col Posted September 8, 2007 Report Share Posted September 8, 2007 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. How do you know ? What effective difference it makes, I really am not sure. Once you have reached saturation, any extra energy you pump into the thing will be mostly wasted... One of the design techniques for some wound components is - assuming you have some target inductance - to design the thing so that it will be working just under the saturation point when achieving that target... this lets you minimize materials and space whilst maximizing efficiency... 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? I think a 1 ohm impedance would fry the LM386 at the lower end of the guitars frequency range, the reactance is ~ 1ohm, but the impedance is around 7 or 8... (assuming no significant capacitance) the impedance is a combination of the resistance and the inductive reactance. The inductive reactance is 90º out of phase with the resistance, so we can use the sums from pythagoras theorem to do the calculation.... one side of the triangle (opposite side) is the resistance R, the other (adjacent) is the reactance X... so the impedance Z is the long side (hypotenuse) So we go Z = sqrt( Rsqd + Xsqd) what this means is that with a constant resistance of 7ohm, a reactance varying from 0 to 8 gives a impedance varying from 7 to 10, which is the correct ballpark for us So the total impedance will never be 1ohm (although, the reactance will be) 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. Yes, I have been considering the permutations today fwiw, two coils in parallel each with double the number of turns (assuming the same core dimensions) would result in the overall impedance being doubled.... double the number of turns means 2 times the resistance and 4 times the reactance for 1 coil, but if you connect them in parallel, the combined impedance is (using a nice complex impedance calculator i found) : @82 hz, roughly 7.5 ohm..... at the point where reactance equals resistance in our single coil driver giving that 10 ohm total result it would be: roughly 17.5ohm ! also, the total impedance will be more reactive (whatever that means) Question: do you know how to manipulate complex numbers e.g. 2 + 34i (where i is the square root of -1) ? (getting any further into an understanding of the relationship between reactance and resistance will be tricky otherwise. In this respect I'm lucky in that I got the basics internalized years ago when I got involved in writing fractal generators) Another thing that I've been thinking about is how all this stuff effects the idea of a bi-lateral driver ! I think that there are a number of advantages to this layout given this slighly improved understanding - and probably some disadvantages. Now I have to work out what impact (if any) the reactive component of the impedance has in our system... e.g. does a mostly resistive 8ohm impedance behave the same as a mostly reactive 8 ohm impedance ? * see note at end 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? No, the opposite in fact ! assuming the 8ohm driver is only different in that it has more turns of wire, its reactance will be higher, so the point at which the impedance starts to rise with frequency happens at a lower note frequency e.g. at the 10th fret instead of the 15th... that makes the 7 ohm driver better on those high notes ! Also with other differences in driver construction and materials, it's 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". Hehe, no. I don't mean how often people have problems... I mean the note frequency at which the problem effects theresponse of the sustainer device ------------------- * Of course, an inductor acts as a low pass filter - similarly to a capacitor acting as a high pass filter.... So even if the impedance stays roughly the same level, if more of it is an inductive reactance, more high end content will be rolled off ! I guess this is what the extra circuitry in the commercial units is for - to counteract the filtering characteristics of the driver. Once we have a more complete picture of this whole mess, and have experimented more with the different options, it should be possible to go some way towards a similar 'compensation' circuit... we basically need to tune the zobel network at the LM386 output to work specifically with the characteristics of the driver.... For optimum performance, each person will have to go through a process to arrive at the correct zobel values for their own unique driver ! ("A Zobel is a series resistor-capacitor (R-C) network that is connected in parallel with a loudspeaker driver in order to neutralize the effects of the driver’s voice coil inductance L(e).") Link to comment Share on other sites More sharing options...
psw Posted September 8, 2007 Author Report Share Posted September 8, 2007 Grrr...another coil failure!!! The jig is good (though got slightly damaged due to over enthusiastic tightening) and the coil was looking superb...this kind of stuff just wont set without air. (waiting 24 hours to find out if it works is a bugger too!) I guess I have no choice but to use epoxies of some kind, but buying specialized stuff will be expensive...shame cause I had some interesting stuff before the "move" and I don't know if it was thrown out or where it has got too, or even if it may have passed it's shelf life! Getting a coil to exactly 8 ohms is a bit tricky too...it is possible, but each additional wind adds to the precision required to produce it within the size restrictions. 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~ 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. 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. Thanks MJS... this illustrates the potential problem here. Measuring the best epoxy coil from the new jig I get a maximum internal core length (distance between outsides of pole edges) ~2 above, as just slightly under 56mm so almost 2mm short! This is a typical F-spaced bridge pickup spacing, I wonder if the matching pickups are the same (ie was this a neck pickup) or the neck slightly wider (all my jap-strats are thinner). I wondered too about Ibanez guitars, they tend to have wider necks don't they? Bobbin width ~3 is also a problem, 14.36 is pretty skinny. Measuring this same coil, it measures 16mm...there was a little "spring" in it once it was released from the mold. This press fitted into my standard squier/fender SC covers nicely. ~4 I think is typical, there were some concerns raised about the look of the driver and the way the leads exit it, plus concerns about irreparable damage should they break. This is a concern and does effect the look and professionalism of the coil over simply a working DIY prototype. I think I am refining a good idea on this while keeping the thing small enough to fit between the covers. Bear in mind that there is only 1mm of bobbin height and less between the covers to make make this happen! Could you also measure the length of the bobbin in total there MJ, if this were to be a common variation, then perhaps a special jig may need to be made for a special order to accommodate them. Otherwise, I will have to be very specific about what this thing will fit...this is not something you can simply sand the edges off to make it fit! So...I am not sure what you used to arrive at those particular figures (they are extraordinarily accurate for metric, did you convert them from imperial perhaps?) but on the face of it, these coils wouldn't fit on this pickup Not Good! The other thing is that it may be an old tele neck pickup, these are a fair bit narrower...I would need to make a different coil for these at some point. The thing is, I can not lengthen the length of they won't fit a standard pickup, nor narrow it down too much (though I have been trying) and the core is set for this particular jig. I know 1mm is very small, but considering this is the total height of the coil itself, these things are huge! ====== Some really interesting stuff going on there Col...I recognize that you are feeling your way, but I simply don't have the "mind" for these types of things and the concepts are difficult for me to juggle as well. That is why I need to "make up" metaphors to keep a handle on them. You are doing a lot better than me! How do you know ? Tricky...ok I don't "know" (but the hex designs had elements of less than 3mmx3mmx10mm with powerful neodyminium magnets attached to both ends!) but I assumed. Got me there! Once you have reached saturation, any extra energy you pump into the thing will be mostly wasted... One of the design techniques for some wound components is - assuming you have some target inductance - to design the thing so that it will be working just under the saturation point when achieving that target... this lets you minimize materials and space whilst maximizing efficiency... Perhaps that is why they worked as well as they did...against the odds! what this means is that with a constant resistance of 7ohm, a reactance varying from 0 to 8 gives a impedance varying from 7 to 10, which is the correct ballpark for us .... assuming the 8ohm driver is only different in that it has more turns of wire, its reactance will be higher, so the point at which the impedance starts to rise with frequency happens at a lower note frequency e.g. at the 10th fret instead of the 15th... that makes the 7 ohm driver better on those high notes ! Ok...good! I was thinking that I may need to lower my expectations to around 7 ohms for the new driver as squeezing that extra ohm is pushing the limitations a little. The last one, although failed, was as compressed as this thing could ever be, and measure 7.5! Question: do you know how to manipulate complex numbers e.g. 2 + 34i (where i is the square root of -1) ? NO!!!!!! * Of course, an inductor acts as a low pass filter - similarly to a capacitor acting as a high pass filter.... So even if the impedance stays roughly the same level, if more of it is an inductive reactance, more high end content will be rolled off ! I guess this is what the extra circuitry in the commercial units is for - to counteract the filtering characteristics of the driver. Once we have a more complete picture of this whole mess, and have experimented more with the different options, it should be possible to go some way towards a similar 'compensation' circuit... we basically need to tune the zobel network at the LM386 output to work specifically with the characteristics of the driver.... For optimum performance, each person will have to go through a process to arrive at the correct zobel values for their own unique driver ! ("A Zobel is a series resistor-capacitor (R-C) network that is connected in parallel with a loudspeaker driver in order to neutralize the effects of the driver’s voice coil inductance L(e).") Will PM you about my circuit's qualities that may effect this. I have long advocated the use of a "Zobel network" and used the one from the data sheet consistently...other's have not. The Floyd Rose Patent has a good description, even some values from memory of an automatic phase compensation (lead/lag) circuit that very much looks like a variable zobel network that is triggered by frequency response. This is exactly what I was alluding to in my hose pipe, faulty analogy. It is just my way of comprehending and holding this kind of information in a way that is useful to me. The higher the frequency, the more the response of the driver "lags" in response, requiring compensation (just as my water current, magnetic fields need to collapse, and reverse and that this is not instantaneous and varies with frequency). Hence the "fast driver" to me, is one that will operate at such a speed at all relative frequencies that this lag time is relatively inconsequential. I used to fret about these things and was unable to do something "electronically" with my limited abilities, and at best try to emulate what had already been done. LK encouraged me to develop the drivers (where I had more skill) and the hope was to develop something that could work without compensation, etc. Hmmm.... So, before I am even completed a fully production prototype I am coming up against the limitations of the application...bugger! Still, from other research I seem to be finding from the big pickup makers, my pole spacing for neck strat pickups does seem to be correct (even if I have to convert their imperial measurements to metric). Metric is much better for things like this, I was brought up with inches and feet but Australia changed to metric. For things like this metric is so much easier (dividing a centimeter into mm and decimal places is much better than fractions of an inch). Measuring the height of someone I will always do in feet and inches...horses for courses I guess. I will have no choice but to find a better "glue" for my application or work a lot faster with what I have got. The forces in the coil are considerable, especially when tightly compressed. The air tight environment makes other systems impractical. It is the viscosity that seems to be the biggest problem with epoxies...will need to search further for a suitable substance. ($$$) Will keep working on it... pete Link to comment Share on other sites More sharing options...
curtisa Posted September 8, 2007 Report Share Posted September 8, 2007 Grrr...another coil failure!!! The jig is good (though got slightly damaged due to over enthusiastic tightening) and the coil was looking superb...this kind of stuff just wont set without air. (waiting 24 hours to find out if it works is a bugger too!) I guess I have no choice but to use epoxies of some kind, but buying specialized stuff will be expensive...shame cause I had some interesting stuff before the "move" and I don't know if it was thrown out or where it has got too, or even if it may have passed it's shelf life! What about using modelling cement as the curing agent? The other thing I thought about when I was at Bunnings the other day, there is some two-pack product they have there in the craft section that comes in tiny little bottles all the way up to big 2 litre tubs. It's called Glasscote or Gloss coat or something like that. It's really meant to be used to pour over flat surfaces to give a super high glossy clear finish without brush strokes. It's also super hard wearing. 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~ 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. And mine on my Yamaha Pacifica 112 are the following: 1. 5mm 2. 57mm (outer edge of bass polepiece to outer edge of treble polepiece) 3. 15mm 4. Zilch, see number 5 5. Don't care either, but the design of the Yamaha pickups are such that the polepieces are flush to the top surface of the pickup with the cover in place, so anything under the cover will make the polepieces disappear under the surface of the cover. The cover also has a fair bit of slop in fitting over the pickup so you could easily get some small guage wires between the bobbin and the cover to make the external connections. The only drawback is that the cover is aligned properly over the pickup by the polepieces which then stop it from wobbling around when the whole assembly is put together and held by screws. So raising the pickup cover by even a small amount by putting the driver underneath will put the whole thing out of alignment and it will allow the cover to move around over the pickup. The resulting "sunken "polepieces, which don't bother me much aesthetically, may bother other users. Also the fact that the polepieces are below the level of the pickups may cause dirt and gunk to enter the pickup/driver and cause rust and damage. Looking at the construction of the Pac112 single coil pickup it looks identical to the pickup I used to make my original sustainer (pickups were from an early trashed Yamaha RGZ). On that one I hammered out the 6 slugs and cut a new slot into the bobbin to accept a taller steel blade, so I didn't have to worry about sunken pole pieces or misaligned pickup covers. I think the idea of being able to universally fit the driver under a single coil pickup cover may need to be rethought as not everyone will be able to drop in the driver under the cover without some major re-working of the pickup itself - say replacing the stock picup slugs with longer ones. Question: do you know how to manipulate complex numbers e.g. 2 + 34i (where i is the square root of -1) ? Complex numbers, vector numbers etc. The "2" bit is the "real" part of the number. The "34i" bit is the "imginary". It's a way of experesing vectors in one mathematical "term". Think of the two parts as the "run" and a "rise" or a triangle - the 2 is the horizontal base of the trinagle running from left to right, the 34i is the vertical side of the triangle running from bottom to top, both lengths being at right-angles to each other. The hypotenuse of that triangle (ie the third line joining the start of the 2 on the bottom-left, to the end of the 34i on the top-right) is the value given by the expression "2+34i". When talking about impedances the "2" is the DC resistance and the "34i" is the reactive impedance. Volts and current can also be expressed in such ways, showing amplitude and phase shift. You can also re-write the expression in more useful ways by working out via trigonometry the angle and amplitude of the hypotenuse of the triangle. So by trig, 2+34i would equal 34.06 at an angle of 86.63 degrees. Errrrr...dunno if that really helps or not... I'll do the impedance measurements on my driver today and get back to you. Link to comment Share on other sites More sharing options...
psw Posted September 9, 2007 Author Report Share Posted September 9, 2007 Ok...thanks Curtis... And mine on my Yamaha Pacifica 112 are the following: 1. 5mm 2. 57mm (outer edge of bass polepiece to outer edge of treble polepiece) 3. 15mm 4. Zilch, see number 5 ok it looks as if I will have to remake the jig with a slightly longer inner core dimension for these "non-fender" variations. This is important information to know, but further puts pressure on the specs of the driver. While making a longer core is not a problem (other than having to rebuild the entire jig ) , it will add length to the total driver meaning that the specs will have to fit the smaller total bobbin width (say of mine and typical fenders) and a wider core. I think it can be done, but it will require some even more "clever" manufacturing techniques. I don't think there can be "zilch" space between the cover and the bobbin. There must be some or it would be extordianrily tight. The driver wires are 0.2mm so they could be squeezed with my proposed method of security to about 0.6mm. If there really isn't that much room, then a small nick in the bobbin or the inside of the cover may be necessary on some pickups. These results also seem to rule out the practicality of "cover-drivers" as pickups do vary. Such a scheme has many advantages, but is possible to an adventurous DIY'er and addresses point 5 as well. It would be tricky, but not impossible, to accurately thin down the inside of a plastic pickup cover using the edges of the cover as a guide for a dremmel type router...with a lot of care. This thin top could then be reinforced by gluing the solid 1mm coil onto the underside. The outcome (which I intend to do for myself) will be a cover that only "sinks" the poles a small amount <0.5mm and put less cover between the device and the strings, adding to efficiency. My staggered pole pickup has a b string pole level, and this does "sink" but this is the only one that is below the top of the cover with the coil fitted. Making the coil thinner, is not going to happen! Thanks for the tip on the glasscote or whatever it is, will check it out. Went to a paint shop in search of information and viable products. Unfortunately, the product I bought proved not to be suitable. It did demonstrate what could be however.I don't think anything that is not "two pack" (chemically hardens) will be suitable. I have a few other ideas that I plan to test out today, there are also other 2-pac paints and stuff that are pretty tough, but these things are expensive. Will construct another jig with dimension modifications and try a few more experiments. One idea I had was to try and work out a way that the wire passes through an epoxy "bath" immediately before entering the jig, coating the wires as it is threaded into the bobbin...this would be neater and apply a more consistent coverage and require less excess to be squeezed out, making it easier to get an appropriate dimensioned coil from the outset...also speeding up the process. Will also aim for a 7 ohm coil rather than 8 as this will make the alterations to the dimensions a little easier to achieve. So..this next jig should have a core of about 57mm (57.08?) and accommodate 5mm poles and if possible wind it to a total width of 14mm and a bobbin/total coil length, of some 66mm...hmmm, it may well be doable! pete Link to comment Share on other sites More sharing options...
curtisa Posted September 9, 2007 Report Share Posted September 9, 2007 I don't think there can be "zilch" space between the cover and the bobbin. There must be some or it would be extordianrily tight. The driver wires are 0.2mm so they could be squeezed with my proposed method of security to about 0.6mm. If there really isn't that much room, then a small nick in the bobbin or the inside of the cover may be necessary on some pickups. OK, not "zilch" as such, but the pickup cover on the Pac112 is a pretty snug fit from side-to-side, and you'll have to run the 0.2mm winding wire down the inside of the cover to some secure point so that the wires aren't easily damaged either during manufacture, installation, and everyday usage. In my situation you couldn't make a termination board on the driver itself and then run thicker wires down the inside of the cover, it'd bulge-out the side of the cover. Making the coil thinner, is not going to happen! I guess the guarantee that has to be made here is that the coil really con be no bigger than 1mm thick under any circumstances! And for Col: Using that inductance calculation page, with my driver I calculate 1.97mH inductance at 2KHz. I also tried it at 440Hz (was that the frequency you used? I can't remember...) and got the same result. Pretty small inductance, and not surprisingly either - A core consisting of a solid steel blade will have a pretty low permeability to start with, and 100 turns of 0.2mm wire will give a pretty low inductance. My driver also has a DC resistance of 13.8ohms, which is almost double what the apparent ideal is. Just for shits 'n giggles, I measured the inductance on the single coil pickup winding that the driver is attached to - the DC resistance is 4Kohms and I calculated an inductance of over 3 Henrys. Given my original results on the driver that must be pretty close, as the pickup winding will have thousands of turns (as opposed to my 100 turns in the driver), and inductance goes up exponentially in relation to number of turns. All results measured with a Fluke true-RMS multimeter. I've been out of the loop for too long - do we want a high inductance at a given resistance or a low one? Link to comment Share on other sites More sharing options...
psw Posted September 9, 2007 Author Report Share Posted September 9, 2007 OK, not "zilch" as such, but the pickup cover on the Pac112 is a pretty snug fit from side-to-side, and you'll have to run the 0.2mm winding wire down the inside of the cover to some secure point so that the wires aren't easily damaged either during manufacture, installation, and everyday usage. In my situation you couldn't make a termination board on the driver itself and then run thicker wires down the inside of the cover, it'd bulge-out the side of the cover. Ahh, no...you cant have thicker wires anywhere inside the cover. My intension is to permanently attach the wires and shrink wrap them with enough length to escape the cover, either from the ends or sides. These wires are reasonably strong, but won't take repetitive bending (which wont happen in use). I have a plan to reinforce them with flexible material and prevent excess bending in any one point. We'll see. I guess the guarantee that has to be made here is that the coil really can be no bigger than 1mm thick under any circumstances! That is the plan. I do know that it will fit, I have wound about a dozen now but it is very precise and the coil needs to be super tight (not stretched, but compressed)! The extreme pressure the coil exerts on the framework in winding and in pressing to this size has really tested these jigs I have been making. I have just spent another day making version number 6. Each one gets a little better and I am mindful of the potential of having to make multiple jigs if slow reacting "glue" is used. Jig 6 is made of solid aluminium and is super accurate with a core length now of 57mm. Making it so solid also allows some heat to be used. Any thoughts on heating the coil up with a current...does it have to be AC? Such a scheme may make a slower reacting epoxy set faster when it is ready...any thoughts? Making these things so accurately and with the experiences of the others and using them, I think I have something that will wind a lot faster. You can imagine that winding a coil so thin is difficult to keep the wire "in the groove". The winder has guides to aid in this, but it is a danger. The coil also needs to be pressed, not only at the end, but several times through the process. Several of the earlier plastic jigs (though quite thick) have started to fail or flex. You really could not produce these things by hand or without this kind of equipment. I will aim for a 7 ohm coil and I think I will eventually have the method down. Thanks for the Glasscote tip. I have looked it up and it is made in Melbourne and certainly looks like it could do the job. It does take a while to set completely. They seem to have a good advice line, so will be contacting them for details before I commit any more cash to it. I hope it all turns out OK. I intend to modify a new Fender Noiseless pickup with one so it will be interesting to see how this all fits. This is really interesting work that Col is doing here, thanks for weighing in on the discussion. I am committed to this particular course, but I think the general idea that both you, Curtis, and I have succeeded with, is a valid one no matter the theoretical "best" driver. We might be surprised to find a real reason for it working as well as it does. Keep it up... Pete Link to comment Share on other sites More sharing options...
curtisa Posted September 9, 2007 Report Share Posted September 9, 2007 I have just spent another day making version number 6. Each one gets a little better and I am mindful of the potential of having to make multiple jigs if slow reacting "glue" is used. Jig 6 is made of solid aluminium and is super accurate with a core length now of 57mm. Making it so solid also allows some heat to be used. Any thoughts on heating the coil up with a current...does it have to be AC? Such a scheme may make a slower reacting epoxy set faster when it is ready...any thoughts? AC would be my guess. I believe polarising any transformer circuit with DC is a bit of a no-no unless it's designed to take it. Pickups may not like it if you burn a DC current through the winding for a long period of time. But I'm only really going on my experience with input and output transformers in pro audio gear, and industrial current transformers in my day job. Pickups may not matter either way. Still, AC current would be my preference, just to be certain. However, you'd need to watch how much current you're passing through the coil to make sure you're not over-heating the coil. The last thing you want to do is build a perfect driver coil, burn it in, and in doing so soften the insulation on some of the turns and cause a shorted winding, rendering the driver useless. If you were super careful (and I mean REALLY super careful - epoxy is highly flammable) you could bake the assembled coil in an oven on a low temperature, say less than 100 degrees C. I baked my driver that way, but I was using PVA as the curing agent. Link to comment Share on other sites More sharing options...
col Posted September 9, 2007 Report Share Posted September 9, 2007 (edited) Using that inductance calculation page, with my driver I calculate 1.97mH inductance at 2KHz. I also tried it at 440Hz (was that the frequency you used? I can't remember...) and got the same result. Pretty small inductance, and not surprisingly either - A core consisting of a solid steel blade will have a pretty low permeability to start with, and 100 turns of 0.2mm wire will give a pretty low inductance. My initial measurements give 7.02mH. so quite a different reading from you, I do intend to repeat the experiment with different values to double check. My driver also has a DC resistance of 13.8ohms, which is almost double what the apparent ideal is. The LM386 is rated for 8 - 16ohm, so I suppose 13.8 won't cause too many issues.... I think whats more important, at least fo r the high note response is the reactance, and to what extent it is corrected by the zobel.. All results measured with a Fluke true-RMS multimeter. heh, mine is a Fluke, but it's an old 75 seriesII... not sure about 'true RMS' but it does provide good accuracy figures up to 500Hz. I've been out of the loop for too long - do we want a high inductance at a given resistance or a low one? Well, this is one of the many questions I am trying to answer There are only really two things we want: to maximize magnetic pull on the strings for a given current*voltage to minimize flux radiation everywhere other than at the strings... At this stage, I'm trying to get a better understanding of the relationship between: (Edit: i can follow the concepts and understand the equations.... what I mean is that I don't have a complete mental picture yet - not got near to the stage where it's intuitive yet) Magneto Motive Force Magnetic field strength Flux density Flux linkage Frequency Particularly the first two and how they relate to inductance Also want to get a better idea of how the extruded form of the coil effects the functionality and what impact eddy currents have - with a very thin core, not much I guess, but a thicker core can give potentially better performance, but more eddy current losses... it would be good to get some idea of what percentage though - 1 - 3% I could live with, 15% upwards definitely not... Assuming that it is important (!) We also need to work out a simple sequence of steps using the minimum of equipment that enables people to calculate the correct Zobel network for their driver - shouldn't be too difficult cheers Col Edited September 9, 2007 by col Link to comment Share on other sites More sharing options...
mrjstudios Posted September 9, 2007 Report Share Posted September 9, 2007 Thanks MJS... this illustrates the potential problem here. Could you also measure the length of the bobbin in total there MJ, if this were to be a common variation, then perhaps a special jig may need to be made for a special order to accommodate them. Otherwise, I will have to be very specific about what this thing will fit...this is not something you can simply sand the edges off to make it fit! 65.57mm So...I am not sure what you used to arrive at those particular figures (they are extraordinarily accurate for metric, did you convert them from imperial perhaps?) but on the face of it, these coils wouldn't fit on this pickup Not Good! The other thing is that it may be an old tele neck pickup, these are a fair bit narrower...I would need to make a different coil for these at some point. The thing is, I can not lengthen the length of they won't fit a standard pickup, nor narrow it down too much (though I have been trying) and the core is set for this particular jig. I know 1mm is very small, but considering this is the total height of the coil itself, these things are huge! Well, lucky for you all outside the USA, my digital micrometer has a mm setting accurate to the 000.00 decimal so my measurements are VERY accurate. Hmm, well I know mine are NOT tele pickups, and all 3 of them on that same guitar were exactly the same size. -MRJ Link to comment Share on other sites More sharing options...
thegarehanman Posted September 9, 2007 Report Share Posted September 9, 2007 Pete, Have you considered using very thin superglue to pot your coils? If you're using polypropylene for the temporary bobbin holder, then you wouldn't need to worry about the CA sticking the bobbin to the jig. I think if you saturated the coil well enough, you'd get 100% penetration of the coil. It would also cure rock hard. Just make sure if you try this that you do it in a very well ventilated area (outside) with a decent organic vapor mask. peace, russ Link to comment Share on other sites More sharing options...
onelastgoodbye Posted September 9, 2007 Report Share Posted September 9, 2007 On my absolute crap single coils I get: 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. Link to comment Share on other sites More sharing options...
psw Posted September 9, 2007 Author Report Share Posted September 9, 2007 Thanks for the responses guys. Good to see you back Tim, I was going to quiz you about making these coils and my jig designs as this seems to be your forte. Have you considered using very thin superglue to pot your coils? I really don't like using superglue, I know it is thin and all but very hazardous and short use time. It isn't all that structurally strong...sure it sticks, but it also breaks. I am not sure if it is enough to support the coils. just a thought...maybe a vacuum pump can get the last bit of air Russ suggested a similar thing in an email. I don't see how you could effectively pot these thing post-winding, it really needs to be potted while winding. The problem really isn't air, or at least gaps in the winding. To make coils this small the windings need to be heavily compressed (that's why my jigs have had to get progressively stronger). Forcing the sides, can push out the ends for instance and cause air gaps there potentially, but this is a very tight squeeze...it will only work if there are no gaps and everything goes on tight! There is a type of wire that has an enamel coat and an epoxy coat over it. The epoxy softens with heat, so by winding the coil, a current can be applied to heat it (after winding) enough to soften the windings and so stick the thing together. How structurally strong this would be (able to support itself without a bobbin) I am not sure, it really is a potting method...but exploring such a method is impractical because of the cost of the wire and the necessity of buying it in bulk. I wound a coil last night with another epoxy with a one hour use time...unfortunately, I forgot to use a release agent and it completely glued the jig and coil together. Using heat and then sticking the thing in the freezer allowed me to destroy the coil and rescue the jig...a lesson learned. Some of these test coils have shown a few problems with the way I make the inner core, so this may require a little more work. It may be advisable to fit plastic face plates to this jig to ensure separation of the coil as a lot of these epoxies have no trouble sticking to aluminium....hmmmm On my absolute crap single coils I get: 1: 4.5 mm 2: 55 mm 3: 14 mm 4: a whooping 1 mm 5: polepieces stick out 5mm above the bobbin (!) so would be no problem. OK...so that looks like my coil would fit. If wider, the cover should still go on with that much gap. Well, lucky for you all outside the USA, my digital micrometer has a mm setting accurate to the 000.00 decimal wink.gif so my measurements are VERY accurate. Ok...I probably need one of those over my guess-0-meter!!! They are unusual pickups, but there are so many variations. 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. So for these, I am really pushing the design. Oddly enough, I am having more trouble now than with my earlier ones getting a clean accurate coil out of it. I think this new wire has slightly more insulation on it making it a little more thicker and less flexible. I wonder if there is a way to "soften" this enamel with a kind of solvent and then try and re-bake it or something after winding. I had also been thinking that perhaps some kind of synchronized tensioning to pull the wire tighter on the elongated sides. Or how about some kind of wheel that pushes (perhaps with springs) the wire in that rolls along the 1mm groove...maybe something like this could also apply the glue? One problem is that under pressure of winding and the jig itself, and further pressing of the sides and such, the "glue" has to force it's way through the windings. The viscosity of most of these things makes that very difficult and the pressing of the side technique that Tim advocated forces all the glue to try and exit through the ends which is very difficult. I will conduct a search to see if there are any left over stuff in the garage. A substance as thin as paint but dries hard and quick by chemical reaction (the coil is effectively air tight in winding so solvent based products have proven very unsuccessful) is what is required. I am not sure if this "glasscoat" fits that bill or if some kind of epoxy paint may work. I may need to do another dry run or so with this new wire to see how much I can get on there and neatly if I take the glue out of the equation. Basically, I only have room for glue between the round profiles of the wires themselves. There is no leeway for air gaps and even overlapping wires from uneven winds are a problem. Any loose wires, even if subsequently pressed in, will loose tension when pushed in and create little gaps. I know it can be done, I have made a few successfully to the right specs, but not in a production mode or of a very professional "finish". A little disheartening and a lot frustrating, but it just strengthens my resolve to work out how it can be done....grrrr I am open to suggestions or ideas on the making of these things, jig ideas, glues and such...may have to think outside the box a little. It is entirely possible that I will have to make yet another jig despite the accuracy and build quality of this one, if ideas come that lend themselves to better building techniques. Link to comment Share on other sites More sharing options...
col Posted September 10, 2007 Report Share Posted September 10, 2007 (edited) hmm... another thought to broadcast to the group.... I've been doing some calculations to try and get close to reasonable specs for a coil using 0.375 wire. I don't want to make a coil only to discover that its miles away from the desired impedance..... So I have most of it worked out, apart from one thing... any thought anybody ? -------------------- one of the variables that the inductance and (therefor impedance) is dependant on is a thing called 'core factor' this is a combination of the dimentions of the core and the permiability of the core. imagine a coil that is 2mm thick and where the core is level with the top and bottom of the coil... so the core is 2mm high. Now imagine that there is another identical coil 2mm thick, same wire, the core length and thickness stay the same, the only difference is that the core sticks out 2mm at the top and 2mm at the bottom... so there is exactly the same core mass actually 'inside' the coil, but in the second coil, the overall core mass is 3 times that of the first coil.... So what if anything does this do to the inductance and therefor impedance of the coil ? I want to know this because if I make a driver using much thicker wire, the coil will be bigger and so the mass of core 'inside' the core will be greater - this in itself would make a difference. However if the coil just 'used' a bit more of the same 6mm high core, would that help to balance things..... It comes back down to the odd shape and specs for our projects 'wound component'. Of course, the 'easy' thing to do would be to make a stack of coils each with one variable changed from the last, but I would really like to avoid having to do that One option is to make up a bunch of cores, and then wind some coils, but without potting them. Then just use them for analysis.... hmmm. I might get around to that I suppose. Anyway, based on my test data from my old coil, A new coil using 211 turns of 0.375 wire around the same core would have a DC resistance of 4ohm, and an reactance of 7ohm @ 82Hz so the initial impedance would be ~8 ohm ! Of course, the impedance would rise more quickly with frequency. However, it may be that the shape of the frequency response curve would be similar, and that it could be balanced with a suitable zobel network. This new coil would certainly be bulkier than the old one, but not way too much - something like 4 - 5 mm deep (for a square bundle cross section) So I think it will be worth some experimentation cheers Col Edited September 10, 2007 by col Link to comment Share on other sites More sharing options...
col Posted September 11, 2007 Report Share Posted September 11, 2007 hehe I worked out how to do coils in femm. Some pretty pictures coming soon... here's a taster - it's a single coil with a smallish alnico bar at the base and a partial 'cage' that seems to help focus the flux into that nice 'flame' shape. As far as the numbers in femm show, this seems to be a good thing, the area where the strings would be has a higher 'tesla' value than the same model minus the extra 'cage'. I'll try to get around to modelling some more accurate versions of the various drivers soon. (btw, I tweaked the colour mappings so there is more detail in the flux surrounding the driver instead of it all being in the core - thats why it looks cooler than usual ) cheers Col Link to comment Share on other sites More sharing options...
mrjstudios Posted September 11, 2007 Report Share Posted September 11, 2007 Nice work Col. That "cage" design is what I was trying to do with my shielding project a ways back in the thread. I figured that there was a way to use magnetic (Steel) 'fins' to channel the magnetic field, but I just couldn't figure out how to apply my theory. You seem to have figured it all out perfectly! I did try some channeling on my single coil design and got some improvement in response, etc., but I had no way to really tell what it was that I was doing that was actually helping. Now I think I will have to get out the angle iron that I was using before and cut some new fins to stick on my dummy pickup / PSW kit driver "unit"...... -MRJ Link to comment Share on other sites More sharing options...
psw Posted September 11, 2007 Author Report Share Posted September 11, 2007 It comes back down to the odd shape and specs for our projects 'wound component'. Of course, the 'easy' thing to do would be to make a stack of coils each with one variable changed from the last, but I would really like to avoid having to do that Not so easy...and it wouldn't necessarily provide the "proof" you are seeking. I am approaching it from the angle of usability of course. However, I still am "attached" to the "thin coil". From experimentation I am sure there are advantages (perhaps the small core mass, or something) for EMI reduction and efficiency that has not been fully explained. Even if it isn't the "ideal" it appears to be a fairly good compromise. It is interesting to see the way the two companies, and all others who have approached this, have built drivers similar in dimension and shape to conventional pickups. This is one obvious area in which my proposed "thin coils" vastly differ from other propositions and products...why they work as well as they do is still elusive. Obviously, there are so many factors involved in these things. Some of them very hard to measure with any accuracy. From the stand point of application, such a proposal does commit you to a dedicated driver of dimensions approaching that of a pickup, probably necessitating the loss of this to the instrument. I do think there could be some interesting ramifications for this for dual coil drivers, something I will want to develop once this thing is underway. Keep up the good work Col... ===== As for me, would you believe I had to completely redesign the jig...now I am on version 7. Back to using plastic reinforced with aluminium and a completely revised core for better shape and separation qualities. This one is very strong but easy and cheap to duplicate. This is important as it is now apparent that fast acting epoxy is not going to work. I have done a few tests with slower acting epoxy with a use time of 1 hour and a drying time of about 24 hours. This gives me adequate time to "work" the coil to the required shape and dimensions, and takes a lot of the pressure off making the things. It would also allow time to make a few with multiple pressing jigs from a single batch of epoxy. SO. I just "cracked open" last nights midnight effort and it came out pretty good. Because of the increased work time, I was able to get the core up to a true resistance measure of 8 ohms (actually went to far and had to unwind a bit as the coil calculator program over estimated turns required [ it also predicted that there was no way of fitting that many turns into this space ] ). The coil bulged a little on one side, making it slightly over-sized and therefore not fit the pickup, but this was during curing and measures can be taken to prevent that. I was also able to add material to the inside and outside of the coil which will aid in protecting these vulnerable edges both in use and during the manufacturing process. This is also used to strengthen and protect the driver wires exiting the coil to prevent excess bending or pinching as it is squeezed between the bobbin and the cover. This will enable a more professional look too, so I am pleased that this feature could be incorporated into this one. The core is approaching 58mm in length so should accommodate varying pickup width, and the width was 16mm with the bulge. Easily get that to 15mm but will aim for 14mm. The total length is exactly 66mm, which is the exact length of my, and typical fender sized bobbins. These dimensions, created with this jig, should therefore create a coil that would fit all the variations of single coil pickup so far reported. This seems to indicate that, with a little practice and refinement, that jig version 7 will make the required coil and to the standard required, if used with a suitable epoxy. It has shown though that it takes a little more work, and the epoxy does add quite a bit of added expense to building the things. It does seem to indicate that I may need half a dozen of these things to practically do a cost effective "run" and that would be as much as is practical in a day. Anyway, this is good news. I was getting very frustrated with this jig making, failed coil regime the last couple of weeks. Will make a few of these jigs and then put all the jig making and aluminium filings away for a bit. ===== This new jig is not quite as adaptable as the previous models, however it and some of these others, could serve to experiment with slightly deeper coils (up to about 4mm) as Col is suggesting. It could also serve as a jig for twin-coil HB version of this pickup adaption system. It could be that for this, Col could find a better wire gauge...perhaps a slightly thicker wire with two coils of 4 ohms would be advantageous (maybe a thinner wire of 2x16 ohms would be worth trying too). I am likely to stick to my present wire, as I have got so much of it, for the time being. My apologies if it seems to be taking too long, but I am putting in an incredible amount of work, and learning valuable lessons in manufacturing and coil building techniques that will eventually feed back to allow further experimentation and more consistency for others. Learning from mistakes is all very fine, but it is easier to learn from other peoples mistakes than it is to learn from your own! pete Link to comment Share on other sites More sharing options...
psw Posted September 11, 2007 Author Report Share Posted September 11, 2007 hehe I worked out how to do coils in femm. Some pretty pictures coming soon... here's a taster - it's a single coil with a smallish alnico bar at the base and a partial 'cage' that seems to help focus the flux into that nice 'flame' shape. As far as the numbers in femm show, this seems to be a good thing, the area where the strings would be has a higher 'tesla' value than the same model minus the extra 'cage'. I'll try to get around to modelling some more accurate versions of the various drivers soon. (btw, I tweaked the colour mappings so there is more detail in the flux surrounding the driver instead of it all being in the core - thats why it looks cooler than usual ) cheers Col Opps, posted during the posting of this so missed it and turned the page, so here it is again. That is such a pretty FEMM (btw, I tweaked the colour mappings so there is more detail in the flux surrounding the driver instead of it all being in the core - thats why it looks cooler than usual ) That is similar to the way Steven Kersting modeled his pickups, which is how I got onto the whole FEMM thing. I think I still have some of the FEMM models he created on an old computer somewhere, as he kindly sent the data I think. Now I think I will have to get out the angle iron that I was using before and cut some new fins to stick on my dummy pickup / PSW kit driver "unit"...... Yes...exactly. You can see the potential now for a consistent well made coil. With access to identical coils, it will be possible to easily do such experimentation avoiding all the costs and inconsistencies (not to mention difficulties) of the coil winding part of it, and concentrate on the development side of such ideas. It will allow direct comparisons between ideas that can be repeated by others in all areas of the project. Unfortunately, achieving this goal is taking an inordinate amount of development time and looks like will take more work and expense to create than originally anticipated. Still, it appears to be do-able, and that is the first stage. good stuff... pete I knew there was more to FEMM...looking forward to seeing what you can do with it! Link to comment Share on other sites More sharing options...
col Posted September 11, 2007 Report Share Posted September 11, 2007 Nice work Col. That "cage" design is what I was trying to do with my shielding project a ways back in the thread. I figured that there was a way to use magnetic (Steel) 'fins' to channel the magnetic field, but I just couldn't figure out how to apply my theory. You seem to have figured it all out perfectly! Thanks. I've played about with different 'cage' ideas before using femm, and also making a housing/cage for my driver.. the real one didn't help at all, and the femm models were just using the permanent magnets so they didn't give any really useful information. There are still potential issues of course. Femm is only 2D + a bit, add to that the fact that the materials used - steel type, magnet type are just guesses as is the current phase etc - and you have a useful guide but nothing more ! This line of investigation has (of course) brought up yet another area for consideration and brainstorming: Permanent Magnet vs electromagnet.... As transient explained a few pages back, the permanent magnet provides a constant pull that the electro magnet then helps or hinders depending on the polarity of the signal at time... So without the permanent magnet the system either doesn't work or doesn't work well. This balance between the electromagnet and the permanent magnet is I think critical to achieving good performance and good efficiency..... If the permanent magnet is not strong enough, the electromagnet can still create string pull instead of 'release' on both the negative and positive parts of the signal.... If the permanent magnet is too strong, it sucks much of the flux away from the strings ! (at least according to femm it does !) I think it will be possible to use femm to look at the interaction between the coil and the permanent magnet, and find a 'sweet spot'. However, moving that knowledge to a practical model will require either magnets with known data, or some way of measuring the specifications of permanent magnets.... The alternative (it might work) would be to develop a test circuit that feeds a controlable dc current into the a known coil and you twiddle it until the coil can 'drop' the magnet - will that mean that it exactly oposes the force of the magnet ? - if so we should be able to get roughly the correct balance.... Unfortunately this would be another situation - like the zobel - where folks would have to go through a process in order to 'design' a suitable coil for the magnet(s) they have.... at the very least we could try and get some 'window' outside of which we can say that a magnet is not suitable.... we would need to specify some kind of test rig that would be easy to construct... .......... I had to leave my pc for a while before I hit send... since then I did a little scripting in femm. Heres a little animation with the current cycling through 360º... I think the double peaking effect is caused by the coil being a bit too strong for the permanent magnet... although this was just a first test run for the femm lua script, so it really just for eye-candy purposes (At least part of the reason for the coil being too strong is that I bumped up the current x3 to see the flux more easily - haven't worked out how to tweak the colours from the script yet.) cheers Col Link to comment Share on other sites More sharing options...
col Posted September 11, 2007 Report Share Posted September 11, 2007 Here's an animation of the same basic design but this time the colours are tweaked and the coil and permanent magnet have been balanced. You can see how with the two balanced, the flux around the bottom part where the permanent magnet is is about the same as the flux at the top when the current is at maximum, this makes sense when you think about it (I think?). quick description of the diagram the central part at the top is the core - 2mm thick steel either side of the base of that are the coils, and directly below it is the permanent magnet the L shaped sections at either side of the magnet are 2mm thick steel 'cage' parts that help(?) to guide the flux back around, hopefully reducing EMI and increasing performance. Col Link to comment Share on other sites More sharing options...
col Posted September 11, 2007 Report Share Posted September 11, 2007 (edited) Heres one for comparison, the only thing changed from the last is removal of the L 'cage' pieces the one with the cage seems a little better - not sure if its enough to be worth the trouble - its hard to say from such a roughly thrown together simulation. Edit: but damn, it's great to be able to get a better idea of the interaction between the magnet and the coil Col Edited September 11, 2007 by col Link to comment Share on other sites More sharing options...
psw Posted September 11, 2007 Author Report Share Posted September 11, 2007 Thant's hypnotizing yet oh so pretty... I wonder, have you tried some of our more conventional designs by comparison....fishing for a very thin single coil on top of alnico poles of course. Certainly it looks promising, but you and I and several others have actually built such "cages" and not found appreciable effects (My recent mid-driver design had inner and outer blades and the Hex designs set in pure iron impregnated epoxy, if you recall). Perhaps the EMI flux travels down the metal string into the source pickup and is unavoidable...so many things to consider! Down here, I have been doing a little work making duplicate jigs after the success of the last coil to make "production" a possibility with the new slow curing and pressing techniques developed the other day. This will be the last day (probably) for a bit that I will be able to spend so much time on the project. It is a relief that I seem to have finally found a way for this to be practical. Proving that it can be done, now requires practice to ensure it can be done in a short amount of time and cost effectively. The amount of "hand making" of these specialized coils adds greatly to their "cost". It is vital if the venture is to be successful, that this is reduced. The intention is that these coils be produced in small runs and while setting the circuits can be made. Taking a whole day for each one is obviously not going to work out or push the cost for the amount of effort way up. Col...are those animations made from within FEMM, or do you use separate immages to piece together an animation outside the program? pete Link to comment Share on other sites More sharing options...
col Posted September 11, 2007 Report Share Posted September 11, 2007 (edited) continuation of cols posting flood.... Think some more about what happens when the coil is too strong for the magnet - both the positive half and the negative half (at least the peaks) will attract the strings. Some related thoughts come to mind... #1 if the above is happening, the sustainer will basically be trying to drive the strings at twice the input frequency resulting in another cause of the well known harmonic effect ! #2 due to reactance increasing with frequency, the coil will effectively be at its most powerful when playing low notes. If when playing these low notes, the coil is overpowering the magnet, there will always be some harmonic bloom in this frequency range (sound familiar Pete ?, it is at least another plausible explanation). #3 The 'traditional' way of getting harmonic effect by inverting the signal works, but it has a nasty choking effect as well because it tries to cancel the fundamental note, I wonder if it would be possible somehow to take advantage of this rectifying type of effect ? Probably rather than driving the coil harder, it would make sense to actually rectify the signal in the circuit instead of inverting it ! hmmm (Edit: of course, my AGC circuit already has a precision full wave rectifier, it will just need a buffer, a cap to decouple the DC offset and may some other little do-dahs here or there...) @ transient, Your explanation is partly correct, but the permanent magnet also serves as a flux amplifier - without it we would need a much more powerful coil to get the strings moving. Otherwise we could just add a DC offset to the driver signal and do away with the magnet altogether... (might be an option in a transformer powered system) cheers Col Edited September 11, 2007 by col Link to comment Share on other sites More sharing options...
col Posted September 11, 2007 Report Share Posted September 11, 2007 Thant's hypnotizing yet oh so pretty... I wonder, have you tried some of our more conventional designs by comparison....fishing for a very thin single coil on top of alnico poles of course. I'm not sure how to accurately model the alnico poles - a femm model is just a 2D model extruded into the 3rd dimention, but if we just pretend a solid bar is the poles, then it shouldn't be too tricky - you give me the specs and I'll make a model. Certainly it looks promising, but you and I and several others have actually built such "cages" and not found appreciable effects (My recent mid-driver design had inner and outer blades and the Hex designs set in pure iron impregnated epoxy, if you recall). Perhaps the EMI flux travels down the metal string into the source pickup and is unavoidable...so many things to consider! These new 'cages' are not an attempt at 'shielding' what they really are is the continuatino of the inductor core !. Thats why they are the same thickness and same material- they are there to increase the power of the magnet and to focus the flux on the strings. ( It is important that they are as close to the coils as possible. ) If we do choose to go this route, I imagine that these parts could be used to aid the manufacturing process - maybe they can be used to compress the coil - and of course, they needn't be removed... Down here, I have been doing a little work making duplicate jigs after the success of the last coil to make "production" a possibility with the new slow curing and pressing techniques developed the other day. This will be the last day (probably) for a bit that I will be able to spend so much time on the project. It is a relief that I seem to have finally found a way for this to be practical. Proving that it can be done, now requires practice to ensure it can be done in a short amount of time and cost effectively. The amount of "hand making" of these specialized coils adds greatly to their "cost". It is vital if the venture is to be successful, that this is reduced. The intention is that these coils be produced in small runs and while setting the circuits can be made. Taking a whole day for each one is obviously not going to work out or push the cost for the amount of effort way up. Col...are those animations made from within FEMM, or do you use separate immages to piece together an animation outside the program? I wrote a lua script for femm that generates a series of bitmaps with consecutive numbered filenames.... I then use GIMP to grab 'em all and generate an animated GIF.... now that I have sussed out the most efficient steps in GIMP, the final part is very quick - a couple of minutes at most. So I can make a change to the base femm model, then if it look promising, I can have an animation up and going in a couple of minutes.. Any big changes to the model might mean the script needs a little tweak here or there, but overall its way quicker than trying to do it all 'by hand' cheers Col Link to comment Share on other sites More sharing options...
psw Posted September 11, 2007 Author Report Share Posted September 11, 2007 Wow...we all seem to be here at the same time at the mo... I'm not sure how to accurately model the alnico poles - a femm model is just a 2D model extruded into the 3rd dimension, but if we just pretend a solid bar is the poles, then it shouldn't be too tricky - you give me the specs and I'll make a model. Ok...a coil at the top of a pole. say 5mmx16mm of 1.5mm thick x 14mm total width. (0.2mm wire at 8 ohms). On the cage thing, I did do some coils that had a "cage" made from a strip of steel staples. These went around the driver and were stuck to the bottom of the magnet (so presumably of opposite polarity). I didn't find any improvement or detriment to the operation of the coil and so abandoned these attempts after a while. I did, and still contemplate, the idea of a iron/epoxy band around the outside of my new designs. I even made a trial run and it is possible, but I have doubts that it is worth the effort. These are some of the "further developments" I have built into the design of the current device family and the equipment that makes it. Great stuff, I will have to do a little animation and stuff when I do the CD of instructions for this thing, similar to the way Tim did those excellent jig illustrations. Another learning curve to climb! pete Link to comment Share on other sites More sharing options...
mrjstudios Posted September 11, 2007 Report Share Posted September 11, 2007 Score #2 For Col! Wow. Hey, just an idea.... As I understand it, you will be modeling PSW's driver in FEMM right? If you get that model/scripting working, try making another model file with a solid 'blade' core in stead of 6 poles, but with otherwise the exact same specs. I will beta test PSW's system in its original "stock" state -- but I am also going to try a blade that replaces the center poles in a standard SC pickup. If it works better with the blade, I'll show how to 'mod' PSW's kit, or in other words, the standard SC pickup, or maybe even he will want to offer a 'conversion-to-blade' kit if it works really well. Who knows. But it would be interesting to see what FEMM says about it... -MRJ Link to comment Share on other sites More sharing options...
Recommended Posts