Sustainer Ideas

[Primal]

That (not-so-subtle) fuzz is the problem I had. I cleared it up by turning the gain down on the LM386 chip. That is the distortion I was speaking of, which is why I find it hard to beleive that psw has had such success with using a distorted signal to drive the strings.

[onelastgoodbye]

I have a 10 ohms resistor between pins 1 and 8, and my gain control pot is less than halfway up, so the 386 is not clipping. Any more volume and the squeal gets uncontrollable.

There has been some debate wether it was best to excite the strings with a clean signal or a full-out clipped-compressed signal. I'm still not sure. You'd think a string would be driven the easiest by mimicking it's natural behavior, thus a clean signal.

I do think the lm386 puts out way more power than is actually needed, provided you have a sensitive enough bridge pickup to go along with it. I remember getting tremendous squeal even from 40 cm away from the pickup on my older (badly potted) drivers.

mmm something comes to mind here...could the magnetic pull of the existing pickups have anything to do with driver efficiency? PSW is using three single coils, they probably inhibit string movement more than my one EMG in the bridge. Maybe that's why he benefits more of the 'raw power' attack.

Spazzyone has proven, that with enough power, you can make a regular pickup into a driver. Even sustainiac and fernandez seem to do it jackhammer style ( I really like your 'only there when needed' approach as an alternative).

I'm hoping to build a humbucking driver in the next few days, might just take care of that EMI.

Tim

[col]

I have a 10 ohms resistor between pins 1 and 8, and my gain control pot is less than halfway up, so the 386 is not clipping. Any more volume and the squeal gets uncontrollable.

Don't you mean a 10μ capacitor ? or 10k pot ?

If you really have a 10 ohm resistor there, you will be nearly at max gain !.

There has been some debate wether it was best to excite the strings with a clean signal or a full-out clipped-compressed signal. I'm still not sure. You'd think a string would be driven the easiest by mimicking it's natural behavior, thus a clean signal.

A fully clipped square wave will cost more battery (I think) - for the same peak voltage, the rms is higher. It will also be more obvious if there is any interference/bleedthrough as primal discovered. I have a feeling that a hard clipped signal is also more demanding on the driver - in my tests with the fetzer ruby, I could hear my driver - like it was a really crappy speaker, vibrating with a hefty rasp, with a non-clipped signal, that is not happening. Of course, that might be because I was pumping hundreds of milliamps into it instead of the 20 - 40 its getting now

mmm something comes to mind here...could the magnetic pull of the existing pickups have anything to do with driver efficiency?

I doubt that is will have a significant effect on the driver efficiency before it destroys the guitars tone.

Spazzyone has proven, that with enough power, you can make a regular pickup into a driver. Even sustainiac and fernandez seem to do it jackhammer style ( I really like your 'only there when needed' approach as an alternative).

I wonder If they're not using something similar. If they are getting the claimed > 10 hour batter life without using a similar approach to mine - maybe by using a class D amp, then imagine what it would be like with a combo of class D and dynamic range inversion...

I'm hoping to build a humbucking driver in the next few days, might just take care of that EMI.

Looking forward to hearing how it goes

cheers, Col

[col]

A couple of pics:

Heres my driver:

Here's the bread board, you can just see the driver, battery box and input from guitar sitting around the board. Two of the 3 pots are for tweaking perameters, the other is for volume. In the final version, it will most likely be either 2 trimpots, or one trimpot and a guitar mounted pot. It looks messy on the bread board, but this will fit in my guitar without much hassle when its properly layed out on perf board - something like 35mm x 80mm x the tallest capacitor, based on my rough layout sketches.

Col

[onelastgoodbye]

its 10 ohms...brown black black (which is 10 ohms right? ..damn resistor color charts). As I recall (it's been a while since I've played around with it),pins 1 and 8 open didn't have enough gain to drive the driver, pins 1 and 8 closed was way too much. Note that I am using the little gem, not the ruby. Are you sure about the max gain thing? I thought it would be around gain x 100. There was no audible ss clipping when I hooked it up to a speaker, though again, it's been a while. I'd go and verify but apparently i've cooked the circuit (seems like my adapter puts out 17v instead of the advertised 12v )

on another note, some interesting tidbits I read on the sustainiac site:

On 2-humbucker installations, we don't recommend using a SINGLE-COIL pickup mounted immediately next to the driver. The driver magnetic field will affect the pickup output in such a way that three of the strings will have lower output than the other three. Therefore, on these installations we recommend that you use a mini-humbucker next to the driver, not a single-coil pickup.

Soo, if I get it right, one coil of the bilateral driver is 'cancelled out', while the other is 'summed' by the pickup coil. Another vote for a full dual rail humbucking driver?

Also: The Sustainiac Stealth or Stealth Plus works very well with active pickups such as EMG. We do recommend that you use a separate battery with the Sustainiac. This is because the sustainers use more battery power than active pickup preamps do. So, you could end up with a dead battery on a gig if the battery dies and you are left with no pickup function. Also, if you run both off the same battery, grunge will most certainly be introduced into the active pickup signal (or even oscillation in some cases).

We supply a special 9-pin guitar jack for guitars having active pickups. This allows independent battery hookup and grounding to prevent these problems. EMG active pickups have reverse electrical polarity than most passive pickups. Therefore, you have to reverse the driver wire colors (black and red) from that shown in our installation mmmm... grunge... oscillation... independent grounding...

[col]

its 10 ohms...brown black black (which is 10 ohms right? ..damn resistor color charts). As I recall (it's been a while since I've played around with it),pins 1 and 8 open didn't have enough gain to drive the driver, pins 1 and 8 closed was way too much. Note that I am using the little gem, not the ruby. Are you sure about the max gain thing? I thought it would be around gain x 100. There was no audible ss clipping when I hooked it up to a speaker, though again, it's been a while. I'd go and verify but apparently i've cooked the circuit (seems like my adapter puts out 17v instead of the advertised 12v )

not sure about the resistor stripes, I gave up trying to read them a long time ago - too many variations on 'standard' markings - I always measure them with my meter. The internal resistance of the LM386 (from the datasheet) that controls the gain varies from 150ohm with the pins shorted (gain = 200) to 1500 ohm with the pins open (gain = 20). sticking a 10ohm resistor in there makes it 160ohm which seems very close to the gain=200 end to me ? gain between 180 and 198 ?

The LM386N-4 is rated at maximum supply of 22v, so if you have that one, it should be fine - the others are rated at 15v, so if its one of those, it will probably have taken damage.

on another note, some interesting tidbits I read on the sustainiac site:

...The Sustainiac Stealth or Stealth Plus works very well with active pickups such as EMG. We do recommend that you use a separate battery with the Sustainiac. This is because the sustainers use more battery power than active pickup preamps do. So, you could end up with a dead battery on a gig if the battery dies and you are left with no pickup function. Also, if you run both off the same battery, grunge will most certainly be introduced into the active pickup signal (or even oscillation in some cases).... mmmm... grunge... oscillation... independent grounding...

Very interesting indeed. maybe the lm386 is loading the battery so that the supply to the preamp is dipping with peaks in the signal - that would certainly fit the type of distortion I'm hearing.

Yet another possible cause to investigate. great info

cheers, Col

[col]

OK, after some more testing:

The difference between using a seperate battery for the pre-amp and sharing one battery is negligible, at least on my setup. I havn't tried isolating the earth - I guess that would involve messing around with little transformers, which I don't have, and can't afford right now - anyone else want to try this ?

(My knowledge of electromagnetism is very basic... would isolating the driver and circuit from the guitars earth using a transformer help - so that the strings are not connected to the earth of the driver circuit ? )

As an experiment, I tried feeding the output to my headphone amp directly rather than through the pre-amp buffer. The sound in this configuration was terrible. I wouldn't be surprised at all if the people who have tried this project and got poor results would do much better with a similar twin buffer setup. In fact, I'd urge everyone to consider trying this.

I spent some more time with the driver above the strings, and, moving towards the bridge, the fuzzy distortion seems to kick in at about the 18th or 19th fret, and get steadily worse until it turns to screaming EMI feedback between the bridge and neck pickup positions. Its definately EMI interference from the driver getting to the pickup :-|

btw, the sound quality with the driver held above the strings below the 17th fret was SUPERB

Col

[psw]

Really busy right now, but looking forward to digesting all this...

I do get a little of that "distortion" and I did have a problem with other pickups till I lifted bot the hot and earth from them. On clean settings (ultra-clear though) I do get a kind of "fizz" in the harmonic mode...have been wondering about this...but it isn't quite what you are talking about I don't think. It may be that my strat has a little more separation between the drive and the bridge...my driver makes no noise, that's a worry, still col, your driver looks good and the circuit intriguing...

pete

[axe_2_grind]

Col,

So what is the second IC you have there? And what function(s) does it serve?

Something to do with "conditiong" the signal as previously mentioned?

Also, for your driver, did you use a standard pickup rail cut to size?

Great pics BTW.

[col]

I do get a little of that "distortion" and I did have a problem with other pickups till I lifted bot the hot and earth from them. On clean settings (ultra-clear though) I do get a kind of "fizz" in the harmonic mode...have been wondering about this...but it isn't quite what you are talking about I don't think. It may be that my strat has a little more separation between the drive and the bridge...my driver makes no noise, that's a worry, still col, your driver looks good and the circuit intriguing...

A fizz in harmonic mode sounds familiar. The background distortion is something that really needs fixing though.

Did I read somewhere in this thread that one of the commercial sustainers at one point had a transformer in it? If so, that might have been for isolating the circuit from ground - that would remove the switching hassles for multiple pickup guitars. On the other hand, transformers aren't exactly small

So what is the second IC you have there? And what function(s) does it serve?

Something to do with "conditiong" the signal as previously mentioned?

In the top down pic, there are three ICs, the left most one is the LM386, the furthest right is an LF358N which is just a standard dual op-amp, I'm using that for buffering the guitar input - one half buffers the input straight to output, the other half sets up the signal for the sustainer circuit with about 3x gain (although that is not final, its one of the variables that gets tweaks a plenty )

The big old 16pin IC in the centre is an LM13700 which is a "Dual Operational Transconductance Amplifier with Linearizing diodes and Buffers" - sounds grand eh ? I'm using it as an automatic gain control (AGC). The gain in the input buffer sets the low level threshold for the AGC which basically gives a big output from a small input and a small output from a big input. Thats the "conditioning" I was talking about and it's all kinda hacked together with a rudimentary understanding, guesswork and happy accidents, and I'm certain someone with a better theoretical background could whip it into much better shape.

Also, for your driver, did you use a standard pickup rail cut to size?

No, its cut from a 2mm x 6mm x 1m bar of 'drawn flat iron' (which is cold steel afaict), bought from my local B&Q superstore.

I seem to be developing an urge to build a dual rail EMIBucker®

Col

[psw]

Wow...too much information...

This is all exciting stuff, and I feel bad I can't at the moment do a bit of fiddling about myself...

I'll make a few obsevations picked out from the last slew of posts while I got a few moments (now that the kids have gone, and before work begins again...)

First...I believe the compressor preamp I had tried is based on the SSM2116 (a compressor limitor all in one chip)...works well as a clean compressor with up to 15:1 as I recall...

A fizz in harmonic mode sounds familiar. The background distortion is something that really needs fixing though.

Yes...one of my personal aims is clean sustain. This is a unique effect of the sustainer and could lend itself to explorations in areas where feedback is not generally associated (country music for instance) and new sounds. If high gain and fuzz is your thing, you can always add that later, but for me, being able to achieve a clean sound is where it is at!

So...the fizz is something only found in the harmonic mode, with ultra clean settings and mostly on the initial attack of the note...it is a very high frequency distortion that can be effectively filtered out, however, so are the notes in this mode and the amp needs a bit of treble to be able to reproduce effectively notes at least an octave, maybe two, above the note played.

Even mild overdrive will effectively mask it and is useful to compress and bring out these singing harmonics (which are a little too pure being absent of the fundumentals), as is a wha to achieve the treble boost to the amp. Some digital effects may have trouble processing tones this high too...it seems to vary.

My sustainer-strat though is a very unusual guitar now though. It is substantially hollow so has a semi-hollowbody's acoustic resonance. The action lately has gone a bit skew-wiff but I don't have the tools to adjust it at the moment...there is quite a bit of fret buzz in some places. The fizz is still apparent on open strings so it is not a physical/action issue.

I do not think though that it is a typical EMI problem though. I am very familiar with this, and that does not seem to be the problem. I have the circuit in a back cavity, the driver wires through the trem cavity (away from the pickups) and all other pickups and selector switches completely disconnected when in sustainer mode...

I wouldn't discount the need for filtering, I think the upper ranges of the audio spectrum should be filtered out, as I believe my circuit does (but perhaps not enough) as they do not add to the drive in the guitar's range (remember in harmonic mode it is not sending a high frequency signal, but cancelling the fundumental with a reverse image of it that creates this effect), though you still do need to drive the higher harmonics to get the sustain.

The big old 16pin IC in the centre is an LM13700 which is a "Dual Operational Transconductance Amplifier with Linearizing diodes and Buffers" - sounds grand eh ? I'm using it as an automatic gain control (AGC). The gain in the input buffer sets the low level threshold for the AGC which basically gives a big output from a small input and a small output from a big input. Thats the "conditioning" I was talking about

AGC features as one of the building blocks of the commercial units...is this not simply a form of compression?

Fernandes, for instance, in US patent 5,292,999 (but typical of most from Rose to Hoover) lists the circuit blocks as

Preamp=>Phase shift circuit=>AGC=>Limiter=>Current Boost Circuit

In the top down pic, there are three ICs, the left most one is the LM386, the furthest right is an LF358N which is just a standard dual op-amp, I'm using that for buffering the guitar input - one half buffers the input straight to output, the other half sets up the signal for the sustainer circuit with about 3x gain (although that is not final, its one of the variables that gets tweaks a plenty )

As an experiment, I tried feeding the output to my headphone amp directly rather than through the pre-amp buffer. The sound in this configuration was terrible. I wouldn't be surprised at all if the people who have tried this project and got poor results would do much better with a similar twin buffer setup. In fact, I'd urge everyone to consider trying this.

I can't explain why you would have such a result...perhaps it is a product of your "conditioning" circuitry, but it does sound like an EMI related problem. (you did mean that you sent the guitar's signal directly to the headphone amp, not the driver's signal?) If you listen to some of my sound clips, there are plenty of examples of clean sounds (try The Yearning) and my guitar has absolutely no buffering, cheapo single coils (I figured if it would work on this type of instrument, it would work on anything) and long guitar lead directly into the computer via a effects box (cheapo korg AX1G). The guitar should sound exactly the same tonally with the sustainer on or off, except for the sustain...well that is the aim.

The track Beckistan features the modified neck pickup on the first phrase, the bridge pickup on the second phrase (at the start) before I turn on the sustainer and it is all bridge from then on. This will give you an idea of what I am aiming for here and what my guitar can do in this regard.

The idea(l) is that absolutely none of the signal of the driver and it's circuit should get into the guitar signal chain. The EMI problem is one of inductance...it enters not through wires, but through a transformer like interaction between the driver coil and the pickup's. There are various ways to address this issue that I will explore later in this post. I am a fan of trying to address the issue with driver design over compensation circuitry, as with the signal phase / high string response issues. The main way, and most important and influencial approach is simply distance between the two devices, hence the bridge pickup, neck driver configuration. Whether you drive the thing with a clean source, or square wave, the driver signal will always sound distorted if it is transferred magnetically through the pickup to the output via inductance. The ratio of windings of the driver to the pickups (100's:1000's) will ensure massive gain of even the smallest signals. A microphonic driver will also produce signals of it's own unrelated to the guitar signal and resonances that will sound very unpleasant.

BTW...At one stage I had thought I'd come across a way of using this massive gain to create the effect, but once sound clips were posted, people universally could not accept the resulting distortion that accompanied it. Very, very low power was required but the guitars signal was massively increased. Part of this device was that it sat on top of the middle pickup and leached off of the pickups magnetic field...the driver though was an early hex design though. Although similar to my driver/combo that I use now in some ways, the pickup below the driver is completely inactive in sustain mode, while in this device, the interaction of the pickup and the driver(s) were exploited to produce the effect...here's a pic of just one of the designs...

I really like the idea of buffering the whole guitar, I resisted doing this as a lot of people don't want an active guitar. Mine is totally passive, but I think there is a lot of sense to using active buffering once you start putting a circuit and power like this into the guitar. The next step for my strat was a piezio system that I did test but didn't get to build into the thing. I had thought that I could build a preamp for both this and the standard pickups. That said, the buffering is not essential...I would like to see it as an option though if we were to develop a "standard" circuit for this device (probably a dual opamp as you have done).

I seem to be developing an urge to build a dual rail EMIBucker®

Yes...it's catchy, but not at all tested yet. I have had other ideas too.

Some of my hex designs had alternating polarity miniture pole elements to reduce EMI. These ferrite based things could also have acted as a core to a new kind of pickup design. The use of six alternating reverse polarity cores mean't a great reduction in EMI...but not enough for the "grand plan" . Other ways of addressing EMI issues should not be forgotten and I did make quite a bit of headway with the hex designs, but was stimied by life before I could adapt them to the more conventional "thin coils". These included the use of iron impregnated epoxy to encapsulate the devices. This served a number of purposes...it set the whole thing solid, it created it's own "shape" and set the alloy cover to it, it provided magnetic shielding, it helped focus and contain the magnetic fields, it prevented microphonics...and it allowed me to use very thin wired components (I had to solder it all together with winding wire to get them that small) that without it, burnt out with the heat...so they acted as a heat sink too... (an example of this naked "stuff" is seen in the prototype hex design pictured above)

So magnetic shielding is an element to explore, another is the direction of the driver coil. My last generation of hex designs (still secret) used a unique way of manipulating the north and south polarities of dual magnets (the "elements" were balanced between two poles) but also had a unique arrangement to the electromagnetic elements in relation to the guitars pickups.

A different example of this approach was done in early Fernandes designs. Essentially they used a rail-ish type of design but with the coil tipped on it's side (90 degrees to the pickup coils)

This is a picture of one on a strat (notice that it is not oval shaped)...

The patent for this is 5,292,999 and a good picture is available on drawing 5. The centre blade is magnetised north by the two side coils while both outer blades (so, three blades I think) as magnetised south. The side coils do not have the same transformer effect as two coils parrallel to eachother. I really liked the idea...but it is possibly more quirky than anything, it is no longer in production afterall!

Another approach I have seen in the patents (but I don't think ever used) is like a secondary reversewound coil around the driving coil and fed by the opposite signal. This cancels out the EMI but as it has no core (other than the driving coil) it provides a kind of active shielding without adding to the drive. Seems like an extra use of energy, but interesting approach (same guy as above Mr Tumura; Kenji (Osaka, JP)).

My approach to the EMI problem has typically been to make the coil as compact as I can. The idea is to make the area of EMI smaller and to make the driver react with more speed (to avoid the need for complicated circuitry). This is a unique approach as far as I know...simple as it seems to be, it is an approach that didn't seem to have been explored before and has other benifits related to mounting the device and combining it with conventional pickup designs.

There is still room around a 3mm deep driver to surround the coil with some kind of ferro-magnetic material. On my guitar though, it is not needed but is offered as an area to explore. I wonder with the pickup/combo though, the effect on the pickup itself. Interestingly, such a magnetic shield is a feature of Kinmans stacked humbucker designs, so it may not be detrimental...

So...I don't know why other's aren't getting quite the results I have had. I think primal (till the circuit blew) had similar results to me and even had less of a "pop" (still very annoying and needs to be fixed on mine). But col's circuitry for instance is very different to my approach in many ways. I am not sure if it was mentioned yet, but I did have problems with the neck pickup (or any other) dispite the thing not being selected as it was still connected to ground (hence the need to disconnect the whole thing completely, not just lift or short it out) and if you are testing the driver above, or near such a pickup, this may be causing some of these problems. Interestingly, in my "sustain box" experiments with a Les Paul where the signal was taken from the output jack (after the controls) and the humbuckers were still connected (though the neck pickup no selected of course) it did not suffer to quite the same degree...hmmm

Anyway...all great stuff, and a joy to read about and hope I have given you some food for thought in return... pete

ahhh...back to the monster posts that this thread has always been renown for!!!!!

Oh...and it occurs to me that recent visitors here may not be aware of the Sustainer Sounds Thread from which I linked the tracks in this post. For pictures of my sustainer-strat for those not familiar with it, there are some at G'mikes sustainer tutorial thread

Edited September 16, 2006 by psw

[axe_2_grind]

OK, so I recycled an old stock single-coil using 30 guage wire (approx. 0.25mm or so)

I tried to keep it as thin a possible, but I am lacking the proper clamps and glues and stuff (need to hit up the hardware store), not to mention I was perhaps a little impatient due to the fact that I just got the wire. I used some regular Elmer's wood glue (which was a bit messy) and I did notice that though the edges were tight, the long parts kinda sagged. Anyway, when I hooked up the multimeter, the resistance fluctuated rather wildly. And when I hooked up the driver to the circuit, my LM386 started heating up and (of course) no sustain.

Of course, as previously mentioned, I did not use any of the prescribed methods for winding a driver, I just tried to reuse an old pickup bobbin with 6 pole pieces, and just lowered the top down. I guess I'll just have to be patient and try again when I get the appropriate materials.

[col]

OK, so I recycled an old stock single-coil using 30 guage wire (approx. 0.25mm or so)

I tried to keep it as thin a possible, but I am lacking the proper clamps and glues and stuff (need to hit up the hardware store), not to mention I was perhaps a little impatient due to the fact that I just got the wire. I used some regular Elmer's wood glue (which was a bit messy) and I did notice that though the edges were tight, the long parts kinda sagged. Anyway, when I hooked up the multimeter, the resistance fluctuated rather wildly. And when I hooked up the driver to the circuit, my LM386 started heating up and (of course) no sustain.

Of course, as previously mentioned, I did not use any of the prescribed methods for winding a driver, I just tried to reuse an old pickup bobbin with 6 pole pieces, and just lowered the top down. I guess I'll just have to be patient and try again when I get the appropriate materials.

The 'prescribed' methods are just suggested methods that have worked for others - not necessarily the best or only approaches.

Is "elmers glue" the same as PVA? we don't have 'elmers' here in Scotland, and when I checked their website there are a multitude of different elmers glues. If the one you used is water based, or otherwise conductive, it could cause problems with the checking resistance while winding while potting approach ?

If your LM386 is overheating, that could be because your driver is not at or near 8ohms, if for some reason (e.g. a short somewhere in the winding) your driver has a much lower imedence, prolonged use could easily damage your driver circuit. Basically, If your driver doesn't measure a steady resistance of somewhere around 7 or 8 ohms (definately no less than 4ohms), then you really shouldn't hook it up to the circuit.

A quick rundown of my driver making experience:

My first attempt was using a 60mm x 4mm x 4mm core rounded at the ends. I used the black plastic from inside a cd case for bobbin ends as suggested. I glued the whole lot together with epoxy, lined the gap with insulating tape and started to wind. At about 5 ohms, the whole lot began to seperate - turns out you can't stick that black plastic effectively with epoxy !.

For the second attempt, I used a slightly deeper core @ 6mm ( x 60 x 2 ) so there was a bit of overlap for better glueing, and I glued the bobbin ends to the core with superGlue ( Cyanoacrylate ) This was much better, the superGlue seems to work with that cheapo plastic. With my next effort, I will coat the whole inside of each bobbin end with superglue in the hope that the epoxy I use for potting (slow drying araldite precision ) will adhere to it better (this might help prevent the vibration noise I was getting when driven hard). The only real difficulty on the second effort was that I started running out of room for wire - I recommend making the bobbin ends bigger than you think you will need, at least for your first prototype . If you are using epoxy for potting, buy some disposable rubber gloves - I bought a pack of 10, but the epoxy doesn't stick to them, so I'll use the first pair again. Also, try to rig up anything else you might need so you can use it with minimum fuss - you don't want to get epoxy all over your multimeter.

One thing I would really like to know is how people are winding on a bobbin that is clamped together - is it a huge clamp that they feed the reel of wire through ?? Maybe I have misunderstood something...

Also, do cardboard bobbin ends stuck down with double sided tape really hold securely under the pressure of winding ? I would like to try a 'bobbinless' driver, but from my initial attempt, I'm dubious that cardboard and tape would be enough.

Col

[col]

...

First...I believe the compressor preamp I had tried is based on the SSM2116 (a compressor limitor all in one chip)...works well as a clean compressor with up to 15:1 as I recall...

Seems like they've stopped manufacturing all the non-surface mount SSM series of compressors

...I do not think though that it is a typical EMI problem though. I am very familiar with this, and that does not seem to be the problem. I have the circuit in a back cavity, the driver wires through the trem cavity (away from the pickups) and all other pickups and selector switches completely disconnected when in sustainer mode...

It seems like it's not the driver wiring (I've still to discover what terrible problems await when I try to install the wiring in the guitar cavity) but as suggested, the strings - the Flux using the guitar strings as a conduit between the driver and the pickup. Of course, It might just simply be the interaction between the fields of the driver and pickup, the strings may have little to do with it. ( have to experiment next time I have the strings off the guitar).

Out of interest, what is the distance between your driver core and your pickup poles ? On my setup, there are about 11cm between my driver core and the nearest humbucker poles.

Another variable here is the sensitivity of the pickup - my DuncanJB is very 'hot'. It's impedence is something ridiculous like 16k, and it probably has a huge chunky magnet to match.

I wouldn't discount the need for filtering, I think the upper ranges of the audio spectrum should be filtered out, as I believe my circuit does (but perhaps not enough) as they do not add to the drive in the guitar's range (remember in harmonic mode it is not sending a high frequency signal, but cancelling the fundumental with a reverse image of it that creates this effect), though you still do need to drive the higher harmonics to get the sustain.

Mmmm.. more conditioning

I agree, filtering out the inaudible signal makes sense - reduce noise, heat, increase stability, reduce battery consumption etc, but it may not have any noticable impact on EMI generated fuzz or fizz.

The big old 16pin IC in the centre is an LM13700 which is a "Dual Operational Transconductance Amplifier with Linearizing diodes and Buffers" - sounds grand eh ? I'm using it as an automatic gain control (AGC). The gain in the input buffer sets the low level threshold for the AGC which basically gives a big output from a small input and a small output from a big input. Thats the "conditioning" I was talking about

AGC features as one of the building blocks of the commercial units...is this not simply a form of compression?

compression is a form of AGC.

Fernandes, for instance, in US patent 5,292,999 (but typical of most from Rose to Hoover) lists the circuit blocks as

Preamp=>Phase shift circuit=>AGC=>Limiter=>Current Boost Circuit

I wonder what the Phase shift circuit is supposed to do, and how effective it is. I have played with a few Phase shifter circuits, and so far found nothing particularly useful - certainly not worth the extra cost, board space and power drain. Maybe they're doing stuff I don't understand properly - say aligning the phase of the voltage and current or some such, that has some enormous benefit that we are missing out on - who knows

As an experiment, I tried feeding the output to my headphone amp directly rather than through the pre-amp buffer. The sound in this configuration was terrible. I wouldn't be surprised at all if the people who have tried this project and got poor results would do much better with a similar twin buffer setup. In fact, I'd urge everyone to consider trying this.

I can't explain why you would have such a result...perhaps it is a product of your "conditioning" circuitry, but it does sound like an EMI related problem. (you did mean that you sent the guitar's signal directly to the headphone amp, not the driver's signal?) If you listen to some of my sound clips, there are plenty of examples of clean sounds (try The Yearning) and my guitar has absolutely no buffering, cheapo single coils (I figured if it would work on this type of instrument, it would work on anything) and long guitar lead directly into the computer via a effects box (cheapo korg AX1G). The guitar should sound exactly the same tonally with the sustainer on or off, except for the sustain...well that is the aim.

I'm not sure exactly - I wasn't using the wrong signal, and it's nothing to so with the 'conditioning' (it's soft and gentle and kind to your tone). Might be that the buffer is preventing some of the nasties caused by a few feet of unshielded cable, dodgy bread board connectors etc. Also possible is that putting a 3x gain on the buffer feeding the driver is loading the signal... I might be able to get better unbuffered feed-through by using the redundant buffer prior to the 3x gain stage.... something to try on a rainy day.

...The EMI problem is one of inductance...it enters not through wires, but through a transformer like interaction between the driver coil and the pickup's. .... The ratio of windings of the driver to the pickups (100's:1000's) will ensure massive gain of even the smallest signals.

This is a VERY interesting point that I had not considered - It makes total sense though - and means that any interaction at all between driver and pickup will be clipped and sound fuzzy/fizzy which is what we are experiencing.

Another approach I have seen in the patents (but I don't think ever used) is like a secondary reversewound coil around the driving coil and fed by the opposite signal. This cancels out the EMI but as it has no core (other than the driving coil) it provides a kind of active shielding without adding to the drive.

Seems something like a driver version of the 'air core' humbucker design (Alembic iirc) also related to Kinmans stuff ie. a secondary coil that is specifically designed for noise cancellation.

I wonder If it there might be some way of building a secondary coil (with a core) that can function as an EMI fuzz-away when in sustainer mode, and as a low impedence active pickup in non-sustain mode... probably not...

cheers, Col

[axe_2_grind]

Elmer's is a fairly common brand in the U.S. , and they make alot of different glues for different purposes. The Wood glue I am using I believe is water based, hence alot of mess when winding. I'll see if I can find some PVA. Thanks for letting us know about the super glue thing. I've been a little afraid of handling epoxy. (don't want to glue my fingers together!!)

Like I said, I need to get the right materials before I attempt to do this again.

[col]

Elmer's is a fairly common brand in the U.S. , and they make alot of different glues for different purposes. The Wood glue I am using I believe is water based, hence alot of mess when winding. I'll see if I can find some PVA. Thanks for letting us know about the super glue thing. I've been a little afraid of handling epoxy. (don't want to glue my fingers together!!)

Like I said, I need to get the right materials before I attempt to do this again.

Epoxy won't glue your fingers together - but it's nasty stuff - if you get it on your skin wash it off asap. If you're going to use epoxy, make sure its not a quick setting one. The one I used stays usable for about 2 hours.

Superglue bonds skin instantly - thats the one to use if you want to glue your fingers together

Col

[psw]

The Wood glue I am using I believe is water based, hence alot of mess when winding. I'll see if I can find some PVA.

PVA washes out in water?! It is probably ok, I don't think that is the problem. The wire does need to be insulated wire (a clear enamel coating) so as not to short as the potting is not the insulator and the wire is protected from the water based glues, or whatever. Could this be your problem for variable resistance of the coil? I did test a few different guages and definitely found that the 0.2mm to have the best results with this design...

Out of interest, what is the distance between your driver core and your pickup poles ? On my setup, there are about 11cm between my driver core and the nearest humbucker poles.

Actually 10cm on the bass side, 11cm on the treble due to the angled bridge pickup...so perhaps actaully closer...hmmm

Certainly, my bridge pickup is a bit on the weak side. It is a cheap stacked humbucker replacement with alnico poles. A hot pickup would be more sensitive and so be more suseptable to this I suppose. I did try a version of this on my Les Paul with humbuckers (not that hot, 60's Gibsons) without problems.

Here is a drawing I found somewhere that shows the interaction between pickups, maybe some food for thought...

It seems like it's not the driver wiring (I've still to discover what terrible problems await when I try to install the wiring in the guitar cavity) but as suggested, the strings - the Flux using the guitar strings as a conduit between the driver and the pickup.

The aim is to have no interaction, but some must get into the strings, being magnetic... As for installation, I found that the device worked better installed...like you it worked better further away from the pickups, but holding the driver above the strings "sprays" the EMI down on the guitar, while mounting it under the strings, on the same plane as the pickups, helps a lot. You also need to do some adjustment of the driver height and pickup height for best effect.

The driver must be pretty close to the strings. I think that in doing so, some of the driver's signal is reduced by the opposite side of the coil's inverted signal, and so helps contain the amount that runs through the strings to an area around the driver. I did try experiments with metal "fins" in a U coming up from the underside of the coil to create a kind of magnetic sheilding effect. I didn't find a lot of difference in the little work I did though on this type of driver.

I wonder what the Phase shift circuit is supposed to do, and how effective it is.

The Floyd Rose Patent explains it schematically a little better than some. It appears to be a series of capacitors to creat a delay that are switched in according to the frequency of a signal. The idea is to compensate for the delay between the vibration of the signal and the circuit/driver's reaction time...hmmm

I worried about this for some time, but then Lovecraft suggested that I ignore the patents and concentrate on the driver. This was very influential advice on my work (thanks LK). Such circuitry itself would add delay and the whole thing seems a little dodgy in some respects...I thought a better driver design would eliminate the need for such circuitry...plus I didn't have the know how to build the circuit anyway.

That said, Dizzy from another forum (Aron's Stombox), who I posted on the Sustainer Sounds Thread, made his mid driver with a copy of the commercial units approach with just such circuitry (secret, I'm afraid) and a bi-lateral driver. (Great sounding clip....production values!!) How important it is, I don't know, but my thin driver appears to do the job with simple amplification....I have reasons why I think this is so...

This is a VERY interesting point that I had not considered - It makes total sense though - and means that any interaction at all between driver and pickup will be clipped and sound fuzzy/fizzy which is what we are experiencing.

Yes it is!!! And the key to why I had to completely disconnect the wiring of the other pickups...even just one end of these pickups connected, or even shorted, resulted in some kind of distorted signal getting into the guitar's signal chain. The leads from the driver also need to be kept fairly short and well away from the pickups as these are an extention of the coil and will emit EMI. Pickups are sensitive devices...

One thing I would really like to know is how people are winding on a bobbin that is clamped together - is it a huge clamp that they feed the reel of wire through ?? Maybe I have misunderstood something...

Also, do cardboard bobbin ends stuck down with double sided tape really hold securely under the pressure of winding ? I would like to try a 'bobbinless' driver, but from my initial attempt, I'm dubious that cardboard and tape would be enough.

Tim's/onelastgoodbye's bobbinless driver is still a favorite. He has a magnet as the core and was able to use plastic protected metal plates stuck to it to wind it on while the epoxy set, then removed them...very special!

My driver used very thin plastic (glue would not stick) for the top piece, the pickup top is the bottom of the driver bobbin). I bent down the edges of the slot and taped it on and reinforced that with some cardboard above. The clamps were applied after to keep the whole thing tight while the glue dried! No, you cant wind it with the clamp on! It still tried to get out (especially on the ends) but once the glue set, it seems to be ok! I tried to mold an epoxy bobbin but with little success, so the art's and crafts approach is perhas better. My first one was thin card and was a little better. With both of these the bobbin material was thin enough that it could be left oversized and cut to shape after winding with scissors.

Another approach could be to wind on a plastic, removable dummy core and, once the epoxy is set, replace it with a blade. This way you could bolt removable bobbin pieces through this leaving just a solid coil. Or, you could use a row of small bolts instead of a blade, and leave them in...just some thoughts, never been tried...

As this is DIY and experimental, it is probably not worth going to overboard, unless a standard design were to come forward and you were going to use the jigs to make a few of them!

Anyway...great posts as always...and Scotland, this thread really has spanned the globe and the time zones. Early monday morning here, and I got to rush out...thanks for the diversion... pete , etc

[psw]

Ah...I found it...

This thread can be tricky to navigate but page 85 was a particularly good one it seems...(clicking on the little pink arrow in the quote below will take you directly to this page, BTW)

Below is some links to the "Aussie Compressor" that I am going to look into...(right now!)

Not to hijack the thread or anything, but this might be important. Pete and I have discussed gain control in theoretical terms several times, and rejected it as too complex and fiddly, but this idea (from , another proud Aussie ) might just make it a lot easier. Brett's come up with an ultra-simple compressor based on the 386:

Brett's Aussie Compressor

Since it works using a super simple feedback sidechain and a single MOSFET, it should be fairly trivial to integrate into Pete's basic power amp. For more info, see this thread over at Aron's DIY Stompbox forum. Unfortunately, I don't have the time to pursue this right now, but it's too good an idea not to share with the group - it could (theoretically) help even out the response curve over the entire frequency range, and make for much smoother sustain. Now, you maniacs take it and run with it - both BS170 and 2N7000 MOSFETs should be easy to find and fairly cheap, and the rest is stock components. If anybody gets truly superb results, please let us know!

Might be another way to achieve the kind of effect that col is developing.

Also on page 85 is a complete wiring of my guitars internal electronics, and a diagram of the method used to build Tim's bobbinless driver pictured in my post above... pete

Hmmm...links are down...stand by for more information.... p

OK...I found some more links....

The link to this thread is...Oz Compressor VERO Layout...there are a few other threads, mods and troubleshooting threads over there if you use the search function...d'uh... p

Be interested to here what people think, would imagine you would still require some kind of load prevention buffer as I did have problems with the LM386 circuits without. This is not a recomended circuit for the sustainer BTW, but something that could be adapted perhaps for the purpose!

Edited September 18, 2006 by psw

[psw]

Just another thing to add...while surfing around I found this PCB Layout creator freeware...also at this site under projects are layouts of the RUby and a number of other guitar related projects, perhaps one day a circuit from us could be launched into cyperspace...looks interesting, have downloaded software and will be playing around with it soon. I think the above layout was done with it... pete

Now that is really great...highly recomend this little free program...it is especially good if you download a bunch of the layouts...has various Ruby's and the Fetzer Valve circuit layouts as well as the better known DIY stompbox projects and the like!

Edited September 18, 2006 by psw

[onelastgoodbye]

Some thoughts on the bobbinless driver...don't bother with the plastic stuck to the metal; just use the metal by itself. Regular epoxy won't stick to it, unless you use jbweld or something. I didn't use clamps, the magnetic force of the err..magnet held the (temporary) metal bobbins in place. You have to be careful while winding, as the whole assembly can shift and slide if you pull the wire too hard.

My jiggamathingie makes things a lot easier because the bobbins clamp to the core. The core extens 2mm from the coil at the top, and about 5mm from the bottom, which equals the height of the ceramic magnet that's supposed to sit up against it. In fact, my humbucking driver should look pretty much like the diagram 3 post's up (interaction between pickups), only without the bobbins.

I wound my first coil on the jiggamathingie yesterday, there's some good news and some bad.

The good (1): the jig worked like a charm.

The bad (1): I used PVA because my scales mess up when I try to mix small batches of epoxy (the digital scale isn't all that accurate so I can't get ratio's close enough, resulting in a faulty mixture which refuses to harden). Only to find out the PVA takes verrrry long to harden (the air can't really get to it).

The bad (2): a 4 ohms coil is a lot smaller than I expected, so I'll have to adapt the jig.

The good (2): a 4 ohms coil is a lot smaller than I expected, so the driver can be a lot smaller!

Ultimately, If one could find a suitable internal magnet, it could be as small as 56 mm long x 10 mm wide x 2 mm thick (yeah, 2mm!). I'm thinking flexible magnetic sheets ? (actually even sourced some in 2 mm thickness)...

[psw]

Some thoughts on the bobbinless driver...don't bother with the plastic stuck to the metal; just use the metal by itself. Regular epoxy won't stick to it, unless you use jbweld or something. I didn't use clamps, the magnetic force of the err..magnet held the (temporary) metal bobbins in place. You have to be careful while winding, as the whole assembly can shift and slide if you pull the wire too hard.

My jiggamathingie makes things a lot easier because the bobbins clamp to the core. The core extens 2mm from the coil at the top, and about 5mm from the bottom, which equals the height of the ceramic magnet that's supposed to sit up against it. In fact, my humbucking driver should look pretty much like the diagram 3 post's up (interaction between pickups), only without the bobbins.

I wound my first coil on the jiggamathingie yesterday, there's some good news and some bad.

The good (1): the jig worked like a charm.

The bad (1): I used PVA because my scales mess up when I try to mix small batches of epoxy (the digital scale isn't all that accurate so I can't get ratio's close enough, resulting in a faulty mixture which refuses to harden). Only to find out the PVA takes verrrry long to harden (the air can't really get to it).

The bad (2): a 4 ohms coil is a lot smaller than I expected, so I'll have to adapt the jig.

The good (2): a 4 ohms coil is a lot smaller than I expected, so the driver can be a lot smaller!

Ultimately, If one could find a suitable internal magnet, it could be as small as 56 mm long x 10 mm wide x 2 mm thick (yeah, 2mm!). I'm thinking flexible magnetic sheets ? (actually even sourced some in 2 mm thickness)...

I love it when a plan comes together...even if there are some unexpected surprises...

I wouldn't have used PVA for a bobbinless driver. I will effectively inhibit vibrations, but I fear it will never dry hard enoough...at least not like epoxy. If all else fails, you could try coating the whole thing completely in epoxy as an outer shell...if it doesn't come appart of it's own accord...

I suspected the good (2) may happen...

I'm thinking flexible magnetic sheets ? (actually even sourced some in 2 mm thickness)...

Now be careful...I had thought of this too! As it happens, the place that I used to source my magnets is pretty close to where I am staying (a little shop in amoungst the houses near a railway line as it happens). Before I returned to work, I was doing some long, soul searching walks, and there before me was Aussie Magnets! (I'd only mail ordered before).

Anyway...they have a metal wall with all these magnets and stuff on there, including flexible magnetic strips. Now, I say be carefull because, most (like they print business fridge magnets on) are polarised in north and south strips along the surface. I was able to find one that seemed promising, but it was very expensive because you had to by an industrial roll of it (A$75 as I recall) and I couldn't get a sample...ohhh! So...if it is that you want one side north, the other south....test it to be sure, cause these guys didn't know. Fortunately, I had plenty of magnets around so that I could run them over the surface and feel the alternating polarities, etc.

I am not sure of the power of these things also...They are substanially rubber/plastic, impregnated with some kind of magnetised particles. I did however see somd patents of a pickup design that proposed their use...nothing came of it though.

I know you like the idea of internal magnets...but they are very hard to source. I was thinking when I proposed the idea of winding the coil and placing the core in after in a post above, that you could use the stuff I have/had (iron impregnated epoxy) to squeeze into the coil space (it is non conductive) making it even more solid. I would propose an external magnet below this.

At one time I was thinking of a row of very small rare earth magnets (2mmx3mm discs) that I used in the hex designs. The combination of the dozen magnets to simulate the blade would be far too powerful, even though the right size.

Although I did use 12 mags of this type in the hex designs, they alternated in polarity and were spaced between the strings and no pole pointing at them. The result is what I called a "balanced field" that could be 'unbalanced' with my minature electromagnetic elements, pulling and pushing at the string. Such radical ideas will only work with carefull alignment, and are so not that practical...though they do work!

So...in a shop full of magnets of all types and sizes, I didn't see anything that quite fit the bill.

I did think also of using the flexible magnets as some kind of magnetic shield though...that could be of some promise...save up those fridge magnet cards perhaps.

Good luck on the magnetic quests and post some more picks of how you operate the jigamathingie if and when you can...intrigued... pete

[col]

edit: oops, didn't see petes post, and repeated some of what he said

Some thoughts on the bobbinless driver...don't bother with the plastic stuck to the metal; just use the metal by itself. Regular epoxy won't stick to it, unless you use jbweld or something. I didn't use clamps, the magnetic force of the err..magnet held the (temporary) metal bobbins in place. You have to be careful while winding, as the whole assembly can shift and slide if you pull the wire too hard.

My jiggamathingie makes things a lot easier because the bobbins clamp to the core. The core extens 2mm from the coil at the top, and about 5mm from the bottom, which equals the height of the ceramic magnet that's supposed to sit up against it. In fact, my humbucking driver should look pretty much like the diagram 3 post's up (interaction between pickups), only without the bobbins.

How do you wind the coil with the clamp in place ?

Is the clamp huge with a big 'throat' so you can feed the reel of wire through each time ?

Please more details

I wound my first coil on the jiggamathingie yesterday, there's some good news and some bad.

The good (1): the jig worked like a charm.

The bad (1): I used PVA because my scales mess up when I try to mix small batches of epoxy (the digital scale isn't all that accurate so I can't get ratio's close enough, resulting in a faulty mixture which refuses to harden). Only to find out the PVA takes verrrry long to harden (the air can't really get to it).

The bad (2): a 4 ohms coil is a lot smaller than I expected, so I'll have to adapt the jig.

use this to give you an estimate !

The good (2): a 4 ohms coil is a lot smaller than I expected, so the driver can be a lot smaller!

Ultimately, If one could find a suitable internal magnet, it could be as small as 56 mm long x 10 mm wide x 2 mm thick (yeah, 2mm!). I'm thinking flexible magnetic sheets ? (actually even sourced some in 2 mm thickness)...

The only flexi magnetic sheets I've found have a wierd arrangement of poles - alternating stripes of NSNSNS... accross the surface. weak magnet strength as well. If you've found some sheet that has a usable polarization, where ?

cheers

Col

Edited September 18, 2006 by col

[spazzyone]

Ill be posting some new soundclips in the next day or so

i have been getting overwhelming results using guitartrax pro

the tremolo effect from the rp6 is incredible

and no effects are in the driver signal chain

so the quality is very good compared to my last vid

and to Pete...so far so good on reliability as i have left my setup on

for over 24 hours last week with no probblems

untill then see ya's

[psw]

edit: oops, didn't see petes post, and repeated some of what he said

This whole thread is filled with repeats, don't you just hate it when it turns another page the whole thing could be condensed to 20 posts in the end...the best of the sustainer thread, thread, maybe...

BTW...any thoughts on the LM386 compressor circuit?

How do you wind the coil with the clamp in place ?

Is the clamp huge with a big 'throat' so you can feed the reel of wire through each time ?

Please more details

from Tim...

I didn't use clamps, the magnetic force of the err..magnet held the (temporary) metal bobbins in place. You have to be careful while winding, as the whole assembly can shift and slide if you pull the wire too hard.

yet

My jiggamathingie makes things a lot easier because the bobbins clamp to the core.

inscrutible...

from me...

No, you cant wind it with the clamp on!

I was able to make one with cardboard, reinforced with plates that were stuck on with tape and removed later and it stayed together really well...the paper card absorbed the PVA and stuck tight...

the one on the right, the other is a single coil wound as a driver, this worked well too despite not being so thin...hmmm The little silver things on the bottom are rare earth magnets (3x2mm), compare with the ceramic on the other of similar power. The problem with them for this application is that the field is so condensed due to their small size...

so far so good on reliability as i have left my setup on

for over 24 hours last week with no probblems

I wondered where you'd got to man...24 hours...good to see you are spending the time playing and recording... ...you got something going on there, but I don't know what it is!

BTW, if people have a mind to, that Layout program is a snap to use and comes already equiped with the Ruby and Fetzer circuits. It will create PCB, Perf and Vero board layouts and you don't have to waste a lot of time with getting really specific about the components from a library. It wont draw schematics, but for converting a schematic or perfboard experiment to a board that can be easily posted, it's tops...plus it can be saved and posted on here if you have something to share or need checking. The veroboard layout on the previous page was done with it for instance....

got to run...5am and time to get to work!... pete

[col]

BTW...any thoughts on the LM386 compressor circuit?

I've checked that one out before - seems like an interesting circuit as a guitar compressor - simple is good . However, I'm not sure how useful it would be to us. I think it would require some major modifications before it could drive an 8ohm load ?

Also, it really only mashes the louder signal, doesn't seem to bring the quiet parts up in volume ?

If it can be made to drive an 8ohm load and is fed by a gain stage, it should work ok for our purposes. I'm not sure how easy the first part of that 'if' is though - might make the thing a whole lot more complicated - anyone ?

How do you wind the coil with the clamp in place ?

Is the clamp huge with a big 'throat' so you can feed the reel of wire through each time ?

Please more details

from Tim...

I didn't use clamps, the magnetic force of the err..magnet held the (temporary) metal bobbins in place. You have to be careful while winding, as the whole assembly can shift and slide if you pull the wire too hard.

yet

My jiggamathingie makes things a lot easier because the bobbins clamp to the core.

inscrutible...

I think I just understood - instead of one clamp with jaws pressing on the top and bottom bobbin ends, each bobbin end IS a clamp that bites the 1 or 2 mm of the core that sticks out (of course, you can't then have the coil flush with the core), Then when you're done and the glue is dry, you just remove the clamps and hey presto... is that about right onelastgoodbye?

What about a jig where even the core is part of the jig, so a bobbinless, coreless potted coil can be constructed, this could then be tried with a veriety of different cores...

I wish I had a machine shop with drill press, lathe, forge, welding kit, computer controlled 3D milling machine etc. Just so that once in a blue moon when I get a crazy idea, I can prove its crazy

Ill be posting some new soundclips in the next day or so

i have been getting overwhelming results using guitartrax pro

the tremolo effect from the rp6 is incredible

and no effects are in the driver signal chain

so the quality is very good compared to my last vid

and to Pete...so far so good on reliability as i have left my setup on

for over 24 hours last week with no probblems

untill then see ya's

Cool, looking forward to these clips, are any of them using a nice crystal clean tone ?

cheers, Col