Sustainer Problems!

col · April 13, 2010

then all i have to do is just divide it by 4,the overall turns for 0.1 mm wires to achieve 8ohms is just 31,25 turns.........is this correct?
and also i have tried that pickup calculator,it shows 34 turns...not so different

No, its wrong.

You are correct that roughly 34 turns of 0.1 will give you an 8ohm DC resistance. The problem is that the fewer turns you have, the weaker the magnetic force of the driver will be, and unfortunately, this is not linear with number of turns.

If 300 turns gives an inductance of 2mH (miliHenries) then 34 turns will give you (2 /sqrt((300^2)) * (34^2)) which is 0.026mH - 100 times weaker !!!

if you are using an 8ohm single coil drive with an LM386, then you want roughly 1 - 1.2 mH. any less means not enough drive. any more means fizz and distortion will become more likely, and high frequency response will get worse.

3.,the starngest thing about my driver is,no matter how many turns of wires and the sizes of it (i have treied 100,200,300,150,64 turns of 0,1 mm and a thicker wire,which i dunno the dia. but still on the same numbers of turns mentioned above) the driver itself generates sound! hum,squeal,and even the sound of the string itself....meaning that the driver acts as speaker even without a CONE.....is this normal?...

What you discribe suggests that your coil is not 'potted' properly, so some of the windings are vibrating. This is BAD. it is important to make sure that all the windings in the coil are solidly held together with some sort of glue, and that the bobbins and magnet are also held firm.

Otherwise you will get vibrations that can cause various problems, and will reduce the efficiency of the driver.

kind of sharing currents or something?sustainer driver actually only need one coil isnt it?

What FFs suggestion would do is to use a doubled strand of 0.1 wire as though it were a single strand of 0.15 wire. To get it to behave like 0.2 wire, it would need to be a 'quad' stranded approach.

There is another way you could try to use 0.1 wire. You could try making 4 separate coils each of 32ohms (~120 turns). Then wire them in parallel. This is nothing like optimal, but you will be left with an 8ohm dc resistance, and the inductance will be maybe 0.3mH. Not ideal, but a lot better than what you would get with a single coil driver. You will also get a much better performance with this than if you used a single coil 0.3mH driver! The min difficulty (apart from having to construct four good coils) is to get them all wired up correctly so that they are in parallel and the magnetic fields are all in the correct orientation.

My advice would be to get some 0.2 or 0.23 wire and make a single coil with 120 - 150 turns. When you have that working, try other stuff.

cheers

Col

psw · April 14, 2010

1# don't worry about turns, worry about getting close to 8 ohms....with the correct wire. Trying to make things equivalent to the right wire instead of getting the right wire, seems absolutely pointless and fraught with problems.

2# the "potting", the glue...is fundamental to the thing working...that's why the instructions say glue and not playdoh!!! sheeeesh! Yes, you have so many loose windings you are hearing the coil vibrate!!!! If the energy is going into this how do you expect there to be enough left over to drive the strings or for these weird signals not to get into the sensitive pickup that detects magnetic signals???

3# the power amp like the LM386 is not built to take the high impedance pickups, so some kind of buffer is important. The fetzer was never the best choice, but it will work. Other options which all have their pluses and minuses are things like the tillman or an opamp or just any common stomp box that usually has a buffer stage in it.

The fact is that you need to have this stage for this circuit and this project. If you have built a buffer and it isn't working with a proper LM386 circuit (tested with a speaker as your driver is obviously useless) then you have done something radically wrong with the circuitry too! This will not work, it is not just the tone suck on the guitar, but that this effects the amount of power you are getting out of the system too!

...

There have been posts that suggest almost any wire guage will work and such, this may be true within some kind of system, but for my design, for it to work as expected, and in anyone else's working on this principle, there are standards that need to be adhered to if you expect it to work. So...absolutely no vibrations in the coil can be tolerated...no playdoh potting or anything like that. I seriously advise forgetting about formulas and theoretically if you run several strands to make the equivalent of...just get the right stuff. If you want to add a buffer stage or preamp, the fetzer will work for instance, not having that transistor is not an excuse to leave it out...there are equivalents that will work just as well that can be adapted. Otherwise, there are all kinds of kits and circuits about that describe simple buffers or preamps that will work.

Before anyone starts yelling at me in bold type about it, I am not saying that my wire gauge or anything is the only one that will work, but it has been proven over and over to have struck the right balance to get the job done effectively and efficiently in this kind of simple design. At no stage did playdoh enter the equation...where did that come from! Glue, there is a step by step real time pictorial that shows exactly how easy it is to do this by hand with safe wood glue and be effective. Many have suggested that this isn't good enough and to use epoxy or something, or even let toxic superglue get everywhere...but really, if we are playing with playdoh, keep safe and use something like PVA...it only confirms why I suggested this solution in the first place. Also, have on hand tape and stuff so that you can get the coil tight.

The turns thing, seems to vary a bit even with the same wire gauge, a lot depends on the size and type of core you are using for instance. I don't recall details of that either...perhaps a picture of exactly what we are dealing with would help. But there is no need for formula and suppositions or counting turns...just measure the ohms till you get to about 8 ohms and you are done...it certainly is the easiest way and most accurate approach and again, clearly showed in the tutorial...

I suppose we may as well go back to basics...is there something you don't seem to be able to follow form the driver winding tutorial I wrote and photographed? It seemed pretty clear to me, but then maybe I am wrong. I certainly didn't include any playdoh or even counting of turns.

Perhaps the best way to proceed is to give us a full picture of what you are doing, perhaps even pictures...but there just seems to be way too many shortcuts taken here and wondering why it doesn't work...it doesn't work because you can't take that many shortcuts with something as sensitive as this. The reason that these threads go around and around, and that the main thread got so big, was that people just keep taking short cuts or following assumptions or simply being far to naive about things.

Then you end up with a bunch of people chiming in suggesting there are problems with the basic principle when in fact there are none, everyone that has replicated my basic design fundamentals has gotten reasonable to excellent results. Many have gone on to create fine variations of their own. It really is not that hard, but it does mean following instructions or at least principles...non of which includes the use of playdoh to pot a coil!

sorry....had to let that out...

...

sigh...I have been able to get a tone sucking circuit like an unbuffered LM386 to drive a string...perhaps most I don't recall...but the driver is a key element and you don't have one yet that could anywhere like be capable of working. However, it is a tone sucking inefficient idea not to have this part of the circuit and you will never get the results without something of that kind in the circuitry. RoG circuits are notoriously skimpy in their components, this project requires clean headroom with all manner of precautions to prevent internal circuitry oscillation at high outputs and loads...these details are in the data sheet and advisable to add in. Many judge the LM386 running at 200x amp without any of these things included which are specified by the manufacturer...hardly fair, nor the best for a project like this that places high demands on the elements of a complete system. Remember, that when a string is vibrating it is pumping a lot of continuous input and output into these things. Fortunately the LM386, for all its downfalls, can take it, but it may well not be happy about it!

For alternatives to the fetzer, try the tillman or any other simple buffer, look for preamp kits...the champ and prechamp kits can easily be adapted though make for an unnecessarily large circuit, opamp circuits should work fine (may take more power to run perhaps).

Your biggest problem is clearly the driver...my advice still stands, use the correct formula, magnets (remember mike G's failed in part because of the bizarre magnets used) and potting to make a solid coil...measure the resistance rather than count turns as it is near impossible to calculate without an accurate bobbin and core dimensions, the stretching and wire, etc. Those turns calculators are a guide and incredibly inaccurate...ignore them.

Fortunately you now have the attention of people that are pretty straight about this stuff, Col and FF have succeeded without the nonsense that has been about recently. There has been advice that any wire will work...i'm sorry if you took that to heart when selecting the wire you bought, but this is a good example of where this kind of advice goes wrong. They are correct that you could use multiple coils and multistranded thinner wire, even I played with this stuff, but if you can't wind it with a single strand or even one coil staisfactory, imagine the problems of winding multistrand or multiple coils together...and none of it proven!

Stick to what you know will work, don't cut corners and you will succeed. The driver is and always has been the key to this thing and you are far from having anything that any circuit could save...yet...but you can do it...

col · April 14, 2010

1# don't worry about turns, worry about getting close to 8 ohms....with the correct wire. Trying to make things equivalent to the right wire instead of getting the right wire, seems absolutely pointless and fraught with problems.

The important factors are DC resistance and Inductance.

The number of turns is the most crucial factor in getting the correct inductance, followed by core specs.

The 'correct' wire gauge is just a by-product of this.

It has been demonstrated that it is not difficult to use the 'correct' wire guage and wind to 8ohm and still get a driver that doesn't work well. This isn't just down to bad construction or luck. It's also down to the fact that inductance and flux density at the strings are the important factors for drive

(DC resistance is important for optimum power amp performance).

Using the 'correct' wire will give people a better chance of succeeding, but unfortunately they can still follow your instructions and fail.

I'm not saying that folks shouldn't use the 'correct' wire, or that they will make a great sustainer if they understand about inductance. However, the more people know and understand about how the thing actually works and about what is actually important and why, then the more chance they will have to get a result, or at least work out why they didn't get a result. Spreading misinformation like "don't worry about turns" is just unhelpful. better to say something like:

"if you don't have access to an inductance meter, you should get *roughly* the correct inductance by sticking to 0.2 or 0.23 wire - be prepared to try different core materials"

cheers

Col

Edited April 14, 2010 by col

psw · April 15, 2010

Fairto a point...my point is that these calculators are not accurate and the only way to know exactly the resistance of the coil is to measure it. Inductance is of course the important thing, but that is hard to measure and not necessary. If people go only on the number of turns, other factors like the bobbin and core size, the tightness of the coil or if they stretch the wire as they turn, can all have an influence. If you use the correct specs the DC resistance will give the required turns to make the thing work.

So, the advice was meant simply to measure the coil resistance and not go through mathematical calculations or rely on calculators to predict it...or counting turns...it is just the easiest most accurate way to get the required result.

As the design has been replicated with a range of cores, from my simple thin blades to pickup poles and everything in between with success, I can reasonably assume that this is less of a factor. The wire size and coil qualities, though I have not measured them nor the ability to do so with any kind of sensible and useful data, do seem to make the biggest difference. Again, my advice was in line with that.

A known design, replicated with reasonably wide tolerances has shown to consistently work. That is not to say other ideas can't or won't work; I didn't try everything. I did try thinner and thicker wire to get to this design, this is what worked best as an average of factors. I did try the same wire with different cores, and unless ridiculously thin (like a 1mm core) this had far less of an impact.

I say 'correct' wire size because this kind of specs in this kind of coil seems to produce the kind of inductance that can work on all the strings if built properly.

By far the most failures were from failing to adhere to the basic tenants of any such device...

if there are loose windings, non potted or playdoh or such...the thing will vibrate internally and fail

if the pickups and circuit are loaded by not having a buffer stage...the thing will fail

if you use a radically different coil specs, like 0.1mm wire in the same design...the thing will fail

if you mismatch the 8ohm expected impedance a lot by extra turns...the thing will fail

We have seen this time and again, and this is exactly the kind of thing that is mis-information...that people have trouble with the design. The problems only occur when people stray from the important elements of any such device.

So, I apologize for any "attitude", but all of the things that would make this project and design fail have been committed. Mike-G used different wire and strong magnets on the ends only of a stainless steel core...the result was he had some success, and yet, some problems...but then, if he had stuck to the original design or discussed these things before writing a tutorial and fixed them, a lot of misinformation and controversy could have been avoided. The result is the myth that there is a problem getting high strings to sustain.

So, the original proven design and failures and poor circuits like the F/R solution have been attributed to me and of the project generally...and every time a failure like this occurs, it provides mis-information about the project and ammunition for those who what to suggest that there are fundamental problems with the design.

It has been a while obviously, but I am not sure that anyone that has done these things as intended has failed...certainly a lot have failed...but all I recall are through not sticking to specs, bad workmanship and following peoples own ideas about things like circuitry before getting things to work in the most basic form.

The problems presented in this thread are classic examples of why people fail and not at all atypical.

So, while the more technically minded may wish to debate inductance or even suggest you need elaborate winders and scopes to build these things, it is not so. Simply measuring the driver as you would a speaker so that it matches the circuit output is more than sufficient. Building a basic amplifier with an input that will accept a high impedance pickup, is more than sufficient. Building a driver that is potted in glue to prevent any internal vibration, be efficient at all frequencies and not spew all kinds of EMI, is more than sufficient information.

Mis-information is when someone pops up, puts up a tutorial of basically the same design, claims that any wire is fine (then reveals he used the same wire gauge as the original), claims that it is different in the use of a different amplifier chip or buffer stage instead of the F/R which I have never used (though can work) and say there are significant differences...that's mis-information. The 0.2mm wire was arrived at to produce a coil that worked, it can be explained through inductance or turns or whatever, the end result is that this spec for this design is relatively important to achieve the result desired. You don't need to worry about the theoreticals if you stick to what works...one should not at all be surprised if almost all of the important elements are neglected if the thing 'fails'.

With someone (OP) at this stage who lacks the knowledge to measure a coil even when demo-ed succinctly with photos, seems to me to be not at a stage where considering the complexities of inductance or even the mathematics of making a coil that might be the equivalent to the intended design; it is a little much to get one's head around compared to just getting a wire size that is known to work and using it as intended as I suggested.

So, my comments were meant to help...the problems now of going along some of the suggested routes will likely lead to more problems along the way and into unknown territory for which the OP is ill equipped to trouble shoot. You will note for instance, that we still have no idea what core or magnets are being used, the stability of the bobbin construction...in fact we only found late in the stage that the thing is vibrating itself silly because of playdoh potting!

Without sufficient full disclosure, it is impossible to answer these questions. One might assume that people are following the instructions and work from that basis, one can see where they failed to do so and see the problems that occur as a result...but really...yes number of turns is important, but this will work itself out if wound to the specs and the DC resistance is measure as I suggest.

Basically...I'd be simply doing a bit of reading or even asking here what are the important parts of the design and getting a handle on that without the 'theory' and 'maths' that have been brought up. Simple things, you have a circuit, was this tested with a speaker...is it working ok? Did you use the right wire that will give the right inductance, did you pot the coil while winding, did you completely isolate pickups other than the bridge...all common advice repeated ad-infinitum because of the failures. What isn't seen here on the open forum are all the many people who did follow the advice and built the thing to spec and got immediate results. May even with the limitations of the F/R, but all that is seen are the problem ones and I really can't recall failures that didn't involve variations, different specs or poor workmanship or wishful thinking.

Fresh Fizz · April 15, 2010

Therizky is telling that he is experimenting with 0.1 wire. I don't see what is wrong with col's reply.

You could try making 4 separate coils each of 32ohms (~120 turns). Then wire them in parallel. This is nothing like optimal, but you will be left with an 8ohm dc resistance, and the inductance will be maybe 0.3mH

My own proposal was a push in the right direction, but still not optimal (r², it's 4 indeed!)

and:

My advice would be to get some 0.2 or 0.23 wire and make a single coil with 120 - 150 turns.

That got me thinking, 0.3mH. By splitting up a coil into multiple coils in parallel one could reduce the self inductance. That would be ideal, because we want the self inductance to be as small as possible. We want a driver that is 'ohmic'. I only think that in our case the inductance doesn't equal 0.3mH because the 4 coils are coupled on the same core. But what if the 4 coils have their own core?

Then the next logical step would be to wind 4 coils 32 ohms/4.0 mH. Thicker wire and more turns.

Cheers

FF

col · April 15, 2010

Fairto a point...my point is that these calculators are not accurate and the only way to know exactly the resistance of the coil is to measure it. Inductance is of course the important thing, but that is hard to measure and not necessary. If people go only on the number of turns, other factors like the bobbin and core size, the tightness of the coil or if they stretch the wire as they turn, can all have an influence. If you use the correct specs the DC resistance will give the required turns to make the thing work.

Not true!

If people go only on the number of turns, then they must use exactly the same dimensions of core made of exactly the same material to get the same result as a driver they are copying.

e.g.

A 52mm 2mm thick core wound to 8ohm with wire with very thin enamel will have about 136 turns

A 56mm 4mm thick core wound to 8ohm with wire with thicker enamel will have about 122 turns

if the 136 turn is optimal - for this hypothetical example lets say that's 1mH, then the other will be underpowered by 20% at 0.8mH

Can you see that that is significant (and probably more than you would expect from the slight changes in specs.

Now that difference may be offset by the extra mass in the 4mm core, but then there would also be significantly more losses due to eddy currents in the larger core - and the core material would come into the equation.

In the end, the only way to be sure is to get the DC resistance AND the inductance right. String gauge is only one factor in achieving that goal.

(It would be better to match the inductance exactly and have the DC resistance vary somewhat - anything form 7ohm to 9 ohm should be fine)

So, the advice was meant simply to measure the coil resistance and not go through mathematical calculations or rely on calculators to predict it...or counting turns...it is just the easiest most accurate way to get the required result.

It is simple, but very inaccurate.

As the design has been replicated with a range of cores, from my simple thin blades to pickup poles and everything in between with success, I can reasonably assume that this is less of a factor. The wire size and coil qualities, though I have not measured them nor the ability to do so with any kind of sensible and useful data, do seem to make the biggest difference. Again, my advice was in line with that.

That conclusion doesn't take into account how the effectivness of a driver depends also on the path length of the magnetic circuit and the flux density at the strings.

All these things could easily conspire to give you similar result from different core specs. Without measurements and data, drawing the conclusion that core specs don't matter while string gauge does is silly.

Mis-information is when someone pops up, puts up a tutorial of basically the same design, claims that any wire is fine (then reveals he used the same wire gauge as the original), claims that it is different in the use of a different amplifier chip or buffer stage instead of the F/R which I have never used (though can work) and say there are significant differences...that's mis-information.

Misinformation is when someone provides information that is false an that may in some way make things more difficult for the recipient of the information. One such example is saying that only DC resistance and wire gauge are important, and the the rest can be ignored.

The 0.2mm wire was arrived at to produce a coil that worked, it can be explained through inductance or turns or whatever, the end result is that this spec for this design is relatively important to achieve the result desired. You don't need to worry about the theoreticals if you stick to what works...one should not at all be surprised if almost all of the important elements are neglected if the thing 'fails'.

All that might be true IF

The system that was presented and has been recommended was a complete system. Unfortunately, the driver is only part of the system, and your driver specs are only optimized for the rest of your system which AFAIK has always been secret ?

If someone provides a complete design with exact specs - including stuff like core materials, wire manufacturer and type and (significantly) circuit diagram with parts list, then it can be cloned be someone else without needing knowledge and calculations.

Otherwise, knowledge is needed for success. Unless folks are willing to go through months of trial and error spending loads of money in the process.

So, if you aren't willing to give folks the design for your complete system, you should probably keep quiet about it unless you are willing to accept that they will need the knowledge required to tweak the driver to match the rest of their system (which is unlikely to be the same as your secret one)

With someone (OP) at this stage who lacks the knowledge to measure a coil even when demo-ed succinctly with photos, seems to me to be not at a stage where considering the complexities of inductance or even the mathematics of making a coil that might be the equivalent to the intended design; it is a little much to get one's head around compared to just getting a wire size that is known to work and using it as intended as I suggested.......................

No need for complexities of inductance, just a cheap inductance meter would do. No more difficult than measuring resistance.

In fact, it would be really useful if you would measure the inductance of your driver as that would give folks a target to aim for.

That got me thinking, 0.3mH. By splitting up a coil into multiple coils in parallel one could reduce the self inductance. That would be ideal, because we want the self inductance to be as small as possible. We want a driver that is 'ohmic'. I only think that in our case the inductance doesn't equal 0.3mH because the 4 coils are coupled on the same core. But what if the 4 coils have their own core?

Then the next logical step would be to wind 4 coils 32 ohms/4.0 mH. Thicker wire and more turns.

Cheers

FF

Yes, its an interesting path. I have been tinkering with this approach for a while.

At the moment I'm waiting for Maplin to get 0.19 wire back in stock.

then I'm going to wind two 16ohm coils and wire them in parallel. this will give 8ohm DC, and the combined inducance will be roughly 1.2mH. The current through each coil will be half, but the magneto motive force (therefore field strength) should be plenty more then double, so should provide a more powerful magnet while still preserving the correct input impedance....

Theres a big post about the idea lost somewhere in the old ideas thread.

cheers

Col

col · April 15, 2010

Can't EDIT previous message, so here what I was going to add:

EDIT: the basic idea is that magnetomotive force (proportional to field strength) is I x N (current times number of turns).

So two coils set up so each has twice the desired overall inductance and resistance should give roughly (not taking into accoulnt that they will interfere with each other somewhat depending on how closely mounted they are):

if 1 core at 1mH takes 120 turns, then a 2mH coil will be 170 turns (in this hypothetical case, we would choose wire so that 170 turns gives 16 ohm or as close as possible)

Lets say we had 100mA going through our single core 120 turn driver, that would be 12 ampereTurns of magneto motive force.

now with two coils of 170 turns, each with 50mA we get 17 ampereTurns.

So we get an increase in magneto motive force of just over 40%

Now, it is important to remember that the field strength is the magneto motive force divided by the 'effective path length'. This means that for a given magnetoMotiveForce, the further the flux has to travel, the weaker the field will be.

I think the best approach with this in mind is compact coils (another point for you Pete ).

It also helps if we have core material everywhere but at the strings, as the field will build up in the 'air gap'. I guess this is why my old dual core driver worked well.

It also explains (one of the reasons) why my more recent driver built using standard humbucker bobbins wasn't so good - too big, with too big a gap between cores.

So to reiterate, I'm going to build a compact driver with two 16ohm cores wired in parallel, aiming at a combined inductance of around 1.2mH.

This worked out at 0.18 or 0.19 wire being the closest to optimal. I'll fiddle with core specs to fine tune the inductance.

cheers

Col

psw · April 16, 2010

I take your points...and you are right...just perhaps too technical for this OP to make that much use of it perhaps...

if I had an inductance meter I'd gladly measure it, there are so many variations that I am sure that it will be quite a bit of variation in the various versions. But, there is a wide margin of effective driver impedances and inductances that have shown to work.

My "complete system" evolved, I have stated that I used modified kits originally based on the CHAmp (an LM386 circuit pretty much identical to the Data Sheet LM386 on x200) and usually a 100uF output cap...this has been repeated often enough, but I have shown a vast number of amplification versions with different chips and variations most of which worked as well...its an amplifier.

For the preamp, in the early days I also tinkered with the PreCHAmp kit, designed to match the CHAmp and used a two transistor push pull kind of thing. Tinkering allowed the control of gain for instance that varied widely over the years and in some versions allowed for variations with trim pots and such. As they were running together, some things in the powersupply could be shared.

These were inside the "sustain box" and the original "sustainer strat" but were a stop gap solution on the way to developing better purpose built circuits, I anticipated with the input of people more expert in such areas, like your good self. I even tried to commision a professional at one stage, but they came up with the exact same thing that I was already using. These circuits were developed for the Mag, 'Silicon Chip' many years ago and CHAmp stands for...Cheap and Handy Amplifier...the kit is widely found and works out cheaper than the components alone.

But, in kits or simple circuits, opamp versions can be made that work just as well with less components and size.

These things were also used for the bulk of the Hex Drivers as well...it's just a small non-loading amp...no magic circuit system at all!

The F/R is not as bad as people make out for this either, it is just missing all kinds of filters to help prevent internal oscillation at high gain, all the things in the data sheet, and these things I have specified many times to improve it.

The Fetzer in the preamp...my main problems with that is not that the thing has some warmth to it, but the need to have trim pots for biasing the transistor and such make the design not the best and clunky and not the best for beginners compared to say an op-amp solution. Again, like my early 'any amp will do to start with' approach and concentrate on the driver as suggested to me by LK very early on, and proved to be sound advice.

So...any amp could work if matched to the driver and pickups...some might be better than others, but I kept things simple and constant to test drivers with identical circuits and conditions and explored alternatives to work towards something more purpose built for this purpose.

I really don't think the circuitry is as big a deal as people make out. I have had other reservations that I feel are true as well about my own more recent circuits...much of which I have given away in the principle and general component blocks. I was never happy, especially with the amount of scuttlebutt and posturing that started to emerge, about promoting other peoples designs like the F/R or the reaction to any design anyone proposed. I saw a lot of things both online and privately, and still do, that seek to take such circuits, lambaste them as not being "perfect", changing a few components, put down the originator, then call it their own. So, that attitude alone is enough to put me off personally. For people who do know about these things, more than enough has been said to extrapolate what I have been doing later and what designs have influenced me.

One of the things that influenced my recent purpose built circuits of course was your work and ideas col. The use of AGC though known, but the discussion around types and advocating the things, got me looking at different options and making something basic of my own using the LM386...nothing as complex or elegant as yours of course, but elegant in it's simplicity. I was also influenced to cut right beack on preamp power, so that with improvements to my drivers, I found that I could scale right back to a buffer only in the preamp, where early on I sometimes used massive amounts of gain. With AGC, it will of course drop even lower when not needed, but I also found that I could run it without effective AGC for that sound at higher drive levels...creating a different kind of drive control with more range than I used to get. But, when at these kinds of levels, we are really looking at my disclosed LM386 with 100uF output cap mod and a buffer...so a very simple replicable circuit.

...

PSW: If people go only on the number of turns, other factors like the bobbin and core size, the tightness of the coil or if they stretch the wire as they turn, can all have an influence. If you use the correct specs the DC resistance will give the required turns to make the thing work.

COL: Not true!
If people go only on the number of turns, then they must use exactly the same dimensions of core made of exactly the same material to get the same result as a driver they are copying.

hmm...maybe you mis read it, but I was clearly saying exactly that...so it is true...the core size and such make a dramatic difference, I believe the overlapping of coils do as well (hence the thin coil) and I recall you feeling that there is perhaps an optimum cross section of coil as well...maybe wrong there. There are of course other elements to the "thin coil theory" in my design philosophy and that element is contentious.

But...yes, I am saying that only going by the number of turns, as the OP and others have obsessed about is no good...reading the resistance is the easiest thing with the most common of tools...and is effective to get a resistance that is in the range the circuit wants to work at. We know that 7-9 ohms will work, most of mine of the standard 3mm design work out to be on or slightly less than 8 ohms for instance.

So...with the abilities and tools available and the experience to interpret them, reading the resistance of the coil with the formula provided will reap a result in the workable range and is entirely reproducible and has been done with all kinds of variations in core dimensions and such...by me and by very many others.

...

Col, some really cool work on the drawing board there...yes, I have always thought the gap between the cores on dual coil drivers may cause problems. Shame, as I wanted to take the 'wafer coil' things I last built to that obvious conclusion with a fit on cover for HBs...probably would be able to get it to work though.

My attempts at multi-coil devices (my hex coils were very different in every way to these ideas btw), like my mid-driver rail thing the size of a single coil...seemed to have some inherent flaws that I didn't like. The only reason that that mid driver rail thing was not successful btw way that I was ambitious in the application...putting it between two active pickups and closer to the bridge, which was the point of the exercise, was pushing things a little far...I had hoped to get something like Dizzy got with his bi-lateral design.

I have not been impressed by my attempts at bi-lateral and stacked designs either...and they take a lot of work.

My pursuing of the single coil is that they do seem to have more "throw" to them, working a decent distance from the string and so evening out some of the effects of action that dual coils seem to have problems with (they seem to need to be close and constant, or at least more so, distance from the strings).

The reasons for pursuing them is largely, for me, was in special circumstances that required less EMI effects, like the mid driver...unless you were going to go the full hex or some other variation for better polyphonic or even response or something.

If, and I seem to be able to, be able to make a single coil driver compact and meet the other criteria that I set for myself, I personally don't feel the drive to develop too far into this area. The only thing that I did have on my mind was the HB wafer coil adapter like I did for the strat pickups but for HB neck pickup guitars...but there again is a special circumstance.

Once you sacrifice the neck pickup, like your own or FF has done, and many others, a lot of complexity is stripped away, especially in switching and installation and EMI problems associated with nearby coils not in use but perhaps still connected to ground and potentially introducing noise.

So, if the single coil creates the effect, has more throw and is far easier to make and to make compact...I don't feel the need personally to go too far into that area. I've always felt that the more compact the driver is, the less EMI spread there is, and having the bulk of the coil as close to the strings as possible, is also advantageous. All these things are easier with a SC device and are happily in line with my global low mod ambitions in my own work that for me determine the degree of success.

But the things are intriguing and could lead to more adventurous designs down the track I guess and i am always interested to see what people make of them.

I'm not sure of the calculations for a 40% increase in force, I think other factors may influence things...but I can see where you are going there I suppose, more force means less power required, cleaner headroom, battery life, etc. Just not sure if other factors of parallel coils are not going to negate things a little. Certainly, my dual coil drivers seemed to be less effective, not more...and have their own problems besides more complex and perhaps bulky construction. More power of you (pun intended) if you are successful though!

It does occur to me, thinking back to a lot of my dual coil designs, that the cores were radically thin to make them that compact and these may well have been a very bad strategy on my part.

...

Just over viewing your posts, enamel thickness and such can make a difference, in more recent times with wire obtaining difficulties, I bought a big roll...and this had thicker enamel. But, the ability to make a very tight and neat solid coil with experience and practice, made far more difference than the insulation thickness.

Two coils of 16 ohms with almost 0.2mm wire is going to be significantly bigger in every way...not sure if the increase in force isn't only going to be realized with more power...then there are other factors like the resonance of this coil structure and things. Probably will work, but like the thicker wire guys, this will give you the option of more force, but in comparison tests, they only worked with more power applied and their characteristics seemed to have problems with higher strings.

I was not suggesting that wire gauge and DC resistance is all that matters...I pointed out we know nothing of the core or magnetics in this case. But when people are using playdoh and not knowing how to measure the resistance of the coil (let alone the inductance and knowing what to make of such a measurement as there is no data anywhere as to what to aim for)...of course all these other factors matter and are more important technically, but in the end, a close enough working device can consistently be made with these things as a guide. Without core dimensions and factors like the winding style (how tight, etc) turns are the most leading guide to making a successful driver...IMHO...for this kind of level of exploration and motivation.

...

One of the reasons for "secret stuff'...well, it can be a complex of reasons really. I did not set out to dictate "a way" of doing these things...I set out to explore and share my explorations and in so doing, and with a lot of ongoing interest that fueled that, encourage others to help develop things further. To a large part this did occur, but far more was the call for "the system" to be made and exposed...then in recent times be criticized or co-opted or even sold as someone else's idea. This is not a new thing either, but it has happened enough...and a conclusive circuit by me I felt may well have killed a lot of exploration. The reality has always been that there are people better equipped and skilled than I am to come up with appropriate amplifier circuits, that's all that was ever required. Also, we all have different aims in response, mine have always differed from yours col, but both are valid, just different sounds and responses from the technology. I was never aiming to replicate the fernandes or other systems and I have interests beyond that of "endless sustain" that most people associate with these things.

I don't have "a system" as such...well I do, but that kind of evolution gets close to commercialization. But there is no "secret" or mysticism about it except by the suspicions of others and the superstitions people have about circuits. A good non-loading amplifier circuit that can be stable at high constant loads (without shutting off for overheating, etc) that matches the driver and the input is all that was ever required. People give me too much credit even in the fostering of the "secret" conspiracies in this regard concerning circuits for what I am capable of doing I suspect. Privately col, I wouldn't mind so much with you, maybe I have already suggested to you the kind of things I use these days, I don't recall, but open forum is inappropriate in my book.

The only thing that I think that some people neglect, that I have always been very open about, is that with a LM386 circuit, I find that a 100uF output cap gives a sound and response I particularly like. No secret there, this mod will make the circuit more treble biased and perhaps a little less laggy. The upside, perfect high string response and less power to drive them and a more even overall drive, the "down side", though I like it, is that in high drive in normal mode, the lower strings, and notes up to about the second C (5th fret, g string), these notes bloom to a harmonic of usually a fifth above. As I say, I like it, it will drive mainly fundamentals if drive is lowered if wanted, but then high string response wears off...but then I have a control to turn it up as well, so, why not. But then a lot of this has been often discussed, so no secret at all.

More important to success seems to be installation on multi pickup guitars, but again, I seem to have done most of the work in this area and my findings and solutions have often been posted...no secret there either.

The real thing is the driver to me, and there I have been more than open. The more open I am the more criticism I seem to attract...this put me right off my circuit design exposure. Now it is that I am making people use 0.2mm wire and attributing magic qualities to it and spreading needless pain on a design that does not work...patently untrue. This simple design will work, significant variations to it will provide significant variations in performance....that I regard as a failure. Saying any wire will do is far worse...not that there is something magical about the wire, just the end result.

Oh...core material...no problem. It is known that typically I use ordinary 3mm steel that I bought in the hardware store for a couple of bucks and cut to size. On strat type pickups and adaptions, I use the standard alnico poles...I have tried adjustable bolts and laminated steel from computer chassis, I have used exotic ferrite materials and powdered pure iron and epoxy...all work, but the steel seems to be plenty good enough. I don't like to go much bigger for a few reasons, but I do aim to keep things compact. The blades I used were cut from fretsaw blades I think...but that was less enjoyable to do...

... Not sure if that helps with any perceived "secrets" of not...the fact that others have been able to replicate similar results shows that my secrets are not that well hidden. I can assure people that playdoh was never a secret 'ingredient' though I have made liberal use of PVC and double sided tape over the years to stick things down!

col · April 16, 2010

I take your points...and you are right...just perhaps too technical for this OP to make that much use of it perhaps...

I think it's for the OP to decide what's too technical for him ?

Col, some really cool work on the drawing board there...yes, I have always thought the gap between the cores on dual coil drivers may cause problems. Shame, as I wanted to take the 'wafer coil' things I last built to that obvious conclusion with a fit on cover for HBs...probably would be able to get it to work though.

I was slightly wrong with my explanation, but didn't have time to edit.

The width between cores of my old dual driver was the same as with a humbucker - dictated by the magnet.

The difference is that the humbucker has much deeper cores, so the overall distance the flux has to travel is considerably more, spreading out the flux over a bigger area reduces its density, and therefore the field strength. The extra core mass may also contribute to eddy current losses.

I'm not sure of the calculations for a 40% increase in force

They are correct as far as I understand for Magneto Motive force.

Whether they provide a more efficient pull on the guitar strings will depend on the physical design and layout of the driver. Personally, I feel that it might be possible to get a better than 40% improvement over a simple single coil driver, but not without a lot of R&D, and maybe access to a gauss meter (a not so simple single coil driver on the other hand could be better than a simple one - I've posted about that before).

cheers

Col

psw · April 17, 2010

Sorry...i didn't mean your calculations are wrong, but when translated into a device with all the other elements, you may not get that much more efficiency. But any more efficiency is a good thing, less power required means less EMI along with the dual coil design to contain it, plus far easier to get more clean headroom and run off a battery for longer, etc.

So, well worth pursuing. For me though, without extensive modifications or doing the kind of thing the commercial guys do, such a driver kind of negates the neck pickup or low mod aspirations I have...but that isn't what everyone wants of course. I look forward to seeing the results.

My reservations are that where you get two opposing magnetic fields, the shape seems to be contained a lot between the two blades, closer it gets more contained. As a result, it seems that the influence of the forces seem to be more between the blades than above them...but I could be wrong. A bit early in the morning here and a while since I was building them.

...

I think it's for the OP to decide what's too technical for him ?

Maybe, but every indication is that there has been little research or knowledge yet about the requirements of such an undertaking...if the basics are not down (potting coils, reading resistance, buffering the amp...even with all the information out there on these things), then more technical information may indeed not help things...well, I think that is not that unreasonable to surmise...but we may already have scared him off...or perhaps the OP is busy getting a handle on these things and exploring options as I hope.

...

There certainly is a lot more that can be done in the area of the driver. Well, circuitry too if one has a mind to design things for the response they want or tailor it to a specific drivers qualities. But it is good to see more innovation into alternatives.

The single coil driver is not necessarily "the best"...one must assume that these simple options must surely have been explored on the way to making the various commercial versions about. They can work surprisingly well though and various dual coil options significantly more technically challenging, at least for me. In the end, I didn't see a big pay off, but maybe I was doing it wrong.

Moving the blades closer seemed to make things like "throw" a little worse, you needed to get the things very close to the strings. I did make at least one bi-lateral compact driver which was an "interesting" experiment but didn't work quite as hoped either.

...

Still I am conscious that this is all a bit off topic as always seems to happen, so apologies.

I have been posting elsewhere on some sustainer stuff and may well be getting back into it after this next move in a couple of weeks...I have at least one project that really needs completing and a few others guitars with special features that should be done.

One thing I have noticed is that many report that their fernandes systems are very even and 'polite' for want of a better word, and that is what was judged between the sustainiac and mine in comparison. The DIY versions do have the potential for a lot of dynamic range if that's the kind of thing people are after.

These aspects are often left out or not fully explored in general. Often clouded by assumptions that there is a "best" or "ultimate" device or sound. It would be good if there was more of a statement of aim with some of these devices as to what the builder is aiming for.

For me, I have always liked the dynamic response that can be got from these things, I feel there is a lot of expressive potential there...others may just as validly be pursuing a constant even response that never drives the strings too hard or varies too much to the players touch. Personally, I like the more ebow like sounds that can be got with something that is very sensitive to the player, a lot more bloom, a lot of the harmonic effects. The simple devices can do this with simple circuits pretty well, and fill a lot of the other criteria I set myself.

Still, there is some amazing potential really...I did a bit of brain storming the other day and thought that it would be fun to make a "stick" like tapping instrument, dampened at the nut, ad the strings driven by a driver just from tapping the thing. Perhaps in an instrument like this I would desire a more controlled and predictable even sustain that would be more forgiving of uneven hammering with different fingers. Perhaps here too, a multi-driver kind of thing, like hex systems, would be advantageous to explore as well...it's always been fun to think about the possibilities of this device.

col · April 17, 2010

My reservations are that where you get two opposing magnetic fields, the shape seems to be contained a lot between the two blades, closer it gets more contained. As a result, it seems that the influence of the forces seem to be more between the blades than above them...but I could be wrong.

If you were correct, then humbucker pickups wouldn't work, or would have a low powered response. Instead, they are the ultimate for anyone wanting a high powered output - the only passive pickups that can cause an amp to break up a bit without a booster of some sort....

...

I think it's for the OP to decide what's too technical for him ?

Maybe, but every indication is that there has been little research or knowledge yet about the requirements of such an undertaking...

It's still for him to say not you !

He has asked for help, I - among others - have provided him with information. For you to come in and state that the information presented is too difficult or technical for him is extremely patronizing.

......

Please don't take this the wrong way Pete, but this isn't the old 'Sustainer ideas' thread, and huge posts that contain very little of relevance to the thread don't help anyone.

cheers

Col

psw · April 18, 2010

No...I realize that I was being patronizing with the playdoh stuff, but I hope that the OP will at least look at the linked tutorials and stuff that I provided initially and the information provided and do a little study and get to grips as to why things like a buffer is important, potting is vital, and the drive design is important. Being able to read the resistance of the coil is pretty basic stuff, most of these things are well covered with photos for instance.

The last person posting recently, used pickup wire and asked whether when it breaks it could just be tied together.

I am not saying that the information is too difficult or technical at all...I am saying study the information presented and base at least the core principles on that understanding...so clearly potting has always been paramount, there is a reason for the wire and resistance formula, there is a reason for the buffer...and this information was given here and over and over before.

These threads do go a little off topic. I don't think going dual coiled at this stage for the OP is particularly a good idea, we all go off into side topics since everything got closed down...it's a shame, but that's the way PG wants it. I may start again elsewhere or reactivate things on other sites if I do much new work, but PG has traditionally been the home of this thing.

...

No...humbucking pickups work but are well known to get the full power of them close to the strings, pulling them back dramatically reduces the effectiveness of them. My LP, when tested with a driver and so lowered the neck pickup 3mm, hardly produced much output at all. Strat pickups conversely cause trouble if they get close to the strings and will work a surprising distance from the strings.

I know that plenty of single coil pickups pack a lot of output out enough to break up a bit without boosters...there are plenty of classic rockers that used them without booster and got that kind of sound. Of course if these pickups are the source sound, then that same potential break up is going to dirty up the sustainer circuit, so a bit of warmth in the fetzer seems mild compared to a super hot HB. The reality is that there is a margin of tolerance that can still get good clean results.

So...what my concern is that these things may need to be very close to get the AC pulse working on the strings themselves and that the shape of the magnetic field is more contained...this is part of why they can be so quiet as well as the RWRP aspects, the things are very localized.

They can produce far more power not because they are HB, but because there is twice the room on those bobbins as a single coil has to turn wire around, they are not twice as loud however. Their characteristics are also very different and not necessarily advantageous to a driver.

But some of this is conjecture, but some of it is experience...but then I can only test the ones I have made side by side with alternate designs...my dual core blade for instance beside a compact single.

If your dual coil original driver or FF's there or others (like the commercial units), I don't know that I see any massive improvements in the region of 40%. This would surely be a boon, but simply have not seen it in my own work or really in others...but I am keen to see it "proved"

For instance, I was under the impression that you had initially made a single coil driver similar to those suggested, and then chose to go on to the dual coil designs you favored. Did you get a massive boost in performance in comparing the two.

My reservations are based on my own real world experiments, but admittedly there are quite different beasts in many respects, I certainly experienced that close bladed drivers seemed to need to be very, very close to the strings and did not provide any more EMF than the single coil designs. In Bi-Lateral designs, there did seem to be a bit of a dead spot between the coils...in some therefore I overlapped the coils in the centre like a z-coil.

But there is nothing wrong with the Dual coil designs or other ideas at all, it all depends on the results you are aiming to get to. As I can make single coil drivers compact and meeting the criteria I set, and not have excessive problems with EMI, that is still my preference and the suggested starting point for getting into the project.

Fresh Fizz · April 18, 2010

EDIT: the basic idea is that magnetomotive force (proportional to field strength) is I x N (current times number of turns).
So two coils set up so each has twice the desired overall inductance and resistance should give roughly (not taking into accoulnt that they will interfere with each other somewhat depending on how closely mounted they are):

if 1 core at 1mH takes 120 turns, then a 2mH coil will be 170 turns (in this hypothetical case, we would choose wire so that 170 turns gives 16 ohm or as close as possible)

Lets say we had 100mA going through our single core 120 turn driver, that would be 12 ampereTurns of magneto motive force.
now with two coils of 170 turns, each with 50mA we get 17 ampereTurns.

So we get an increase in magneto motive force of just over 40%

'...'

So to reiterate, I'm going to build a compact driver with two 16ohm cores wired in parallel, aiming at a combined inductance of around 1.2mH.
This worked out at 0.18 or 0.19 wire being the closest to optimal. I'll fiddle with core specs to fine tune the inductance.

This appears to be a theoretically sound idea to me. If that 40% increase can be achieved remains to be seen. But since nobody else ever did this experiment I can only say: go for it.

I've read about mutual inductance, but I have no idea how it plays out.

These threads do go a little off topic. I don't think going dual coiled at this stage for the OP is particularly a good idea, we all go off into side topics since everything got closed down...it's a shame, but that's the way PG wants it.

make love .and. war

cheers

FF

therizky · April 29, 2010

hi...whats up?

sorry the reply tooks too long....the personal life suck my a**.

1.one night,ive decided to quit joking around...get some 0,2 wires (as everybody said would work most effective),a dozen of "power glue",wound it to the tiny plate stell as the magnet as it core, tide it to a dmm and get a 8,0 ohms driver....done ......but by the way its still humming,i think its impossible to make it completely silence,except i dont' connect it to the circuit

2.for the circuit,i added bjt buffer in front of little gem from the link that fresh fizz has provide (thanks man,really appreciate that) ,but i dunno, when i connect it to the speaker,the high frequencies still sucked up,and never sound smooth or as normal as it supossed be....i think im gonna switch to another circuit....will "stereomicroamp" from the tonepad.com works? its opamp based and stereo (got two channel).

3.anyway,i got a TA8210 based circuit,the power is crazy,20 watts,and it can take loads of 32 ohms,to be honest i have tried it,i run it on 12volts 2 amperes adapter and use 32ohms driver...it works no doubt...all 3 low strings has moved even before i touch it....but yet the high frequencies response not that good,i think theres a "X" factor that come with every guitar that make the thing wont work as expected...i dunno...maybe the neck,wood,magnetic fields,bridges,etc.theres no way to cure it except move to another guitar.

maybe in the next post ill post some pics about my builds and the "blink sustainer" (named it after the name of pop-punky teenager band,you know who and why) ,just so that someone would find the error of the builds and free me from opening dictionary for my lack of english......and maybe in the end,if finally icouldnt make it right,i quit the 9v version,get stereo guitar sockets,get a box for the adapter,and own a unmatched magnetic sustainer. the only downfall is i must bring the box all the time with my guitar,and have to bug stage crew for ac line....not a very good idea.

thanks....

Fresh Fizz · April 29, 2010

maybe in the next post ill post some pics about my builds and the "blink sustainer" (named it after the name of pop-punky teenager band,you know who and why) ,just so that someone would find the error of the builds and free me from opening dictionary for my lack of english......and maybe in the end,if finally icouldnt make it right,i quit the 9v version,get stereo guitar sockets,get a box for the adapter,and own a unmatched magnetic sustainer. the only downfall is i must bring the box all the time with my guitar,and have to bug stage crew for ac line....not a very good idea.

thanks....

yes, a picture of your driver, blink sustainer, whatever would help a lot.

Hey man, you're the star. Let the crew work for you

Cheers

FF

psw · April 29, 2010

Far too much power there....you clearly have some major problems with the circuits you have tried and this is likely to continue...perhaps you could find someone with more experience in these things to get it going properly.

If you put 20 watts into the thing you will not only get hummming, but quite likely melt the driver. The reality seems to be that it has some major problems and I don't think the logical conclusion is that the guitar is cursed.

Not sure what "power glue" is. If it is "super glue" I would avoid it completely. Same with fast epoxies. The wood glue method was recommended by me and extensively tested as being safe, easy, reversible and effective. Even though it won't set "solid" like a proper epoxy coil (which is specialized, thin, slow acting stuff and expensive) and be self supporting...the wood glue method will gap fill and set enough to be very effective as a damping material if the instructions are followed. That means, pushing the sides in occasionally so that the coil has minimum gaps, and that the whole thing is taped tight, clamped and given a long time to "set".

Otherwise, I think the main problem is that your circuit seems to be having trouble even with a speaker...so it is not going to work with the higher demands of a sustainer.

The mods suggested, caps to stabilize a LM386 circuit, the lower output cap and such will work. For testing purposes, most stomp boxes have an active buffer...any boss box will have one even when off...and this can be used as a buffer to an LM386 circuit.

AC powered sustainers are a very bad idea IMHO. You have to realize that a coil of wire with 20 watts into it is effectively a bar radiator = heater...see the first post of the sustainer thread...these things can get seriously hot, just before they literally melt!

So...it's not the guitar...post some pics and some real details. I don't know what the buffer FF offered was, so I can't comment...but any number of things may have gone or are wrong with your circuit and potentially the newer driver is ok, it just isn't getting an appropriate signal.

This kind of "it doesn't work" posts are exactly the things that others latch on to and then suggest that the design is at fault. This clearly isn't the case, but it has come up a lot lately, over at another forum as well, and it seems to stem a lot from simply not following the design or bad workmanship to an overwhelming degree.

The best thing is to completely show exactly what you are doing...the driver, the guitar, the circuit and the diagrams you are following. Don't think that the guitar is cursed or fall for the more power will fix all ills approach, it only ever makes things worse!

therizky · May 10, 2010

1."Far too much power there....you clearly have some major problems with the circuits you have tried and this is likely to continue...perhaps you could find someone with more experience in these things to get it going properly."

yeah...since we're not talking about circuits that run on 2 amperes currents and provide 20 watts of power i think its out of disscussion...and im gonna think it over and over again to put it in my guitar.

2."Not sure what "power glue" is. If it is "super glue" I would avoid it completely. Same with fast epoxies. The wood glue method was recommended by me and extensively tested as being safe, easy, reversible and effective. Even though it won't set "solid" like a proper epoxy coil (which is specialized, thin, slow acting stuff and expensive) and be self supporting...the wood glue method will gap fill and set enough to be very effective as a damping material if the instructions are followed. That means, pushing the sides in occasionally so that the coil has minimum gaps, and that the whole thing is taped tight, clamped and given a long time to "set""

no...its just daily cyanoacrylate (is that spelled right?)....but the wood glue method seems to be easy and cheap too,im gonna try it in the next build.

3."Otherwise, I think the main problem is that your circuit seems to be having trouble even with a speaker...so it is not going to work with the higher demands of a sustainer.

The mods suggested, caps to stabilize a LM386 circuit, the lower output cap and such will work. For testing purposes, most stomp boxes have an active buffer...any boss box will have one even when off...and this can be used as a buffer to an LM386 circuit.

AC powered sustainers are a very bad idea IMHO. You have to realize that a coil of wire with 20 watts into it is effectively a bar radiator = heater...see the first post of the sustainer thread...these things can get seriously hot, just before they literally melt!"

yeah...the lm386 never being a good boy to me....little gem circuit relly need some improvement too...i wonder why rog didnt give some notes there...they just suggested to use another input and grounded the second one...it doesnt give much help......for the buffer,im gonna use 741 chip based,found somewhere on the net,but forgot the site,gotta google it again now.

and the 20 watts driver become hot....not only the driver,but the ic also (guy who sells it recommend me to use heatshrink)...not as heat as radiator,but yes,its hot...dont have anything againts that one.

4."So...it's not the guitar...post some pics and some real details. I don't know what the buffer FF offered was, so I can't comment...but any number of things may have gone or are wrong with your circuit and potentially the newer driver is ok, it just isn't getting an appropriate signal.

This kind of "it doesn't work" posts are exactly the things that others latch on to and then suggest that the design is at fault. This clearly isn't the case, but it has come up a lot lately, over at another forum as well, and it seems to stem a lot from simply not following the design or bad workmanship to an overwhelming degree.

The best thing is to completely show exactly what you are doing...the driver, the guitar, the circuit and the diagrams you are following. Don't think that the guitar is cursed or fall for the more power will fix all ills approach, it only ever makes things worse!"

the buffer ff offered is right <A HREF="http://http://www.muzique.com/lab/buffers.htm">here</A> ,and i didnt said that the design is fault,i just said that diffrent guitar would yield different results also...its simple,different brand,different pickups,different geometry...each will give different result and also need special care to be worked on...maybe thats why people keep facing different funny problem over and over again.

well,anyway this is my whole work for 4 months...its flying around my room...i dunno where the rest.

1.first,the driver

currently in use is the left one,with black electrical tape.

2.the circuits.

has built some circuit to be worked with...up there is ruby,fetzer,bjt buffer ruby,little gem,amz muffer,,tiny audio,utylity boost by tim escobedo,and the biggest one is blink sustainer i was talking about.

3.now,the guitar.

this is my current setup,the little gem loaded with amz muffer....looks to be worked,but it didnt.

some zoom in.....dont need to phase the driver,my strat has individual pickup control and phase switches...so i could try it on each pickup..or all three...

now some picture of the circuits...running out of hookups wire,i use the wire instead.

curent sustainer

the two circuit.

what should i do now? can someone suggest me another chip or circuit to be worked with? what is the disadvantages if using a little bigger wire,other than it takes more spaces? ive been built some stompboxes project before and also wound my own pickups...this project take me beyond the two...if only i succeded.

please help.

Edited May 10, 2010 by therizky

col · May 10, 2010

3.anyway,i got a TA8210 based circuit,the power is crazy,20 watts,and it can take loads of 32 ohms,to be honest i have tried it,i run it on 12volts 2 amperes adapter and use 32ohms driver...it works no doubt...all 3 low strings has moved even before i touch it....but yet the high frequencies response not that good

32 ohm, but what inductance ?

For a driver with a ~55mm long core wound with 0.2mm wire to be 32 ohm, you would need roughly 470 turns.

A few basic rules:

inductance increases with the square of the number of turns, so if 120 turns gives 1mH inductance, 475 turns will give about 15.5 mH

In the same way that capacitors act as high pass filters, inductors act as low pass filters, the higher the inductance, the lower the cutoff point.

e.g.

at 660 Hz (12 fret high E string), an 8ohm 1mH inductor (driver) will have an impedance of about 9ohms

a 32 ohm 15.5 mH inductor will have an impedance of 71 ohms

What this means is that you need to be very careful about how many turns of wire are on your driver. Too few and you won't get enough drive, too many, and the higher frequencies will be wiped out.

For a 32 ohm driver to have a low enough inductance, you would have to use very thin wire. this would not be able to carry the current you are using without burning up, or at least severely damaging the performance.

i think theres a "X" factor that come with every guitar that make the thing wont work as expected...i dunno...maybe the neck,wood,magnetic fields,bridges,etc.theres no way to cure it except move to another guitar.

The only 'X' factors in the guitar that really matter are: the action, what strings you use, what the bridge pickup is and how much room there is in the wiring cavity. If you get the design of the driver and drive circuit right, it should be tweakable for any normal electric guitar.

cheer

Col

McSeem · May 10, 2010

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

col · May 10, 2010

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

That's pretty difficult, because so much depends on the physical layout and the cores you use.

If you take two single string coils that are each 1mH when alone and place them next to each other, the effect of mutual inductance with either increase or decrease the effective inductance of each coil (depending on if you have them in or out of phase with each other). How much they effect each other depends on proximity and, (i suppose but am not _sure_) how much of the inductance is being created by the core material.

As far as what inductance would be optimal per string, you can work that out for yourself. There are a few online calculators that can help a lot.

pickup calculator can help you work out roughly what wire gauge you would need to get a desired DC resistance from a particular number of turns.

the Inductance Calculator here can help you work out what the impedance will be for a particular driver (assuming you can calculate or measure roughly what the inductance is)

the RLC calculator here can help you work out what response a particular combination of cap and inductor (including the DC resistance) will give you... so do some research to find the frequency ranges of each string, then use the calculators to work out some ballpark figures to get started with your experiments.

To get you started, a 55mm driver wound with 150 turns of 0.23mm wire wound round a thin steel core gives and inductance of roughly 1.2mH. You can use the relationship between number of turns and inductance to calculate rough inductance values for similar drivers with more or fewer turns:

e.g.

how many turns would I need if I wanted a 3mH driver ?

150^2 = 22500 (inductance varies with square of number of turns so square our initial number of turns)

22500 / 1.2 = 18750 (we know 1.2)

18750 * 3 = 56250 (we want 3.0)

sqrt(56250) = ~237 ( square root gets us our answer)

so using the same core, I should get near to 3mH by winding 237 turns instead of 150.

now, assuming that 8ohm is what we need in this case (it might not be, depending on the drive circuit) what wire would get nearest ?

using the pickup calculator, we find that 237 turns of 0.275 (SWG 32) wire gives 8.2 ohms which is close enough

Of course, for this to be useful, you need some way of measuring a baseline inductance for your driver wound on your core material. Cheap inductance meters can be had for less than £20, and although not high precision, are plenty good enough to guide this kind of process.

good luck

Col

Fresh Fizz · May 11, 2010

1.first,the driver
currently in use is the left one,with black electrical tape.

Can you tell me what this new driver looks like. I can imagine the driver on the right doesn't work properly. The wire is not wound tightly around the magnet/core. Of course it's impossible with a core shaped like that. You need a bar magnet!

Is the cover (or top part of bobbin) of your driver made out of iron? That shouldn't be, it should be made out of plastic, or at least a material that isn't magnetic.

2.the circuits.
has built some circuit to be worked with...up there is ruby,fetzer,bjt buffer ruby,little gem,amz muffer,,tiny audio,utylity boost by tim escobedo,and the biggest one is blink sustainer i was talking about.

Like psw told 20 Watts is too much power, the challenge of building a good sustainer is that it's a balancing act. With 20 Watts you have enough power but 1. on the long run you'll fry your driver, 2. string rattling, 3. squeal problems

3.now,the guitar.
this is my current setup,the little gem loaded with amz muffer....looks to be worked,but it didnt.
some zoom in.....dont need to phase the driver,my strat has individual pickup control and phase switches...so i could try it on each pickup..or all three...
what should i do now? can someone suggest me another chip or circuit to be worked with? what is the disadvantages if using a little bigger wire,other than it takes more spaces? ive been built some stompboxes project before and also wound my own pickups...this project take me beyond the two...if only i succeded.

please help.

If you say it's the driver that's causing the problem then my advice would be build a new one, use a bar magnet (ceramic) and 0.20 wire. You could show us what you're going to use (picture) before you start building.

Cheers

FF

Fresh Fizz · May 11, 2010

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

The 'solution' is to drive each driver coil with the required amount of power. So maybe the low E needs only half the power the high E needs. Build a driver that works on the high E and you'll be fine.

FF

psw · May 11, 2010

i didnt said that the design is fault,i just said that diffrent guitar would yield different results also...its simple,different brand,different pickups,different geometry...each will give different result and also need special care to be worked on...maybe thats why people keep facing different funny problem over and over again.

I don't think so, many of the problems seem to be exactly the same...or people go way outside the specifications of known working designs.

Thanks for the pics...

http://en.wikipedia.org/wiki/Cyanoacrylate

Cyanoacrylate = super glue

This kind of stuff is dangerous and not suitable to this application regardless of what others have suggested. It has no "filling" properties necessary for 'potting' coils...PVA types of wood glue have a lot of filling qualities and are safe and water based and reversible and don't dry instantly causing voids in the winding of coils with glue.

Though I have used epoxies in recent times, I am experienced in their use, but even then went through quite a few formulas to get something that had the right qualities...hardware store 5minute epoxies don't do the job too well at all and certainly not for a beginner.

It is hard to tell what the actual construction of the driver is, but are the bobbin plates metal? Where is the magnet...is the magnet the core, what size is this magnet?

32 ohm drivers are far beyond the specifications I deduced, I bow to Col's descriptions there, but thinner wire as col points out may not produce the power required.

You have some very dodgy wiring there...I would not be using enameled winding wire to solder circuit modules to one another. It's hard to tell what you are using for a guitar lead there, but is that even shielded? All these things seem to be indicating a lot of places to pick up a lot of noise and so effect the performance of any circuit. You really should be taking a direct wire from the bridge pickup itself before the controls. On a guitar with this many mods and I can't really tell what in it...there are going to be a lot of problems I predict. My first testing guitar for these things had cheap low powered single coil pickups (as shown in the pictorial) and phase switches on each string...but for this to work all this needs to be bypassed really. It should work in testing mode, but I'd be running a shielded lead from the bridge pickup out the back trem cavity to test these things...all leads should be reasonably short and be shielded before the circuitry, a decent gauge of plastic insulated wire should be used to the driver and other components...especially with dealing with things like 20watts of power, you ahve to have wire that can effectively carry the current to the components like the driver.

There are significant problems with varying from the design to this extent, and this is where things go wrong. I have built and test these things on numerous guitars from high end custom made guitars with name brand HB's, my telecaster, no name teles and various strats...even bass guitars. There are some quirks between guitars, but all have "worked" and with little or no mods required to the basic design.

If you have 9's on the guitar or very light strings, this can cause a problem...my tele would probably be ok, but generally I use 10's anyway and this is the recommendation from the commercial units for good high string activation.

...

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

The details of the Moog guitar and even it's performance other than in promotional material is yet to be established and so very much based on assumptions on hype. We can assume such a system to some degree..certainly hex with signal probably generated from the piezo bridge and individual amplification. It does however provide for a complete all on sustainer kind of thing...so effectively no longer "hex" and one wonders why you'd do this if the hex system was "all that" and more. The muting thing is very much up to conjecture. I would not be surprised if this was a sensor on each pickup that initiates a direct current into the coils with a low threshold signal compared tot he string played, resulting in the strings being pulled to a stop by a strange electromagnet action...but this is all guess work. I could find no patent or technical data or independent reviews that detail much.

I built many hex designs...I felt they were innovative enough to keep 'secret'...I took a dramatically different approach than any that I saw proposed because none seemed to address particular problems I could foresee and experienced. Mainly, that these things interact. I settled on components that I could modify that dispelled with winding and developed a novel magnetic array of balanced fields. The "coils" weren't magnetized as such, but disturbed these fields around each string pulling them in a more orbital manner. Even with such miniaturization and successful testing on individual units...a hex array still interacted. There were something of a nightmare, but I was building at least one a week of different variations for over a year before abandoning the concept.

One thing that wasn't built that did show promise was something I called a "drive engine". Basically working like the cylinders of a car, in my imagination. You have an amp circuit and a sequencing switch chip take the signal and very quickly sequence through say, pairs of drivers far apart, or indeed individually. So, at no time would all six be operating at a time and string momentum would carry the "sustain" through the sequence...as I say though, very fast. It was clearly beyond my ability to successfully build and troubleshoot the inevitable clock noises and such...but there might be something in this approach, and potentially only need one small amplifier, probably smaller than that required to drive all six strings.

I think FF comments are equally valid...the appropriate power and perhaps circuit bias for each string is probably the go and certainly the most obvious starting point with a true hex system. My hex systems generally assumed a mono source and intended to work towards optimizing the natural power ratios with the drivers to create an even balance and better polyphony...or so was the hope. I just developed individual miniture drivers that worked on any string to start with and intended tweaking from there. But, I failed to create a system that provided the necessary independence and I think this really is the first step to overcome. So far, I have not seen the required solutions offered up.

However, FF's comments seem to be enough if all you are seeking to address is getting strings to vibrate.

The other thing...and relevant to the OP's problems...once outside of the known design parameters, thigs are very much up in the air. There amy be a way of detirmining things with formulas and I am not known to be any good at that at all, but when presented with such things, I have to consider if these proposals could physically be made to the size required to even prove them. So...

using the pickup calculator, we find that 237 turns of 0.275 (SWG 32) wire gives 8.2 ohms which is close enough

I'm thinking...say you wanted this value for a single string driver...this calculation seems to be for a 55mm full coil...so, for a single string you would need say a 5mm coil and given string spacing of 10mm, then there is only 2.5mm to wind on...so how deep at theses things if you have to put that kind of wire (o.27) and what effect would this have...I mean the variables and practicalities are enormous with this kind of approach. I have wound hex coils a little like this, or as FF did, very rough, and made neat little single string drivers with thin wires and small cores that worked fine (I wound the original around a 4mm ferrite bead with spectacular results, see page 2) So it can be physically possible to make the kind of thing FF did of course, and for it to "work". If they can work as a true hex system is another matter I suspect.

My feeling was some of the big picture problems meant that a far more innovative approach was necessary...things like the magnetic array thing, the drive engine and countless other flights of fancy. The ironic thing is that many of these things were based on "fears" about the effectiveness of a simple design, many assumptions that I sense are being repeated before me in this subject and that I once had, a drive to make something "better" and "novel" and, an ability that I ahd developed to make these hex systems without winding coils and significantly smaller and with less EMI than anything else I had conceived of. They were even smaller than the smallest simple designs I had created till the newer wafer coils...but I don't know that these would work without the core of the host pickup.

So...unfortunately, I couldn't guess as to what is "optimal"...it is still not entirely sure what you are trying to do...but I would start by using FF's assertion that if it work on the high e notes, it will work on all, and altering the power could provide "balance". Regardless of anything you might do, there will for sure required a whole lot of experimentation...as I say, I built well over 50 drivers in my "hex era" on these things. There should be plenty of pics about...I could refind them from photobucket, but there are a lot of them.

col · May 11, 2010

I'm thinking...say you wanted this value for a single string driver...this calculation seems to be for a 55mm full coil...so, for a single string you would need say a 5mm coil and given string spacing of 10mm, then there is only 2.5mm to wind on...so how deep at theses things if you have to put that kind of wire (o.27) and what effect would this have...

You are totally missing the point of my post Pete.

I am explaining the process you need to use to work this stuff out. I used a 55mm driver as an example because thats what I have measurements for.

The same process will work in the same way for a single string driver, however, the OP would need an inductance measurement from a driver with known number of turns as a starting point - I explained this in my post, and that a 'good enough' inductance meter can be had for very little money.

cheers

Col

psw · May 11, 2010

I'm sorry if you feel I misrepresented you there col...I think I did get the point of your post. It showed that you could calculate things and used a 55mm driver as an example, and I accept your point about inductance meters, I looked into such things a little, not sure that I personally want to go to far down that road and have yet to find one accessible or affordable to me or that would reap the benefit. Reviews I read in regards to using them with pickups, particularly cheap ones, were not at all positive. But I may well not picked up on all that you were trying to communicate, I was trying to add the physical requirements of the application into the equation.

My point was that, looking ahead on the hex driver question...the physical size of each coil is very important regardless of the way the physics of the things are calculated. There may have to be a trade off. I do know from experience but it was many years ago and I don't know the specs unless I posted them early in the thread, that a small coil, about 4mm deep on a 4mm ferrite bead and a total diameter of 10mm with a tiny neo-mag below it and wound with thin wire was remarkably efficient at driving any string with an LM386 type set up. However, the full power of the amp was being used solely to drive the thing.

The complications arise if you were to build 6 of these things all close together in the manner that FF has clearly done.

However, as is well known, I worked by trial and error and built upon successes as well as novel ideas. So, I did not calculate out an optimized coils or anything at such an early stage, I was more concerned to create something that would work and "fit" into the physical application. Just suggesting that perhaps, the "optimum coil" that could be calculated would exceed the space required.

But a small point, as my experience and intuition suggests that a major problem might well be the interaction between such close magnetic coils and cores and fields. Perhaps I am over stating the problem, it remains to be seen what people can come up with if they were to go in that direction.

The 'solution' is to drive each driver coil with the required amount of power. So maybe the low E needs only half the power the high E needs. Build a driver that works on the high E and you'll be fine.

I think FF's 'solution' is the more obvious answer or place to pursue things, far more economically put than I seem to be capable of.

I am all for the calculating of things and working on that basis, don't get me wrong, but at some point these ideas need to be put into a real world application and cover all the bases. My belief still tends to be biased towards actually putting some things into practice before getting too far down the road of assumptions.

A single string driver is a fairly easy thing to do, it can even be made to fit within the space for the application...this has been shown many times...not sure if they are 'optimal', I was going to tweak these things till they were or seek more advice if the more significant problems were overcome.

...

Since it may be hard to find...here is an example of an early single string driver from the hex era using the so called balanced field approach...

One of these were sent to Bill Love for independent evaluation along with it's workings...

Here is an example of the same drivers put into a compact array of 6 in test on what became the sustainer strat with the pickup/driver...

In addition to the novel magnetic array and tiny electromagnetic elements...these things were encased in a thin alloy shell that provided a mold for an epoxy iron shielding compound that held it all together.

...

The point of posting this stuff is not to big note the work I have done, but that a lot of work has been done, some problems uncovered that need to be address, how far and in what directions I took it, and that to a large extent I "failed". It is intended to be instructional. The work was not lost though, aspects of it that were discussed lead to my work with the "thin coil" or compact drivers to lower EMI spread and allow for novel applications with more conventional simple designs...such as the 'stick on' compact driver say on my telecaster or the 'piggyback' things. What I found is that the Hex things could be vastly more complicated and exhibit more quirks than something like that proposed in the Ebow patent say, might suggest...and that is essentially where people are at with these things in the DIY realm...what quite the Moog guitar that inspires these adventures in more recent times, is still something of a mystery to most but it's promoters and designers. It is an interesting thing, but there is still no technical data or first hand independent reviews that I am aware of. It may well be that people are attracted to the idea of the thing, over the 'reality'...certainly a better job at it than I could have built, but exactly what it is and how well it works is a bit of an unknown commodity as far as I can tell so far.

...

As for the actual topic of this post, a bit of a view of the OP's project shows some significant deviations and poor construction techniques that at least would need to be rectified before any reasonable assessment can be made. I think that most would agree that winding wire is not the best idea for hookup wire, shielded wire can be reasonably important, and I'm not sure, but I suspect that most people know of the risks of winding with superglue and it's innate problems as a potting agent.

As so much of these threads are in an effort to fix such problems or address ideas outside the known working solutions, there has in recent times been a perception that there are "problems" with this project and a bit of a myth has arisen around it. It may well be that these "myths" are trying to drive solutions to problems that are not even there in one that works acceptably. My "attitude" in more recent times then is to react to this by pointing out what should be obvious, most of these "problems" relate directly to variations and naive constructions far outside the boundaries of the basic principles and the known working designs.

The "problems" then, are not in the design, or the guitar it is fitted to...but these very things. Typically we even have yet another example of "it doesn't work properly, so I will just throw more power at it solution" and then wonder why that doesn't help any. The basic project is pretty easy and forgiving, so many variations have shown to work on very many different guitar types. What we hear about in this forum generally are the aberrations that give the project a bad name and false assumptions.

Sign In

Sustainer Problems!

Recommended Posts

Link to comment

Share on other sites

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Join the conversation