2mm acrylic sheet and make all the joins out of plastic welds, but 2mm sheet is probably too thick. I was thinking perhaps I could order something like this: http://www.stewmac.com/shop/Electronics,_p...ckup_Parts.html Cutting the bobbin vertically between the 3rd and 4th slug holes would give 2 (rough) bobbins. A bit of reshaping and cutting out a new rectangular window would give me something not unlike the two side-by-side bobbins shown in that sustainiac stealth dissection pictorial. Where do you get your magnets, Pete? Link to comment Share on other sites More sharing options...
Recommended Posts