psw Posted September 4, 2007 Author Report Share Posted September 4, 2007 ...Each change effects the "count" and so a little trial and error and a lot of coils will have to be done to get an accurate result. 7 ish ohms though is perfectly usable, but I may as well get it "right"... Something else you might want to consider as your manufacturing techniques allow the coil to be tighter is trying some slightly thicker winding wire ! Why? I hear you ask... One of the possible (and fairly likely) reasons why wire in the range 0.2 - 0.25 seems to be near 'optimum' is an effect called 'flux linkage' for an inductor to be optimally efficient, the flux generated by each individual winding must envelop all the other windings. If the wire is thicker, then you need more windings to get your 8Ohms and the size of the coil increases dramatically, as soon as it is big enough that some of the windings are not 'flux linked' directly to some of the other more distant windings, efficiency is lost. Sure, there are many other variables in play here, but I would imagine that a coil where the largest guage of wire that still allows maximum flux linkage will provide for the most powerful and most efficient driver. (thicker wire means more turns which in turn means stronger magnet for the same current.) Due to flux linkage (and ss I have suggested for a long time), the best 'format' for the coil is likely to be not flat or tall, but a roughly square cross section for each wire bundle (on either side of the core) - with 0.23 wire this is roughly 3mm thick for 8Ohm when using my limited hand winding skills... The thickness == height for bundle cross section doesn't help you with your piggy-back approach, however it _*may*_ be possible for you to get slightly improved results using slightly thicker wire assuming your improved manufacturing techniques can keep the size down.... Of course, it may be that using the same wire and reducing the size will improve the flux linkage even more and that that is the way to go.... Still a worthy area for experimentation - it would be crazy to start a manufacturing run without eliminating all the reasonable options first { @ everyone, not just Pete } It is possible to calculate a whole bunch of different effects that may effect driver design, unfortunately the odd shape of the core means that you REALLY need a deep understanding (I don't have one) as the ready rolled equations tend to be for the type of coils used in transformers and for radios, so its difficult to know what difference the elongated core will make.... Some other potentially important effects/variables are: eddy currents core saturation permeability of core material hysteresis loss read about all this an more at this excellent website: Producing wound components I also wonder what effect the 'Q-factor' may be having... Anyone who's read up on pickups will know that more powerful pickups with a higher impedance also sound 'darker' this is because the coils have a lower resonant frequency.... (looky here) I wonder to what extent the resonant frequency of the driver coil(s) has an effect on efficiency and performance ? For example some folks have more trouble getting the higher strings to sustain... Some find the low E sustains very well while for others it is not so strong... some drivers tend towards blooming harmonics no matter what the circuit is set for.... maybe we need to look closer at the resonance of the driver and controling its 'Q' factor ? Anyway, I hope someone out there has the physics and maths chops to apply some of the info from that wound components website to our project cheers Col Well...very early in the thread I did read a lot about this kind of thing, unfortunately I don't have the "chops" to go beyond the concepts and basic principles. I do think that the Q-factor plays a major influence, especially on the ability to get the high strings to move. I have "the feeling" that this is the major reason for the success of the thin coil design and the choice of specs of the wire. It is also why, once the design is significantly changed (2xparallel coils) I did not believe it would work as well without a re-evaluation of these choices. An example from the earliest days of the thread that went unreported... I had a guitar with a neck pickup only, tune-o-matic bridge and tailpiece (kind of a jazz guitar thing). I made a small coil around a steel rivet and threaded the string through it immediately behind the bridge. I then placed small but powerful magnets to either side of the string (N-S) just in front of the bridge. The idea was that the coil would send a magnetic pulse down the string and that the string and this pulse would be attracted to one or other of the magnets, pulling the string with it from side to side and that the small movement at the bridge end would propagate down the string. In retrospect, a very inefficient system...hmmm. However, I was using a computer "tone generator" at the time and feeding these experiments from the headphone socket of the computer speakers (I had been doing some experiments to try and confirm if different wave shapes (square, sine, triangles...would be more adventageous). Now, the thing is, if you tuned the string exactly to the pitch of the signal, it did kind of work. (such experiments went on to fuel the Hex designs) This is the same principle of breaking a wine glass by hitting it's resonant frequency...there is a lot of power there. However, it will only work at that frequency. The purpose of the phase compensation circuitry of the patented systems, best descried by Rose) seems to be an attempt to address this to an extent. Q-factor may also influence the "fizz" EMI effects too. Given my limited knowledge, skills and abilities and consequent naive approach, LK encouraged me to ignore a lot of these things and concentrate on the driver to alleviate some of the problems. Instead of a pseudo science approach, I took a more hands on practical approach trying to extend working models till a design emerged that "worked". The other factors too are tricky to address, even with the knowledge. By necessity, a DIY project such as this will have enormous variables. And, there are significant limiting factors as well. I know ferrite cores would be better (I used them extensively in the hex designs), but they are impossible to deal with. A lot of the "secrecy" in the hex designs was to do with the exotic nature of materials used, the expense and frustration involved and a feeling I was paddling against the tide and I didn't want others to follow me too far up sh*t creek if it turned out that the ideas were not successful and then blame the navigator (which I think transpired to a degree and so I took what I could from all of that and changed direction and came up with the thin driver as a result). The other thing to consider is "application". In my latest venture, I can not attempt to make the "ideal" driver, but I am trying to make as good a driver as is possible - for this application! For this application there are a lot of restrictions and unknowns. The piggy-back pickup/driver must work with the pickup of choice and not interfere with it's operation. It will need to be a single coil, at least in this incarnation. It will need to produce results without excessive EMI getting to the source/bridge pickup. Permeability factors, core saturation, and such are concerns that are limited by what I can do within the brief of my application of the device. I did attempt, for instance to add a band of epoxy/iron around the outer and inner edges of my new coils. Shielding in this kind of way would be expected to increase eddy currents so close to the coil, but it was worth a shot, this band would also provide some protection for the outer edge of the coil. (another idea was to fill the spaces between the poles with similar material. But, these things would also affect the pickup's response... My "feeling" on the thin coil design is that it works as well as it does in part because it condenses all of the coil as close to the strings as possible and exists on the very edge of the magnetic poles of the core. Its small profile contributing to its small EMI radiation. Also, that the design appears to be "fast enough" to drive even the highest notes without phase compensation circuitry. How or why this is so, or if this "feeling" is correct is for others to work out, but is born out of a lot of experimentation and comparison with other differing ideas. The important thing is that the device works within acceptable parameters for this application. These "feelings" have been informed by my research into the factors you have mentioned but I have achieved more "results" from experimentation than with the theory, no matter the entertainment value ! Right now, I am struggling with the manufacturing side of things, later it will be concerns like the sunken poles on the slightly elevated cover caused by the driver I'm sure. My last two coils failed despite two days rebuilding the form work to create them. Both were due to a failure of the glue and, as it turns out, there was some compression of the core former that would have made the core about 1mm too small to fit around the poles of the pickup! The first of these too, which I reported on yesterday attempted to substitute one pot polyurethane but it takes forever to dry (especially since the solvents cant escape inside the form work press) and appears to be not adequate strong enough to hold the coils internal tensions in place. The second used epoxy, but I suspect the mixed ratios were off or there may have been some contaminant (despite washing) from the previous coil attempt, and so even overnight the epoxy never cured. You can see the frustrations I am dealing with here have nothing to do with theory of design anymore, but on the practicality of construction. In recent years, and with this "venture" in particular, the application requirements and limitations, and the practical construction concerns have put theory well into the background. I think that a lot of the design of the commercial units are informed by the practicality of coils produced in a Chinese factory somewhere once they had settled on a theory and committed themselves to a particular direction. My intention is, that if I can create something that works that can start people off with this thing, then improvements and informed theory may well emerge. Got to start somewhere, and this is the best I can come up with at this time that addresses all the criteria. I freely acknowledge my limitations and strive to overcome them or "work" through them regardless and learn from the countless mistakes and failures and be happy for the successes that would not have occurred without them. It is particularly frustrating in that I have a working prototype of the thing, but the details and the little improvements and ways to make the thing repeatedly, is where the devils reside. Anyway...more frustration ahead with making jigs for this thing. Fortunately, winding the coils I find quite enjoyable. Making the circuits is a little tedious, especially without a PCB, but it is a lot less frustrating than prototyping too. I still have a very low tolerance for the frustrations and obstacles, so not a good frame of mind to be attempting such things. Still, am taking lots of breaks and it distracts me a little from other things, so it serves a purpose. Anyway...thanks for the link col and the suggestions. Any manufacturing tips would be welcome, potting types, mold/jig making, accurate shaping of materials, the like... pete Link to comment Share on other sites More sharing options...
col Posted September 4, 2007 Report Share Posted September 4, 2007 (edited) Well...very early in the thread I did read a lot about this kind of thing, unfortunately I don't have the "chops" to go beyond the concepts and basic principles. I do think that the Q-factor plays a major influence, especially on the ability to get the high strings to move. I have "the feeling" that this is the major reason for the success of the thin coil design and the choice of specs of the wire. It is also why, once the design is significantly changed (2xparallel coils) I did not believe it would work as well without a re-evaluation of these choices. ... This is the same principle of breaking a wine glass by hitting it's resonant frequency...there is a lot of power there. However, it will only work at that frequency. The purpose of the phase compensation circuitry of the patented systems, best descried by Rose) seems to be an attempt to address this to an extent. Q-factor may also influence the "fizz" EMI effects too. Hmm, would phase compensation be there to correct or adjust the resonance of the system ?.... I agree about possible links between resonance and fizz... For these small coils, the self resonant frequency will be very high - way above the guitars range for sure... however, combined with the output section of the circuit, it _may_ be enough to cause the amplification of stray noise in the system due to feedback, if this happens then it will cause loss of efficiency, and may add to the mix of fizz, fuzz and crackle that S*****niac call 'grunge' Given my limited knowledge, skills and abilities and consequent naive approach, LK encouraged me to ignore a lot of these things and concentrate on the driver to alleviate some of the problems. Instead of a pseudo science approach, I took a more hands on practical approach trying to extend working models till a design emerged that "worked". ... Yes, that was good advice, however, that was then and this is now ! since those wise words, this project has developed to the stage where most of the thing is done - unfortunately it's been at this 'refinements' stage for a while now. Thats why I've gone back to 'researching' - I feel its at the stage where I need a better understanding of why things that work do and things that don't don't. Getting even a basic grasp on whats really happening will help to avoid wild goose chases, and at the same time direct us towards possibly useful modifications. Permeability factors, core saturation, and such are concerns that are limited by what I can do within the brief of my application of the device. I did attempt, for instance to add a band of epoxy/iron around the outer and inner edges of my new coils. Shielding in this kind of way would be expected to increase eddy currents so close to the coil, but it was worth a shot, this band would also provide some protection for the outer edge of the coil. (another idea was to fill the spaces between the poles with similar material. But, these things would also affect the pickup's response... My "feeling" on the thin coil design is that it works as well as it does in part because it condenses all of the coil as close to the strings as possible and exists on the very edge of the magnetic poles of the core. Its small profile contributing to its small EMI radiation. Just as much EMI can come from a small driver as a large one.... The compact coil can help the system produce less stray EMI if it is more efficient at turning current into magnetic force, but other than that I don't see that it makes a lot of difference in practical terms. Also, that the design appears to be "fast enough" to drive even the highest notes without phase compensation circuitry. How or why this is so, or if this "feeling" is correct is for others to work out, but is born out of a lot of experimentation and comparison with other differing ideas. The important thing is that the device works within acceptable parameters for this application. These "feelings" have been informed by my research into the factors you have mentioned but I have achieved more "results" from experimentation than with the theory, no matter the entertainment value ! I've never understood what you mean by the driver being 'fast' ? My understanding is: How quickly the driver can change polarity is more to do with the core material and configuration than the dimensions of the windings. The phase is nothing to do with 'speed' although it is important. (It may be that the commercial units sacrifice phase in their driver compromise because they can be corrected for in the circuitry while efficiency losses cannot be 'corrected' for...) Right now, I am struggling with the manufacturing side of things, later it will be concerns like the sunken poles on the slightly elevated cover caused by the driver I'm sure. My last two coils failed despite two days rebuilding the form work to create them. Both were due to a failure of the glue and, as it turns out, there was some compression of the core former that would have made the core about 1mm too small to fit around the poles of the pickup! The first of these too, which I reported on yesterday attempted to substitute one pot polyurethane but it takes forever to dry (especially since the solvents cant escape inside the form work press) and appears to be not adequate strong enough to hold the coils internal tensions in place. The second used epoxy, but I suspect the mixed ratios were off or there may have been some contaminant (despite washing) from the previous coil attempt, and so even overnight the epoxy never cured. Sounds like a big headache You can see the frustrations I am dealing with here have nothing to do with theory of design anymore, but on the practicality of construction. In recent years, and with this "venture" in particular, the application requirements and limitations, and the practical construction concerns have put theory well into the background. I think that a lot of the design of the commercial units are informed by the practicality of coils produced in a Chinese factory somewhere once they had settled on a theory and committed themselves to a particular direction. That is very likely - the manufacturing process is always very high on the priority list for any commercial product. Anyway...more frustration ahead with making jigs for this thing. Fortunately, winding the coils I find quite enjoyable. Making the circuits is a little tedious, especially without a PCB, but it is a lot less frustrating than prototyping too. I still have a very low tolerance for the frustrations and obstacles, so not a good frame of mind to be attempting such things. Still, am taking lots of breaks and it distracts me a little from other things, so it serves a purpose. Anyway...thanks for the link col and the suggestions. Any manufacturing tips would be welcome, potting types, mold/jig making, accurate shaping of materials, the like... pete I am a little surprised that you are not going to etch your own pcbs. That is not difficult, and makes a lot of sense if you intend to do a 'run'. It would also allow you to make the circuit more compact. I'm sure there are plenty of folks over on diystompboxes who etch their own pcbs, so they could probably advise and suggest ?... ---------------------------------- I've been thinking about other possible areas of study and wondering if anyone out there can help with some bona fide knowledge™ My current bugbear is the impedance of the driver. We know that the LM386 is supposed to drive a load of 8 ohms.... The thing is that we don't know what load the drivers we are making really are ! We are measuring the DC resistance of them ! We should be calculating the AC reactance (the signal is an alternating current after all) and combining that with the resistance to get the true value. Unfortunately the one we don't measure is the important one and, what's worse, it changes with frequency ! So not only do we need to calculate and/or measure the reactance of the driver, we also need to decide what frequency range to optimise for.... It may be that we're already pretty close, but we might be out be an order of magnitude in which case we can win a hefty efficiency boost by getting this right. It's even possible that the 'thin' driver with compact coil of relatively few turns of wire 'won' through in the R&D because its reactance is not as far out as that of the larger coils ! 'What if' when reactance is taken into consideration, larger coils with more turns can provide a more efficient more powerful driver that works better over a broader range of frequencies ?.... it would start looking like the commercial units... hmm... Now, please don't blow a gaskit ... this is only conjecture, and even if it turns out to be correct, we can still use the knowledge to help refine and improve our existing systems... So, who knows how to calculate and/or measure the inductive reactance of these drivers ? (I've gone through a bunch of theory and come up empty handed because of all the unknown quantities (I did the sums a few times over, but the results were obviously wrong(I think/hope!?)) - permiability of my core material.... how to measure/guess the "effective path length" and "effective area" of the core taking into consideration the extruded form and the fact that the core 'circuit' is not closed... just so many unknowns... I hope I don't have to build a test inductor from my core material and a special circuit just to test the permiability....) enough Col Edited September 4, 2007 by col Link to comment Share on other sites More sharing options...
psw Posted September 5, 2007 Author Report Share Posted September 5, 2007 Thanks Col... Hmm, would phase compensation be there to correct or adjust the resonance of the system ?.... I am sure of it! For these small coils, the self resonant frequency will be very high - way above the guitars range for sure... Ahh...but the resonant "Q" frequency of the pickup, the driver and anything in the circuit could conspire to amplify high frequency fuzz, fizz of anything else that gets into the system... A "Q" out of the guitars range is desirable I think, because it will not obviously bias a particular frequency (although some evidence of this is apparent)... Just as much EMI can come from a small driver as a large one.... The compact coil can help the system produce less stray EMI if it is more efficient at turning current into magnetic force, but other than that I don't see that it makes a lot of difference in practical terms. It kind of developed from a lot of the Hex things. The most obvious thing about them, no matter how they worked, was their very small size. I built the thin coil to emulate this with a conventional coil construction for DIY. My hypothesis is that the smaller the driver the more "contained" (confined to the area around the driver itself) it is. The shape also helps too. I've never understood what you mean by the driver being 'fast' ? My understanding is: How quickly the driver can change polarity is more to do with the core material and configuration than the dimensions of the windings. The phase is nothing to do with 'speed' although it is important. (It may be that the commercial units sacrifice phase in their driver compromise because they can be corrected for in the circuitry while efficiency losses cannot be 'corrected' for...) You have the idea. The current, being AC is flowing backwards and forwards. The core material, it's permeability is a large factor...the ability of the core to change polarities. With higher frequencies (also hampered by the lighter strings and increased tension) these polarities need to change very fast. The core material is important. We all know if we attach a magnet to a steel nail, it will itself become a magnet for some period of time after the magnet itself is removed. A core of similar material will also try to hold on to a magnetic charge and resist change. This is why ferrite would be a far better material and is used in inductors typically (or laminated steel, Mu metal, etc). Ferrite is highly magnetic, but will not retain a magnetic charge like the nail. But there are factors in the coil, at least from my reading (and it was some time ago). If you look at the current flowing through the coil as water through a hose pipe. Now imagine that you can pump water through it from either end. (In passing: a thin hose will be more difficult to pass a large volume of water through it and provide more resistance, similarly a thinner wire needs less turns or length for the same resistance). It is easy to see that the water flowing in one direction carries some momentum and when it changes direction, it is required to counter the opposing flow...there is a lag. The quicker you attempt to change the direction of the flow, the more crucial this lag is. This lag is tiny, but combined with lags in the core material and in the strings response, they are significant. The idea of the thin driver is to minimize the amount of core material and to place the maximum effect very close to the string and the edge of the permanent magnet. The result is a much smaller than conventional core which responds quicker (i.e. less lag) than a more conventional device. There are other factors in the thin coil, the number of overlapping wires is important, with a coil of so few turns, a thin coil maximizes this also. So...the aim of my present project is to take this to it's extremes, and I was a little surprised at the results. I was surprised for instance that your driver Col had 2mm coils, as did some of Avalon's. The obvious next step was to see how small I could make them and still be effective...hence the building of "the machine" and its tricky, jiggy components to make such coils. Oddly enough, this immediately produced favorable results (lending support to my "thin coil theories" explained in part above. It also opened up, combined with the ideas Tim had been exploring for making coils, the possibility of making coils independent of a pickup, to better recreate my original "piggy-back" driver/pickup idea. Clearly, these coils are not easy to make without the machine and a lot of practice. So, why bother going to all that trouble just to prove a theory or make one for myself? Well, the machine has other uses and I was expecting to use it for a few ideas including stacked coils and different coil depths, etc (anticipating that the first coils wouldn't work!). Making such coils would solve a lot of problems for a lot of people (wire supply, consistency of manufacture, skills, apprehension about winding coils, etc) but there would need to be some standardization at the outset (only works on single coil pickups) till other schemes were developed. Only if a need was there, could I justify working on such things (I don't need an HB driver for instance, but I am sure a lot of people would and I'd be interested in making something). Sounds like a big headache So three whole days later, I am rebuilding yet another jig. Each one gets a little better and simpler, which is a sign of better, right? This latest one has aluminium section to avoid flexing of the plastic parts. Tapped components avoids the fiddly bolts and nuts of previous designs in getting the thing apart. Wider flanges allow for an improved side pressing to make the things even more condensed. Steel inserts in the core part, particularly at their ends, avoids compression from the coils tension and malformation. A lot of work goes into making these things that is completely unrelated to making the device itself, as you can imagine from this description! However, once an appropriate jig is made, this will be able to be reproduced to make even more coils. Making the coils is far easier than making the jigs to make the coils and it will be a relief when this is taken care of. That is very likely - the manufacturing process is always very high on the priority list for any commercial product. Quite right, and my present venture is no different, compromises need to be made. Thats why I've gone back to 'researching' - I feel its at the stage where I need a better understanding of why things that work do and things that don't don't. Getting even a basic grasp on whats really happening will help to avoid wild goose chases, and at the same time direct us towards possibly useful modifications. Good Col, I think there is still a lot to learn. I think I can contribute more by setting a standard and making it available for people to work from. My hope is that all this work can be condensed down to a few basic components and from this point some real developments can be derived. Having a coil and or a circuit to work from will really give people a viable option and encouragement to take this further. Once something has been successfully installed (as my present test guitar is) it is quite simple to swap out a circuit, or to connect up an alternative driver idea...and to make direct comparisons that others can also test for themselves. Using different amps and drivers, and the enormous variances we have here in the effectiveness of the construction, or even getting things to work as they should, is the biggest obstacle. Many more knowledgeable people have expressed interest, but balk at the whole coil winding side of things. Many people here have innovative ideas that have not been fully explored as yet, and a scheme such as mine makes these things a much easier proposition and of more benefit to others if some of the inconsistencies are removed. I am a little surprised that you are not going to etch your own pcbs. That is not difficult, and makes a lot of sense if you intend to do a 'run'. It would also allow you to make the circuit more compact. I'm sure there are plenty of folks over on diystompboxes who etch their own pcbs, so they could probably advise and suggest ?... Well...it is like this. Getting into etching my own boards is a commitment and even more study and experimentation. Making boards doesn't sound easy, all those fine holes to drill and boards to cut...let alone the toxic chemicals and the expense of materials and equipment. The layout of my circuit could not be made any smaller. At 37mmx22mm(x10mm) it is small enough. I was not able to get it smaller on PCB than with vero, the components are very tightly packed with little vacant space. I have exactly mirrored the PCB layout on vero but it does mean a lot of track cuts. However, it is not fully tested and may need modification anyway. Vero takes care of the drilling and is easy to snap to size, then grind off the rough edges. The track cutting jig is quite clever. It is a piece of vero board with track cuts going right through the board (many of these track cuts are between the holes BTW). I have tiny "pins" soldered on each corner of the jig, and these press fit into the vero. I then simply take a fine dremmel like engraver and trace through the slots to cut the copper strips below. accurate and a no brainer to do, certainly easier than etching and drilling all those tiny holes. The vero is cheaper too than an equivalent blank board, so go figure! If this is successful and I need to make a lot, then I have already made the layout and got pricing to go ahead with a PCB. I would need to make about 200 of them to be cost effective (works out to be about a dollar each), so you'd want to know that everything was right. This method will do for now, and the backup plan is in place. Obviously, there is so much expense in the setting up both in time and money, that there will be no profit in these things, but I can see how it could work with even small quantities. The coil is the bigger headache right now, making a circuit look acceptable is not that hard (if you have the design) but scratch building these coils and making them less DIYesque is proving to be quite tricky. The next worry is when people try to fit them and variance in pickups cause them not to fit or need to amke alterations to covers and such to get the things in there and the wires safely from under them. The other thing that has been bothering me, is that the thickness of the coil means that the cover will be elevated about 1mm. This is not a problem with mounting, but aesthetically, it means that flush poles (typical and even with my staggered poles, the b string is flush) will be sunken below the depth of the cover by that amount. One could make it a feature by secreting some SMD LED's between the poles so they give off a glow . You could stick little metal washers to the poles to extend them a little. If I were making pickups, I could build them with 1mm more length to the magnet. Ideally, I would make covers (themselves 1mm thick) but then every pickup's pole spacing is different and such things can get very expensive to develop and have made. Better stop wasting time and get back too it, hopefully this latest jig will do the trick and I can reproduce more (they act as a press as well, so the coil needs to fully cure before attempting to be removed...therefore, a few of these things will be necessary if production is to be more than one a day! pete Link to comment Share on other sites More sharing options...
psw Posted September 5, 2007 Author Report Share Posted September 5, 2007 Oh...and Col And about testing coils. Like you, I felt this need for a while, especially early on. I was trying or intending to try some stuff and there are things that are software (freeware) based that could do the job. One thing is a spectrum analyzer and tone generator programs. Run tones through the coil and use the spectrum analyzer to determine peak q's etc. You can also get software oscilloscope programs if you really want to get into it. Another trick is to use a pickup to transfer the signal emitted from the coil back into a signal...may be useful to test the interactions of the driver and pickup. Could be someone clever could generate a reverse phase of the "fizz" and cancell it out electronically, if we knew exactly what it is! As for my latest coil and jig...well closer. Running out of wire though so have ordered a bulk reel of the stuff in anticipation. The coil is pretty solid and although the core is a little "oversized" this time, it still fitted into the cover. Fitting them into covers and selling these would be a lot easier and more professional looking...any thoughts on this. I have a few spare covers that I bought a while back but none of the pole holes fit any of my pickups! The other problem is that I used side clamps and some of the wires sneaked past and under these...while they were able to be pushed back in, the glue had already set by this stage...tricky. Also, dispite release agent, these aluminium side clamps did threaten to glue themselves to the side of the coil. Eliminating the side bulge is vital to make these fit and to enhance efficiency and the overall look of the thing, if nothing else... More work, and wire required.... All epoxies are not created equal either, I am trying some more expensive stuff that seems to work a lot better...all very tricky to use and I have to work extra fast...hard to do while you are learning along the way... pete Link to comment Share on other sites More sharing options...
col Posted September 5, 2007 Report Share Posted September 5, 2007 Thanks Col... Hmm, would phase compensation be there to correct or adjust the resonance of the system ?.... I am sure of it! I think it is misnamed then ? resonance is about change in amplitude at different frequencies. True 'phase compensation' would have no impact on the amplitude, just on the phase of the signal. I've never understood what you mean by the driver being 'fast' ? My understanding is: How quickly the driver can change polarity is more to do with the core material and configuration than the dimensions of the windings. The phase is nothing to do with 'speed' although it is important. (It may be that the commercial units sacrifice phase in their driver compromise because they can be corrected for in the circuitry while efficiency losses cannot be 'corrected' for...) You have the idea. The current, being AC is flowing backwards and forwards. The core material, it's permeability is a large factor...the ability of the core to change polarities. With higher frequencies (also hampered by the lighter strings and increased tension) these polarities need to change very fast. The core material is important. We all know if we attach a magnet to a steel nail, it will itself become a magnet for some period of time after the magnet itself is removed. A core of similar material will also try to hold on to a magnetic charge and resist change. This is why ferrite would be a far better material and is used in inductors typically (or laminated steel, Mu metal, etc). Ferrite is highly magnetic, but will not retain a magnetic charge like the nail. But there are factors in the coil, at least from my reading (and it was some time ago). If you look at the current flowing through the coil as water through a hose pipe. Now imagine that you can pump water through it from either end. (In passing: a thin hose will be more difficult to pass a large volume of water through it and provide more resistance, similarly a thinner wire needs less turns or length for the same resistance). It is easy to see that the water flowing in one direction carries some momentum and when it changes direction, it is required to counter the opposing flow...there is a lag. The quicker you attempt to change the direction of the flow, the more crucial this lag is. This lag is tiny, but combined with lags in the core material and in the strings response, they are significant. The 'remnance' of the core material is the important thing as far as the polarity switching speed - it is not related to the size of the core, only to the material. I think the analogy of water trough a hose is only valid in explaining the basics of resistance (narrower hose = reduced 'flow) - beyond that the analogy breaks down and causes confusion. I have done a lot of reading on this and as yet have not come across anything that suggests what you are describing, it also seems wrong intuitively to me.... can you give me a link to some reference ? There are other factors in the thin coil, the number of overlapping wires is important, with a coil of so few turns, a thin coil maximizes this also. Its the flux linkage that is important here, the optimum rectangular shape for the cross section of a winding 'bundle' would be a square - see the image i posted in my last but one post.So...the aim of my present project is to take this to it's extremes, and I was a little surprised at the results. I was surprised for instance that your driver Col had 2mm coils, as did some of Avalon's. The obvious next step was to see how small I could make them and still be effective...hence the building of "the machine" and its tricky, jiggy components to make such coils. My most successful driver - the dual core one - has wire bundles that are roughly square in cross section. They are roughly 3.2mm thick and about 3.2 mm wide down the length of the driver. There is some variability due to winding quality.... If I had a setup to compress the bundles while the glue set, then they would probably go to about 2.5x2.5mm or smaller, but remember, these are 4ohm coils ! The coil is the bigger headache right now, making a circuit look acceptable is not that hard (if you have the design) but scratch building these coils and making them less DIYesque is proving to be quite tricky. The next worry is when people try to fit them and variance in pickups cause them not to fit or need to amke alterations to covers and such to get the things in there and the wires safely from under them. Yes, designing for the variability in pickups that end users might have is certainly going to be a major issues - even different fender strats can have differences large enough to cause problems - let alone different types of guitar from different manufacturers. ============================ Oh...and Col And about testing coils. Like you, I felt this need for a while, especially early on. I was trying or intending to try some stuff and there are things that are software (freeware) based that could do the job. One thing is a spectrum analyzer and tone generator programs. Run tones through the coil and use the spectrum analyzer to determine peak q's etc. You can also get software oscilloscope programs if you really want to get into it. Thanks, I am aware of some of the PC software scopes... they are ok for analysing audio frequencies. Unfortunately a lot of what you need a scope for when building and debugging audio circuits is for dealing with ultrasonic frequencies like those occurring in parasitic oscillations... also stuff like we were discussing related to high frequency noise being amplified due to a high resonance frequency.... PC soundcards cant 'hear' those frequencies - the sample rates are just way too low. You can buy 'PC' oscilloscopes with a PCI card that you plug in and then use the PC as GUI, but they are mighty expensive.... There are some really nice looking portable digital scopes available at reasonable prices now - a few hundred £ rather than thousands.... but I can't justify that As far as testing the coils, unfortunately, the 'test' I want to do is not something thats easily testable using an oscilloscope or spectrum analyzer... basically it is the AC equivalent of resistance and unfortunately you can't just connect up an ohmeter I need to do some more googling on the matter - possibly there are simple circuits out there that I could build to use for making this measurement... As for my latest coil and jig...well closer. Running out of wire though so have ordered a bulk reel of the stuff in anticipation. The coil is pretty solid and although the core is a little "oversized" this time, it still fitted into the cover. Fitting them into covers and selling these would be a lot easier and more professional looking...any thoughts on this. I have a few spare covers that I bought a while back but none of the pole holes fit any of my pickups! The other problem is that I used side clamps and some of the wires sneaked past and under these...while they were able to be pushed back in, the glue had already set by this stage...tricky. Also, despite release agent, these aluminium side clamps did threaten to glue themselves to the side of the coil. Eliminating the side bulge is vital to make these fit and to enhance efficiency and the overall look of the thing, if nothing else... More work, and wire required.... All epoxies are not created equal either, I am trying some more expensive stuff that seems to work a lot better...all very tricky to use and I have to work extra fast...hard to do while you are learning along the way... pete Sounds like it's coming together... The potting compound is tricky though - for hand made one-offs, the stuff I use is ideal, but for making a run of drivers, the expense would be prohibitive, and the curing time would certainly require multiple jigs to be used. I wonder if there are potting products that are designed to be 'cooked' in some way to allow fast controlled curing ? UV cured polyurethane would be nice, but i guess the UV wouldn't penetrate deep enough into the coil and so not all of it would cure. OK, I'm off to find out more about measuring / calculating inductive reactance.... cheers Col Link to comment Share on other sites More sharing options...
mrjstudios Posted September 5, 2007 Report Share Posted September 5, 2007 The thing is that we don't know what load the drivers we are making really are ! We are measuring the DC resistance of them ! We should be calculating the AC reactance (the signal is an alternating current after all) and combining that with the resistance to get the true value Ah Col, you are right! I'll see what I can come up with.... One could make it a feature by secreting some SMD LED's between the poles so they give off a glow. Whoa there!!! I thought we were getting away from the glowing S****** thing from about 30 pages back! You could stick little metal washers to the poles to extend them a little. Uhm, not so sure about that. Theoretically, it should work, but when I was spacing my last driver's magnet, even with really good computer steel (very magnetic), the power of the magnet was cut literally by around 60%-75% by the time it got through the spacer steel. I don't know if a washer would do the same to a pickup pole, but it is something to think about. -MRJ Link to comment Share on other sites More sharing options...
mrjstudios Posted September 5, 2007 Report Share Posted September 5, 2007 I think this might help you Col AC resistance Equations Link to comment Share on other sites More sharing options...
psw Posted September 6, 2007 Author Report Share Posted September 6, 2007 Whoa there!!! I thought we were getting away from the glowing S****** thing from about 30 pages back! biggrin.gif wink.gif I was only kidding, but I know people are going to want to, so have include LED power out on the circuit! I readily admit that my "theories" are shaky, I don't have a reference for that one, those electrons are flying through the wire at the speed of light or something, I am not even sure of the ideal flux linkage and such... The work I am doing is to balance the design to it's purpose, and to that extent, a less than optimum driver may result. The important thing is if it is good enough for this application. A lot of my personal "hypotheses" are simply based on many, many experimentations. In the last 3 days I have attempted 8 coils and made 3 new jigs to create them. Each one gets a little better with the jigs (but still not quite there yet...makes a good coil but the aesthetics need work), but with the glue alternatives were giving me problems. Polyurethane made the winding easier but would not set...more costly epoxy works well but finding a best solution to getting a neat and strong exit of the driver wires is proving a little elusive. Now, I have a box here of perhaps 100 drivers of various types. I have tried deeper, cylindrical, sideways, crossways, shielded, multi-cored, multi-coiled, laminated, ferrite, ceramic, alnico, neodymium cores, thinner and thicker wire...stuck them to the top, bottom inside the windings and even sides of a pickup!!! However, the mathematics and theory, no matter how I tried eludes me, as do a lot of the electronics (which I taught myself for this project BTW) so anything I say is based on a "feeling" from extensive reading (mainly textbooks from libraries) and from countless hours, even years of experimentation. I had hoped that my open work would attract someone with the knowledge to guide me in some of these areas, but the application and "design" aspects are equally important. (the closest thing to an expert in this would be R.G. Keen, who I have been led to believe is an expert in transformers in "real life"). I am afraid I have to resign myself to my limitations in some areas and try to substitute these with inspirations coupled with a lot of perspiration! Uhm, not so sure about that. Theoretically, it should work, but when I was spacing my last driver's magnet, even with really good computer steel (very magnetic), the power of the magnet was cut literally by around 60%-75% by the time it got through the spacer steel. I don't know if a washer would do the same to a pickup pole, but it is something to think about. I am not sure why that should happen... Yes, well I think this will become more of an issue once I am able to produce these coils in number and people try to fit them. For people like yourselves, cutting the top out of a cover, or making your own, or a host of other "solutions" could be found, but a lot of people have for some time wished for the device to be as inconspicuous as possible and something has to hold the driver down securely. Ideally. I would fit them inside commercial covers, but that is not really practical. Neither is making them any thinner I suspect. Lets have a Poll... Anyone following this tread with single coil pickups of any variety, could you measure... 1~ the width of the poles (I get about 5mm) 2~ the span of the poles (I have been working on 55mm, but have added a mil on each end to be sure) 3~ the bobbin width (if you can take a cover off...I am working on 15mm max) 4~ the amount of space you estimate there is between the cover and the bobbin for the driver wires to get by (I get less than 1mm) 5~ an opinion on the effect of raising the cover 1mm or so, effectively lowering the pole pieces. While this is not "perfect"...close to this could be achieved by building the drivers into adapted covers, it is the best one size fits all solution that I have been able to come up with. better run now, back later perhaps... pete Link to comment Share on other sites More sharing options...
col Posted September 6, 2007 Report Share Posted September 6, 2007 I think this might help you Col AC resistance Equations If only it were so easy The problem is that the impedance is related to the inductance, and the inductance is dependent on the properties of the core material e.g. permiability and its dimensions (which in our case are so far removed from 'standard' inductors that I don't know if the basic theories can be applied)... I was looking at meters today, maplin sell a DMM that also measures inductance - unfortunately its £45 - I don't have the money for anything like that right now.... If we could measure the inductance, the we can use that to derive the impedance. All the 'easy' ways to measure inductance require kit that I don't have - oscillscope, function generator etc. It may be possible to use a PC soundcard scope for this, but I'm not sure how accurate the results would be because of the extra circuitry that has to be built and linked in. Unknowns like input/output impedance of the soundcard and test circuit... not to mention all the parts for the circuit... I wouldn't be confident that the results were sound. So I guess I'll have to sit on it until I can buy or borrow an inductance meter. What I will do is try using the theory I have read in the last few days to model a hypothetical inductor that can at least give us some idea of how important this area is to us.... Really though we need a meter , that way we can also measure things like what difference the permanent magnet makes (if any) to inductance and impedance... measuring is so much easier - tweaking and measuring really helps to build up a more solid intuition. Heh, I 'know' if the answer from a current or voltage calculation is way out, but I have no idea what order of magnitude of inductance readings to expect from our driver - mili-henrys ? micro henrys ? nano ? no idea cheers Col Link to comment Share on other sites More sharing options...
barboom Posted September 6, 2007 Report Share Posted September 6, 2007 so it would seem you've got somewhat satisfactory results for the plain neck pickup slot sustainer but does anyone have schematics/diagrams of it? Link to comment Share on other sites More sharing options...
psw Posted September 6, 2007 Author Report Share Posted September 6, 2007 Testing ideas is a real problem, and if I were able to do them, it would help to "see" the effects of different designs rather than having to build them all the time and go on "gut feeling." I wonder if someone with access to university equipment could help you out on this quest. Meanwhile...have rebuilt jig 4 and have orded bulk wire as I am rapidly running out and my small reel suppliers have run out. Have built or attempted to build quite a few drivers in the last few days, altering jigs and procedures with each one. Am getting good solid and tight coils within the constraints but the finish leaves a little to be desired. Jig 4.1 should help. Lets have a Poll... Anyone following this tread with single coil pickups of any variety, could you measure... 1~ the width of the poles (I get about 5mm) 2~ the span of the poles (I have been working on 55mm, but have added a mil on each end to be sure) 3~ the bobbin width (if you can take a cover off...I am working on 15mm max) 4~ the amount of space you estimate there is between the cover and the bobbin for the driver wires to get by (I get less than 1mm) 5~ an opinion on the effect of raising the cover 1mm or so, effectively lowering the pole pieces. Any takers??? so it would seem you've got somewhat satisfactory results for the plain neck pickup slot sustainer but does anyone have schematics/diagrams of it? See the links at the bottom of this post for more details and pics of building these things. The project I am working on may also be of interest.... pete Link to comment Share on other sites More sharing options...
col Posted September 6, 2007 Report Share Posted September 6, 2007 (edited) Done some more 'research' I tried modeling an inductor based roughly on a single core driver. The main difference is that the equations are for a solid core while a real driver has a VERY large air gap.... For an Iron core (again a real driver won't have a pure iron core, but this isn't such a significant variation as the air gap business) of 2mm x 2mm x 55mm with 150 turns of wire, the impedance (inductive reactance) is roughly: 3.5kOhm @ 600Hz and 480ohm at 82Hz If the values in a driver are anything remotely like this, it would make the system very inefficient ! Now, adding an air gap into the mix reduces the permiability of the core which reduces the inductance and the impedance, so the real value for our drivers will be lower than this - how much lower I'm not sure though - probably an order of magnitude or more. The equations expect the magnetic circuit that the flux follows to be all core material - no gaps... Here is a diagram to make the point a little clearer (i hope) To be sure about this, we really do need real world measurements for inductance, but at least this illustrates the concept a little better (It's quite possible that the magnet also has an impact on these values) Does anyone own or have access to a multimeter that reads inductance or an LCR meter? if so please take some measurements of your driver (even if it doesn't work well ) cheers Col Edited September 6, 2007 by col Link to comment Share on other sites More sharing options...
psw Posted September 6, 2007 Author Report Share Posted September 6, 2007 See now Col, this is where my brain begins to hurt....and maybe I start making things up...hmmm For an Iron core (again a real driver won't have a pure iron core, but this isn't such a significant variation as the air gap business) of 2mm x 2mm x 55mm with 150 turns of wire, the impedance (inductive reactance) is roughly: 3.5kOhm @ 600Hz and 480ohm at 82Hz So, inductance changes with frequency...now why is that? Of course I have read this to be true, but the same can be said of any coil, say a speaker. That is a huge range of impedances, what does that tell us? We can't build anything with a consistent "impedance"...so what kind of thing should we be aiming for? Can you do a comparison with a different shaped coil or something to illustrate how this could change? You know...the FEMM program will calculate coils and stuff, but it is pretty complicated, if you have this you may wish to look into it, perhaps this could aid in these kinds of models. I only used it to get a grasp on magnetic fields, and only in the most basic ways (shapes of fields). It can import vector graphics too I think, so perhaps you could draw models with a better interface than that of the clunky program itself...just a thought. You are on a steep learning curve there, getting further than I was already. On my side of things the jig 4.1 was an improvement, but not good enough. The coil was over-sized and some of the windings escaped the side presses (which has happened before). Some of the improvements, were an improvement but others were not. My original way of ensuring consistent coil depth and removal around the core was better but the jigs made the core 6-7mm wide. Seeing your illustrations and discussions reminded me that such an "air gap" between the core poles and the coil was excessive. So now a whole new jig needs to be made. One thing I have learned is that these things need to be made very "stiff"...the coil has quite a bit of internal pressure and squeezing it with epoxy that has already started to set exerts even more. I am working with very tight tolerances and need to work necessarily quickly. The last one had a better finish and extracted from the jig pretty well, albeit mis-shaped. Inserting the glue before closing the jig seemed to be an improvement too, but takes valuable time...the whole thing needs to be wound and "finished off"...driver leads and all, within 3 minutes of the glue being mixed...so even 30 seconds is significant. I am hoping the next one will be up to the job, it is a combination of the original design (which made quite good coils) and the improvements of the later ones. It is quite difficult to make these things, although a successful one will aid in making more if required. Mainly it is a matter of accuracy...really I should get a CNC to make the appropriate thing. Any fault is magnified in the finished "product" which is not good! I really am thinking I should start setting the coils into store bought covers and be done with it. I have made lots of coils that will fit and work and this way I could address some of the "sunken pole" phenomenon and completely hide the coil so it wouldn't be seen. It would also address the difficulty with the driver wires. Has anyone else had experience with single coil pickup covers and dimensions? I think now that the dimarzio covers that I found had a bridge pickup "F-spacing" but the neck pickups generally do have a spacing of 50mm between outer pole centres, as mine appear to have. Any opinions? pete Link to comment Share on other sites More sharing options...
col Posted September 6, 2007 Report Share Posted September 6, 2007 (edited) See now Col, this is where my brain begins to hurt....and maybe I start making things up...hmmm For an Iron core (again a real driver won't have a pure iron core, but this isn't such a significant variation as the air gap business) of 2mm x 2mm x 55mm with 150 turns of wire, the impedance (inductive reactance) is roughly: 3.5kOhm @ 600Hz and 480ohm at 82Hz So, inductance changes with frequency...now why is that? Of course I have read this to be true, but the same can be said of any coil, say a speaker. That is a huge range of impedances, what does that tell us? We can't build anything with a consistent "impedance"...so what kind of thing should we be aiming for? Can you do a comparison with a different shaped coil or something to illustrate how this could change? look here for a discussion of the same issue in relation to loudspeakers - the very close relatives of the sustainer driver.... The main difference between a speaker and a sustainer driver as I see it is that because the driver of a speaker is moving, it generates back EMF that causes the initial spike in the graph... apart from that, the rising graph looks something like I would imagine an equivalent graph for a sustainer driver to be. You know...the FEMM program will calculate coils and stuff, but it is pretty complicated, if you have this you may wish to look into it, perhaps this could aid in these kinds of models. I only used it to get a grasp on magnetic fields, and only in the most basic ways (shapes of fields). It can import vector graphics too I think, so perhaps you could draw models with a better interface than that of the clunky program itself...just a thought. You are on a steep learning curve there, getting further than I was already. I've gone back to FEMM a few times, but my impression is that with respect to coils, its just a glorified calculator, it doesn't have the same power that it has for permanent magnets.... what I mean is that when you use it for coils, you need to understand the math in order to set it up... while for permanent magnets thats not quite so important... I am working with very tight tolerances and need to work necessarily quickly. The last one had a better finish and extracted from the jig pretty well, albeit mis-shaped. Inserting the glue before closing the jig seemed to be an improvement too, but takes valuable time...the whole thing needs to be wound and "finished off"...driver leads and all, within 3 minutes of the glue being mixed...so even 30 seconds is significant. I'd say that if 30secs is significant, then you REALLY need to change some part of the process I am hoping the next one will be up to the job, it is a combination of the original design (which made quite good coils) and the improvements of the later ones. It is quite difficult to make these things, although a successful one will aid in making more if required. Mainly it is a matter of accuracy...really I should get a CNC to make the appropriate thing. Any fault is magnified in the finished "product" which is not good! I really am thinking I should start setting the coils into store bought covers and be done with it. I have made lots of coils that will fit and work and this way I could address some of the "sunken pole" phenomenon and completely hide the coil so it wouldn't be seen. It would also address the difficulty with the driver wires. Just don't get into too big a financial commitment before you know for sure that the results will be compatible with end user pickups.... btw, sorry that I can't help with your poles poll... I don't have any single coil pickups kicking around to measure. ===================== Back to impedance nerding I found another site with a different calculation method - this time for solenoids rather than transformer style 'closed' (my definition) cores... Unfortunately, it uses an approximation that is only accurate for 'long' solenoids (I'm also guessing that they don't have multiple winding layers but it doesn't say that....) Anyway, interesting thing is that it gives similar results to the other approach I tried... it does state that the relative permeability of 'magnetic iron' is ~200 while on the other site they talk of '98% pure iron' with permiability of 5000.... taking that difference into consideration, the new result of 100ohms @82Hz is in the same ballpark, and that's without the 'air-gap' uncertainty (just a new uncertainty instead lol) Anyhow, this line of investigation is getting getting curiouser and curiouser.... I wonder what difference it would make either to build a driver with a minimum impedance of 8ohm (over the guitars frequency range)... or to use some sort of driver circuit that is designed to drive a higher impedance load... I wonder if this has anything to do with Spazzyone's success with his rail pickup as driver... It may also be part of an explanation why the DIY driver can work while being so much smaller the the commercial units... making the coil smaller with fewer turns of wire reduces the impedance as does using a thinner core (although it's really the ratio of fluxs 'effective path length' to cross section area of the core that matters, so the path being shorter as well kind of cancels this out...) maybe I should take a driver along to Maplin and ask to 'test' a LCR meter cheers Col Edited September 6, 2007 by col Link to comment Share on other sites More sharing options...
psw Posted September 6, 2007 Author Report Share Posted September 6, 2007 Anyhow, this line of investigation is getting getting curiouser and curiouser.... That's what I found, and the whole subject is fascinating. It is one of the things that most has me hooked about this. Magnets are fascinating (even to play with, I have been showing my 5 y.o. how you can push along his hot wheels cars with them, make chains of paperclips, etc) and the idea that you can influence them, or create the effect with a coil of wire and some power...wow. The idea that you can use the signal of a string to drive the string is like following a path on a Möbius strip! I'd say that if 30secs is significant, then you REALLY need to change some part of the process Yes...or get faster and more organized. Speed is good if it can be done, the whole coil would be finished in 20 minutes or so, complete! The problem is the epoxy. I will need to look into something more suitable with a slightly longer work time (say 10 minutes). I did try polyurethane without success, but it's more liquid nature did enable the coil to be more condensed. I could try some 2-pack paint maybe...any thoughts? Just don't get into too big a financial commitment before you know for sure that the results will be compatible with end user pickups.... True, the materials for the jigs and stuff is all being made from left over scrap, even the winder was very cheap to make from stuff I already had from previous experiments. A lot of the various versions of the jigs re-used the metalwork too. The investment is largely in time, but there is no doubt that if one is going to try to make something like this, a lot of experimentation will have to be done to get things right, if only to develop the skills once a method has been worked out! btw, sorry that I can't help with your poles poll... I don't have any single coil pickups kicking around to measure. Yes...well no takers, but research on the web suggests a 50mm gap between outer poles is typical for neck and middle pickups and the wider F-spacing is for bridge pickups, as is a 5mm core width. I will add a mill or so to the ends and have reduced the core down to 5mm. I wonder what difference it would make either to build a driver with a minimum impedance of 8ohm (over the guitars frequency range)... or to use some sort of driver circuit that is designed to drive a higher impedance load... I wonder if this has anything to do with Spazzyone's success with his rail pickup as driver... I had thought that with enough power any coil could work on some level and that may have been his secret...Shawn did use a PA to run it . I have had thoughts repeatedly about trying to drive a high impedance load, and therefore potentially use a conventional pickup. Micheal Brooks' groundbreaking (yet secretive) Infinite Guitar seemed to work like this with a very large transformer on the back of the guitar and a powerful off board amp...or so it would seem. It may also be part of an explanation why the DIY driver can work while being so much smaller the the commercial units... making the coil smaller with fewer turns of wire reduces the impedance as does using a thinner core (although it's really the ratio of fluxs 'effective path length' to cross section area of the core that matters, so the path being shorter as well kind of cancels this out...) This line of thinking is very interesting...I'll PM you some data from the workbench to ponder on . I do think there is something "special" going on with the thin driver but perhaps my make believe theory to explain it could be replaced with something a little more solid. Experimentation shows it to work a lot better than other designs, so something must be going on! After all, everyone else seems to have gone along with the "concept" despite making stand alone drivers that could exploit any kind of coil... Good stuff, don't get a brain drain and I will try not to inhale too much glue ... pete Link to comment Share on other sites More sharing options...
curtisa Posted September 6, 2007 Report Share Posted September 6, 2007 Does anyone own or have access to a multimeter that reads inductance or an LCR meter? if so please take some measurements of your driver (even if it doesn't work well ) Hello once agin Col (been a while!), I used this site when I custom-wound some multi-tapped inductors for an EQ I built a while back: http://et.nmsu.edu/~etti/fall96/electronic...uct/induct.html You will need access to a sinewave signal generator and an AC voltmeter that can read 1KHz accurately though. However, you don't need to worry about permeability of different cores and air gaps to get an accurate reading, just measure what you see and plug in the numbers into the formula. If I get any spare time over the next few days I'll do the measurements on my driver and see what I come up with. Lets have a Poll... Anyone following this tread with single coil pickups of any variety, could you measure... 1~ the width of the poles (I get about 5mm) 2~ the span of the poles (I have been working on 55mm, but have added a mil on each end to be sure) 3~ the bobbin width (if you can take a cover off...I am working on 15mm max) 4~ the amount of space you estimate there is between the cover and the bobbin for the driver wires to get by (I get less than 1mm) 5~ an opinion on the effect of raising the cover 1mm or so, effectively lowering the pole pieces. Same here Pete - I'll make some measurements as time permits on my SC pickups and report back. Link to comment Share on other sites More sharing options...
mrjstudios Posted September 7, 2007 Report Share Posted September 7, 2007 When I get some free time I've got 2 different SC pickups that I can measure... later. -MRJ Link to comment Share on other sites More sharing options...
psw Posted September 7, 2007 Author Report Share Posted September 7, 2007 Thanks...all I just received an industrial quantity of 0.2mm wire so will have plenty to practice on, have wasted quite a bit in recent days Also...working on jig MK5...this time for sure, Rocky! This one is very solid and accurately made from aluminium and plastic, has a no tools simple release mechanism to save time and can be taken apart so that it's parts can make duplicates! This thing will have a highly polished surface and should make a fine driver. It is also made to the exact shape and size of a typical bobbin...so the coil will have to fit, wont it! The core has been reduced to about 5.5mm wide and about 56mm long. I believe this is the most typical SC size. The bobbin/coil is a maximum of 15mm wide and 66mm long. I'm pretty proud of this one, it's almost solid aluminium so there will be absolutely zero flex. The thing comes completely apart for cleaning and making duplicates and the "locking system" and spindle parts are of hardened steel. Best of all, all the parts for this were about from years of fiddling around with various things and if I don't use them now, I can't imagine justifying keeping all this stuff. Essentially, it was free...bonus! The time it has taken so far...ridiculous. The problem is to make the sides compact so that they fit within the bobbin size. I tried Tim's clamping ideas, but was having a few troubles with this. The machine does help a lot, perhaps I need to add some kind of wire tensioning device that varies between the long and short strokes of the elongated coil...hmmm. better run, but progress is being made, even if a little slow... pete Link to comment Share on other sites More sharing options...
psw Posted September 7, 2007 Author Report Share Posted September 7, 2007 Ok...so wound the second coil from this new jig, definitely on to something which is a bit of a relief. Was able to make a coil to specifications (i.e. will fit within the cover) but worked out to be 7.6 ohms. This is an acceptable resistance, but there is still a bit of "air" in the coils and my inability to remove it is in large part to do with the epoxy setting so fast. Either less viscus, longer setting stuff needs to be found, or I will have to shave a minute off the winding time. I even seem to have solved the driver lead exit thing and this one is starting to look good. As with everything, practice makes perfect...if I have enough of these "dud drivers" (i.e. not of salable quality but still usable) I may be looking to off load them . Hopefully, this will see the end of the jig building and getting back on track with the device itself. I have plans for my strat...have bought a neck and middle fender "Jeff Beck" noiseless pickups on eBay (also a set of locking machine heads). These pickups have a really interesting magnetic structure and it will be interesting to see how my "thing" works with this kind of pickup. Also found a cheap supplier of push/pull pots, so I can do some fancy wiring of the thing as well. The plan is still to use the super switch to activate the device as one of the pickup settings. A push pull to be used to operate the harmonic mode. As I have not yet found an easy source for 1K push/pull for the drive control, I may have to take the pot apart and substitute a resistance wafer from another pot to change it's value. I may also use a switch to change some of the pickup selection combinations when pulled and perhaps a third to operate some kind of cheesy piezo thing as I had been experimenting with on the other strat. There have been several projects at PG on this kind of thing where you attach a piezo element within the guitar (often on the trem block...I had it in the neck socket for a while) and a little preamp. There is some potential for this kind of thing as long as you don't expect an acoustic sound to it. Mixed with a magnetic pickup, the piezo provides a unique attack and texture. Besides these future tricks, this guitar is really supposed to demonstrate how "low mod" this device could be, and practical in it's installation and use. Of course, in addition to the coil and circuit I am developing, switching will need to be arranged, and this could mount up...most people don't have a super-switch (I installed one for the mid-driver experiments) or a push pull pot, and these could cost almost as much as my hand built system...go figure! Toggle switches and rotary controls are cheaper options, but still could cost a bit and have to be found. Anyway...will try and find an alternative epoxy or something somewhere and keep practicing winding. If I use a too slow epoxy I will have to consider making multiple jigs so that I can keep winding while they set and not waste glue. At least with this "speed" winding method, the whole driver comes out without waiting! I even worked out a super novel way of releasing the coil from this new jig...every little innovation helps. later... pete Link to comment Share on other sites More sharing options...
col Posted September 7, 2007 Report Share Posted September 7, 2007 (edited) I used this site when I custom-wound some multi-tapped inductors for an EQ I built a while back: http://et.nmsu.edu/~etti/fall96/electronic...uct/induct.html You will need access to a sinewave signal generator and an AC voltmeter that can read 1KHz accurately though. However, you don't need to worry about permeability of different cores and air gaps to get an accurate reading, just measure what you see and plug in the numbers into the formula. If I get any spare time over the next few days I'll do the measurements on my driver and see what I come up with. Thanks, I hacked together a probably innacurate test rig using my sound card and NI Reactor to generate a sine wave and my DMM for measurements.... The DMM is only accurate up to 500Hz, so I used 440.... also didn't use quite the resistance values they suggested... I probably will try again with different values to double check. Anyway, I got results that seem reasonable - @440 Hz, my old single coil driver has a total impedance of ~20ohm. At 82 Hz (low end of guitar range) the inductive reactance doesn't have a huge effect... The resistance of this driver is 7.3ohm, the overall impedance @ 82Hz is ~8.9 ohm... that probably explains why Pete was finding that 7ohm resistance drivers seems as good or better than 8ohm ones ! here's a thought: If the wire was a thicker guage, a coil with an 8ohm DC resistance would have more turns and therefor a higher inductance and impedance... so it would be further away from the optimal 8ohm impedance ? So Pete's tests would have got closer to the truly correct impedence as the wire got thinner... could be one reason why he ender up with 0.2mm as the 'optimum'? So maybe using thicker wire could give us better results as long as the driver is wound to 8Ohm impedance rather than resistance ? That would give us a stronger magnetic field for the same current == a more efficient system... ? In that case, the limiting factor becomes the size of the coil... thicker wire with more turns means a lot more bulk. Does that sound plausible ? cheers Col Edited September 7, 2007 by col Link to comment Share on other sites More sharing options...
psw Posted September 7, 2007 Author Report Share Posted September 7, 2007 Does that sound plausible ? Now I will have to get the hang of impedance resistance and resistance, resistance.... Checked out your Wiki Link A lower "resistance" may draw more current from the amplifier...will this create a stronger electromagnetic effect? BTW...My made up pseudo-science in the water analogy is my way of understanding impedance I guess...perhaps not a good one...the idea that moving water faster through a given hose size will create more resistance....hmmm Not all my drivers, especially in the early experiments were made with this wire of course. I was using much thinner wire initially (0.125 probably) and did try thicker wires...often wound to "size" disregarding counts and resistance (now there science for you!). So, I probably did do some with thicker wires with a lower impedance, but all the designs changed, so that won't prove anything. So Pete's tests would have got closer to the truly correct impedance as the wire got thinner... could be one reason why he ender up with 0.2mm as the 'optimum'? I most likely approached it from a thinner gauge...but I did go thicker and thinner, as have others and not had the success. Of course, this really only holds true for my single coil drivers, dual coil drivers may benefit from a different gauge. So maybe using thicker wire could give us better results as long as the driver is wound to 8Ohm impedance rather than resistance ? That would give us a stronger magnetic field for the same current == a more efficient system... ? Maybe, but it may consume exorbitant amounts of power to produce the required electromagnetic effect. In that case, the limiting factor becomes the size of the coil... thicker wire with more turns means a lot more bulk. And so now we return to practicalities and if the system works. That is where I am still at for now, the new design seems to work but I am sure that many compromises are being made in the pursuit of my particular application of it. Equally, I think there is a lot we don't know about why a thin coil works as well as it does, and I'd love to know why...good stuff col...keep nutting it out! I wonder, I think I may know some people I could approach to do some testing eventually. I recently heard that my high school science teacher that I have had intermittent contact with over the years, works extensively in tiny coils for medical instruments...uses sound waves to move tiny laser cameras in veins or something, I bet he has some appropriate equipment. Anyway, I got results that seem reasonable - @440 Hz, my old single coil driver has a total impedance of ~20ohm. At 82 Hz (low end of guitar range) the inductive reactance doesn't have a huge effect... Hmmm...you will want to have the most efficiency in the mid to higher ranges I imagine as the heavier strings are easier to move. It is interesting that when people used 0.25mm wire the most consistent symptom is that it couldn't move the higher strings. The resistance of this driver is 7.3ohm, the overall impedance @ 82Hz is ~8.9 ohm... that probably explains why Pete was finding that 7ohm resistance drivers seems as good or better than 8ohm ones ! That is reassuring as getting a true 8 ohm coil in my new design is very difficult. My latest one is 7.4 ohms. The windings are extremely compressed, however you can tell from the width of the coil at the ends, relative to the sides, that a little more is theoretically possible to squeeze on there. A thinner slower acting epoxy would help there, but it is possible that I won't be able to achieve 8 ohms and that it doesn't seem to matter that much. It does seem a little odd that with all this theory that a 1 ohm difference in rating shouldn't have a more dramatic and noticeable effect. I couldn't say conclusively if or what differences there may be. My Strat pickup/driver is in the low 7's but the coil wound to test feasibility (same coil but built off the pickup) was exactly 8 ohms. However, although the core was the same material in both...on the pickup it extends down through the pickup coil to a magnet below and with the stand alone driver, it was only the depth of the coil itself. I'd be good at science, except for the methodology ==== Not to worry, at least I seem to be getting a coil out of the latest rig that is half presentable and have something I can work with to refine the process. The coil I made yesterday is probably good enough for the beta-testers to have a go at. Better do some housework though, there appears to be a lot of metal filings about the place...grrrr! later... pete Link to comment Share on other sites More sharing options...
col Posted September 7, 2007 Report Share Posted September 7, 2007 (edited) Now I will have to get the hang of impedance resistance and resistance, resistance.... Checked out your Wiki Link A lower "resistance" may draw more current from the amplifier...will this create a stronger electromagnetic effect? More turns of wire around the same core means a stronger magnetic force for the same current (I think?), so thicker wire with lower DC resistance allowing more turns might be worth trying... So Pete's tests would have got closer to the truly correct impedance as the wire got thinner... could be one reason why he ender up with 0.2mm as the 'optimum'? I most likely approached it from a thinner gauge...but I did go thicker and thinner, as have others and not had the success. Of course, this really only holds true for my single coil drivers, dual coil drivers may benefit from a different gauge. You're going to like this I just discovered that I made an error in the calculations that my last post was based on ! it turns out that for my 7.3ohm DC resistance single core driver (my first driver) the reactance is a lot lower.... at 82Hz, it is just under 1ohm ! the calculation for impedance - sqrt( resistance^2 + reactance^2 ) means that if either the resistance or the reactance is much larger, then the other doesn't do much ! Sooo, what we have is a driver where the impedance stays at roughly 8ohm from 82Hz right up to about 800Hz due to the DC resistance being dominant - at this point the reactance becomes dominant, so the impedance starts going up linearly with frequency... 800Hz is right up near the top of the guitars range - think high E string somewhere near 15th fret. So if any of this is correct, it might be the real reason for the even frequency response (maybe replacing the theory about it being a 'fast' driver ?) right up to the high notes... ? This error in my calculations still doesn't mean that using a heavier wire guage for testing higher inductance drivers is a bad idea... the inductance increases with the square of the number of turns on the coil - just going up to 0.34mm wire would, assuming the same 8ohm DC resistance, mean 4 times the inductance and 4 times the reactance. At this setting, the inductance would become dominant at something like 200Hz, so we would start to see more problems with frequency response, however the magnetic field would be stronger... The driver would also start looking more like the commercial units ie. bulkier. and the circuitry would probably need at least some sort of filtering to counteract the impact of the reactance above 200Hz.. does compensation circuitry sound familiar to you? Also, we mustn't forget that these calculations are based on some pretty ropey measurements on an old and not optimum driver. Any of the following can have an effect on these features variation in core material - could be very significant (as in 'orders of magnitude' !) variation in core dimensions - not as significant, but still noticeable variation in wire guage - even the difference between 0.2 wire and 0.25 wire can make a big difference !! quality of coil construction - could be significant Fortunately the permanent magnet doesn't seem to make any difference with respect to impedence - so thats at least one less thing to worry about Maybe, but it may consume exorbitant amounts of power to produce the required electromagnetic effect. Or it may just be a more efficient way of turning the current into flux ? If the impedance is the same, then the same amount of power will be used, however, more turns of wire make a stronger magnet assuming the same current is flowing.... ALTHOUGH I don't know when we would reach core saturation - that may be a long way off with this system, or we may already be there - no intuition on that one... ALSO, as the wire gets thicker, the flux linkage is reduced, so there will be a point at which thicker wire starts making performance worse not better... AND as the inductance increases, other stuff like eddy currents will also have more of an impact... and then there's the issue of frequency dependent response I wonder, I think I may know some people I could approach to do some testing eventually. I recently heard that my high school science teacher that I have had intermittent contact with over the years, works extensively in tiny coils for medical instruments...uses sound waves to move tiny laser cameras in veins or something, I bet he has some appropriate equipment. He sounds like our man for sure :-D ...It is interesting that when people used 0.25mm wire the most consistent symptom is that it couldn't move the higher strings. Yes, see above with the new corrected analysis, my 7.3ohm driver of 0.23 wire only starts to increase impedance with frequency at about the 15th fret on high E string... with 0.25 wire, it would start at roughly the 10th fret. That is reassuring as getting a true 8 ohm coil in my new design is very difficult. My latest one is 7.4 ohms. The windings are extremely compressed, however you can tell from the width of the coil at the ends, relative to the sides, that a little more is theoretically possible to squeeze on there. A thinner slower acting epoxy would help there, but it is possible that I won't be able to achieve 8 ohms and that it doesn't seem to matter that much. It does seem a little odd that with all this theory that a 1 ohm difference in rating shouldn't have a more dramatic and noticeable effect.... a 1 ohm difference in a 8ohm coil is 12.5 %... sounds more significant now right ? that 1 ohm difference is almost exactly the same number of turns as the step from 0.23 wire to 0.25 wire, so can be the difference between response dropping off above the 15th fret (where it is being counteracted by the action pushing the string closer to the driver) and at the 10th fret where loss of performance would be much more noticable... think about all that dodgy blues soloing at the 12th fret blues box... the E and B string would start to be affected..... Also with other differences in driver construction and materials, its likely that some folks have ended up with drivers that have this issue arising at lower frequencies.... Anyway for all the guesswork and dodgy measurements, at least I feel like I'm starting to understand the thing a bit better... It will be interesting to see if curtisa has similar results to mine... cheers Col Edit: thinking more about dodgy measurements and how I could be out by quite a bit with these results. It would make a lot of sense if the impedance started rising with frequency at or under around 200 - 300 Hz, this would help to explain why so many people have a lot of trouble with their B and E strings... Edited September 7, 2007 by col Link to comment Share on other sites More sharing options...
psw Posted September 8, 2007 Author Report Share Posted September 8, 2007 Sounds like you are doing some great work that proves the "thin core theory"...at least in the important specifications, if not the shape. ALTHOUGH I don't know when we would reach core saturation - that may be a long way off with this system, or we may already be there - no intuition on that one... ALSO, as the wire gets thicker, the flux linkage is reduced, so there will be a point at which thicker wire starts making performance worse not better... AND as the inductance increases, other stuff like eddy currents will also have more of an impact... and then there's the issue of frequency dependent response Well saturation and such depends on the core materials and the strength of the magnet, and the core size. With some drivers I know I have exceeded saturation, the Hex ones for instance did, and possibly the mid driver with it's small and very thin hardened steel blades. What effective difference it makes, I really am not sure. Permeability does have an impact though, but a very hard one to address in practical terms. So if any of this is correct, it might be the real reason for the even frequency response (maybe replacing the theory about it being a 'fast' driver ?) right up to the high notes... ? Well...I guess that is what I would define as a "fast driver"...one that will react efficiently in the ranges we are operating on. I wonder if this low (1 ohm) impedance in the lower notes has some bearing on the excessive drive of the lower strings, as much as their mass and tension? It also makes you consider a multi-coil design with drivers calibrated to each string. This was the ultimate intention of the Hex drivers. Each driver in the prototypes were essentially identical, but it was the intent that to make the thing more "polyphonic" and even in response from a mono signal by means of some kind of crossover, most likely by altering the values of each individual driver to better match the string it is driving. It also make you think about the effect that the dual parallel coils were having. 4x the number of turns/wire is a huge difference. a 1 ohm difference in a 8ohm coil is 12.5 %... sounds more significant now right ? that 1 ohm difference is almost exactly the same number of turns as the step from 0.23 wire to 0.25 wire, so can be the difference between response dropping off above the 15th fret (where it is being counteracted by the action pushing the string closer to the driver) and at the 10th fret where loss of performance would be much more noticable... think about all that dodgy blues soloing at the 12th fret blues box... the E and B string would start to be affected.....[edit: psw } Ahh...but I thought the point we had both pursued here was that the difference between 8 ohms and 7 ohms is not that noticeable (not that any conclusive results have been put forward). It was even intimated that it could be "better" to be lower (perhaps based on the previous calculation errors?). So...are you now suggesting that an 8 ohm driver is likely to have a better response in the higher strings and that perhaps an 8.5 may have slight advantages up there? Maybe a driver of thinner wire but similar turns taking it up to 10 or 12 ohms? Also with other differences in driver construction and materials, its likely that some folks have ended up with drivers that have this issue arising at lower frequencies.... Quite right! Although, don't you mean "higher frequencies". The difference in construction even if following a specific design and specifications does have a big impact. My new venture/drivers were an attempt to address these inconstancies of course, and to improve the build quality significantly. The reason I have been able to make such thin cores lately is mainly down to these factors. Since I appear to have shrunken the same device essentially by two thirds over my original with the same specifications, you can see that even a successful and well wound coil such as that in the original S*strat still contained a fair degree of internal air. A lot of theory and such is based on assumptions and ideals difficult to attain in the real world. Making a tight coil with windings close and parallel without "air" and internal inefficiencies is tricky. As is the calculating of "scatterwinding" and its use to ameliorate eddy currents within the coil. There are so many variables and considerations and variables to those factors in turn, that eventually, I stopped looking. It is interesting to see that some of the theory supports some of my work, even if my intuitive explanations may be off the mark by a long way. These "intuitions", even if I do struggle to explain them, are real and serve a purpose, even if technically wrong! They gave me a means to explore various avenues with success. There is a limit though to working with hunches and at this stage of refinement, it is above these limits that we will need to explore. Meanwhile, I have committed to my design and specifications for now and will continue to develop that. The hope is that the work I am doing will provide a starting off point for further explorations...not an end to it, but an equal footing for any who wish to take it up. The idea I am pursuing is the best possible adaptation/implementation that I can come up with at this time. The machine and even the latest jigs are adjustable however so thicker coils and different wires could be tried out, without variances in quality...and even compared with the same core materials and magnetic structures, right on the instrument! pete Link to comment Share on other sites More sharing options...
psw Posted September 8, 2007 Author Report Share Posted September 8, 2007 Doing tests on a different "glue" that seems to be an improvement as far as making a good compact coil to the correct specs and allows for some finishing touches difficult or impossible, or no space allowance, by the previous method. How it "sets" will be another issue however but it is much easier to work with and a lot less tonic too. Unfortunately, it is going to take a long time to set so it will have to wait to see how it comes out, or even if it does come out without damage. A coil wound with similar material failed last week, but I kept it and is now setting fairly hard. This seems to be better stuff (a little expensive, but has other uses I guess) with virtually no wastage or shelf life. If it works out ok, it does set a problem for "production" because I will have to make them one by one or make multiple jigs so they can be made in batches. Anticipating that this may occur and in order to make it accurate, a lot of work went into the jig design and templates were made to make it to ensure accuracy. This means that duplicating jigs is possible (the first one took about 16 hours to complete) but a bit of a pain. However, the benefits of not having the time constraints of the traditional epoxy is pretty great...a better coil more consistently and a lot less stress, it is actually enjoyable making this one (which is kind of the idea)! If it works, the coil solution may well be around the corner... pete Link to comment Share on other sites More sharing options...
mrjstudios Posted September 8, 2007 Report Share Posted September 8, 2007 Lets have a Poll... Anyone following this tread with single coil pickups of any variety, could you measure... 1~ the width of the poles (I get about 5mm) 2~ the span of the poles (I have been working on 55mm, but have added a mil on each end to be sure) 3~ the bobbin width (if you can take a cover off...I am working on 15mm max) 4~ the amount of space you estimate there is between the cover and the bobbin for the driver wires to get by (I get less than 1mm) 5~ an opinion on the effect of raising the cover 1mm or so, effectively lowering the pole pieces. These are the measurements of my crappy strat copy's pickups. (They are probably the most standard and generic SC pickups that I can find.) ________________________________________________________________________________ _ 1- 4.95mm 2- 57.08mm 3- 14.36mm 4- Bobbin touches both sides of the cover, but 2 small coil wires could be squeezed through easily. 5- I don't care, and I am 100% sure no one in the audience will either, or even notice. ________________________________________________________________________________ _ -MRJ Link to comment Share on other sites More sharing options...
Recommended Posts