DIY SUSTAINER...hear for yourself

[mathflan]

Hi,

I have some questions before building this great projet from PSW:

1) what is the diameter of the enamel wire I have to take ( in millimeter)
2) what kind of metal element I have to take to put the wire.
3) What in fact the best Preamp circuit for the driver? ruby, feltzer?
4) DO I have to do some modifications on the preamp circuit?

thanks

[psw]

G'day mathflan

I have some questions before building this great project from PSW:

1) what is the diameter of the enamel wire I have to take ( in millimeter)
2) what kind of metal element I have to take to put the wire.
3) What in fact the best Preamp circuit for the driver? ruby, feltzer?
4) DO I have to do some modifications on the preamp circuit?

Sorry for not replying sooner but....hey, read about it here along with this type of info and more...there's been quite a few people working on them with pics schem's and even circuit board plans...check it out at...
http://projectguitar.ibforums.com/index.php?showtopic=7512&pid=239005&st=1065&#entry239005
also at the tutorial thing...
http://projectguitar.ibforums.com/index.php?showtopic=16984
there's also some sound clips...
http://projectguitar.ibforums.com/index.php?showtopic=17852

Anyway, in short, your questions....
1)...0.2mm
2)...I used ordinary hardware soft steel, 3mm thick...basically anything magnetic...could layer thinner sheets, etc
3)...fetzer/ruby hybrid circuit is simple, small and has shown to work...I used a generic LM386 amp and preamp from kits...there's a nice circuit board graphic up just the other day...
4)...the preamp is really just there to stop loading and add gain for the amp to work efficiently. the stock Ruby doesn't really cut it but the fetzer/ruby has a greater gain ratio so should be fine...

I see you've already raised these questions over at the PG forum (page 69) but other's may be interested in hearing the answers...

best of luck...

pete

[psw]

The sustainer thread at PG doesn't stop. More of these things are being made and there are pics to show how easy it is. The sustainer thread has now had over 30,500 visits and over 1000 contributions. It's 74 pages long so don't try and read it all, just check out the last couple of pages or the tutorial and stuff linked in the above post. Here is a couple of pics of the latest go at making one from member Randy:


Made from a steel core, CD cases and some tiny Rare earth Magnets. There are lots of different variations having been made of late and some interesting discussions going on about how to improve them or what's a better theoretical device. Feel free to join in!

http://projectguitar.ibforums.com/index.php?showtopic=7512&pid=243063&st=1095&#entry243063

or ask questions here

psw

[KMS]

Maybe you have already tried this................

Don't turn the unit completely off..

I am assuming that your using true bypass and the pop is an arch in the switch from either the the clean signal or the sustainer signal or both.  Defeating the circuit has worked for me on every project to date where I had a pop in the bypass switching. 

And LED/LDR will also stop the pop (for sure) if an extra 300 or so ohms in your guitar signal won't screw things up for you.  You need one LED/LDR for each throw on your switch, then there is no physical contact made to create an arch (pop).  Also note that an LED/LDR does not completely turn either signal off when switching rather the off mode signals are decreased by say 1M but it is still an operating circuit.

You might be able to do this with a 4M resistor in-line with the signal coming out of an opam just before the pot.  I turn my mixer channels off like that with no pop at all. 

I have used both LED/LDR and big in-line resistors...both do sort of the same thing, which is not allow a charge to build up since the circuit is never completely off or completely bypassed.

[psw]

Thanks a lot for that. I have tried a few things similar without a lot of success. Not very many people have done the full install yet. When I did (true it was a complex wired cheapo strat involving single coils) I had to use a large 4TDT switch to bypass select and power on. The pop is only when switching on. I have a feeling if the little preamp stage were always on the guitars earth would always be connected to power and this might help the situation...this stage could even be used to make the guitar active like the tillman preamp.

There are also more modern and efficient popless chips than the simple LM386 ruby type circuit that would address the problem. The big switch was required because I needed to lift both ends of the coils on the middle and neck pickup to avoid EMI noise from the driver. Lifting only one end was not enough. A good humbucker may not have been as suseptible and of course the strat has three pickups instead of the more common two.

Thanks for your interest though...the project does work as is but has to be adapted to the guitar in question a little. The only difference between this and a stompbox project is that it has to be built into and involves some modifications to the guitar and you have to litterally make the driver. This seems to be the stumbling block for most. Dispite it looking like a pickup and similar in principle, it is much easier to make than that. The wire is not fragile and can easily be wound by hand in about 10 minutes as it is only about 200 turns (rather than 1000's for a pickup with hair thin, hard to get wire). The example above is this fellows first go and experience of doing anything like this...and it looks the business.

The ultimate thing would be to have a dedicated sustainer guitar with only one bridge pickup and one neck driver. The switching, installation and everything would be much, much easier and potentially the space left by the removal of a neck Humbucker could easily accomodate both the driver (which is about 4mm thick) and all the circuitry, only requiring room to be found for the battery somewhere accessable.

I'd love someone to help devise a better circuit than the fetzer/ruby and potentially some clever fellow could devise some simple electronic switching to help with the bypass and wiring complexities too. Unfortunately a little out of my expertise. Thanks for the input...much appreciated...psw

[petemoore]

  Ok...perhaps some compression would help 'even' the strings out, making the small strings work better?
  Something like put clipping diodes before the amp part [of course I don't that'll work with JFet Fetzer, perhaps splice someother clipper/booster in the front.
  Only testing or theory can say whether distortion 'messes up' the string activation device, I believe if applied 'correctly' this could make the small and large strings respond less differently, keeping the overall volume output to the driver more 'same like' by keeping the amplitude 'evened over'.

[psw]

Thanks petemoore.

The thread is so long because in the early days I was very consious that all the patented circuits had some kind of phase correction circuitry (although it's nature was not specific). This was to compensate for the lead and lag times between the driver's response (particularly on high strings/frequencies) and the signal. With fast movements it takes a while for the driver to magnetise in one direction, then change states to the other. With a fast moving string it may physically change states before the driver can respond...in other words be out of phase. With these drivers the signals from all the strings are fed into the one device. In truth the signal is clipped to a large degree by the circuit and the driver's inefficiencies. I have tried diodes, led's a compressor limiter and fuzz style preamps. There was not a lot of difference really I suspect because the circuit is already compressed and clipped as I say. (BTW i used a different pre&amp than the fetzer/ruby). Another idea was to cut of one side of the wave to give the driver time to recover between states.

So...it is not so much the relative "volume" of notes but the phase. The placement of the driver, the width of the "blade" core and such have a far bigger influence. I spent a lot of time on various Hex drivers...individual miniture drivers for each string. But they were still driven from a monophonic source. The current crop of designs are much easier to DIY. The concept is that the "thin driver" (3-4mm thick) and possibly a "narrow core", responds quicker to changes in states thereby allowing the thing to work without complicated phase compensation.

BTW...the device is polyphonic but favours the lowest tones. To sustain higher strings the lower ones must be dampened. On a chord, the bass note is the one that will feed back generally, the other's sound as normal. Now, this could perhaps be improved with better cleaner circuitry or filtering. The device is a little key biased too. If placed where the mid pickup generally is (24th fret) it will favour notes from the key of e minor (the open strings and their harmonic series) in the middle range of frequencies. There was recently some discussion of whether a wider core as in the above pictures would help in this regard. But I suspect the "narrow core theory" will prevail for other reasons.

Anyway...thanks for your interest...psw

[Jaicen_solo]

Going on what I know about vibrating strings (I studied them at degree level ), and Ideal amplifier in this situation would output a square wave signal, swinging from rail to rail. Some low pass filtering may be needed to tame some of the harmonics this would produce, but this would certainly be the most efficient driver.
To make the Ruby swing to rails, you could possibly leave pins 1 & 8 unconnected to allow full gain to be applied. A high impedance buffer will probably still be required to avoid loading the pickups. Cross talk from the driver to a neck pickup can be avoided by simply grounding the hot end of the coil. Power should not be turned on and off whilst the unit is in operation, as this will cause inductive spikes, hence the popping. There is simply no way around this.
That's all I can think of off the top of my head, which may already have been covered but I haven't had time to read that big ass thread!

[KMS]

Thanks a lot for that. I have tried a few things similar without a lot of success. Not very many people have done the full install yet. When I did (true it was a complex wired cheapo strat involving single coils) I had to use a large 4TDT switch to bypass select and power on. The pop is only when switching on. I have a feeling if the little preamp stage were always on the guitars earth would always be connected to power and this might help the situation...this stage could even be used to make the guitar active like the tillman preamp.

There are also more modern and efficient popless chips than the simple LM386 ruby type circuit that would address the problem. The big switch was required because I needed to lift both ends of the coils on the middle and neck pickup to avoid EMI noise from the driver. Lifting only one end was not enough. A good humbucker may not have been as suseptible and of course the strat has three pickups instead of the more common two.

Thanks for your interest though...the project does work as is but has to be adapted to the guitar in question a little. The only difference between this and a stompbox project is that it has to be built into and involves some modifications to the guitar and you have to litterally make the driver. This seems to be the stumbling block for most. Dispite it looking like a pickup and similar in principle, it is much easier to make than that. The wire is not fragile and can easily be wound by hand in about 10 minutes as it is only about 200 turns (rather than 1000's for a pickup with hair thin, hard to get wire). The example above is this fellows first go and experience of doing anything like this...and it looks the business.

The ultimate thing would be to have a dedicated sustainer guitar with only one bridge pickup and one neck driver. The switching, installation and everything would be much, much easier and potentially the space left by the removal of a neck Humbucker could easily accomodate both the driver (which is about 4mm thick) and all the circuitry, only requiring room to be found for the battery somewhere accessable.

I'd love someone to help devise a better circuit than the fetzer/ruby and potentially some clever fellow could devise some simple electronic switching to help with the bypass and wiring complexities too. Unfortunately a little out of my expertise. Thanks for the input...much appreciated...psw

OK I'm going to make one, but I have to finish some of my other audio projects first.  I'm populating a ten band mixer tonight, an Orange Squeezer on Monday and some work to get it all in one box.  Pic of my large stompbox at the end.

I have an old 1952 Kalamazoo that I love but have already made too many mods to it to call it original anymore.
So what the heck, it is a two coil set up.  Only problem is that it is a solid Rosewood body and very thin compared to the Fenders. No room inside the body at all, except just a little around the neck pickup.  I'll finish reading everything I can find on the "Sustainer" and give it a whirl.  My plan is to keep the driver on, but just barley on, and I will call that "off mode".  Note that anytime you turn a magnetic coil off, there is a backlash effect.  Also there is a 4PDT out there that is a latching 5V relay and very small.  I got about ten of them.

[psw]

Thanks guys...

KMS...you'll get all the support you need over at the PG thread. Quite a lot of ideas have been floated for a surface mounted device to aid installation. Some very professional looking concepts. I made one with a remote stomp style box. The design ideas were that the circuit and battery could reside in a tailpiece with controls sticking out, just behind the bridge. You may have to dig deep in the thread (mid 30 pages as I recall) but there were some great graphic realizations put forward.

Jaicen_solo...well, one of the more intriguing things about the sustainer project is that it isn't as intuitive as it seems. If it were, there wouldn't have been such a "big ass thread!" lol. The saturated square wave seems to make the phase things worse. Certainly a polyphonic signal will turn to mush as with a heavy fuzz. You can connect a speaker to the circuit to hear whats putting out, but the real effect of the driver on the signal can not. In the very early days of the thread I tested individual string drivers with pure tones of various shapes generated by a computer program...the results for any being more favourable were inconclusive and only really tested monophonically.

You are intuitively right that a squarewave would seem to be the desirable form. What happens, I suspect is that the driver significantly rounds them off, and/or is too slow to respond at high frequencies due to the pase problems. Potentially the thing could end up inhibiting string vibration if it were to fall out of phase enough.

The LM386..."leave pins 1 & 8 unconnected to allow full gain to be applied"...no it's the other way around, hard wire together for 200x amplification. I use a 10uf cap as suggested to contol oscillation. Although there is a trim pot I've always run the thing flat out to get the gain required so that trim is not really necessary I suspect and usually have omitted it completely.

"Cross talk from the driver to a neck pickup can be avoided by simply grounding the hot end of the coil.' Well, no...but you think it would. I found it necessary to lift both ends of my single coils because of EMI related noise. Humbuckers may not have such a problem, I'm not sure. Hence the elaborate switching. The noise (an unaceptable squeel) can be controlled in one mode...but when you switch over the coils polarity for harmonic mode it returns if pickups are at all connected. I suspect it is a pulsating transformer like effect between the coil of the driver and nearby pickups connected by their magnetic fields and the mag energy pulsing through the strings.

From KMS..."Note that anytime you turn a magnetic coil off, there is a backlash effect." "There is simply no way around this." I think you both may be right. The indication is that the pop only occurs when turning off. I wondered whether having the preamp always on and turning off the LM386 stage would help. My concern is with battery consumption. I wonder though about how much it could consume if the driver coil were to be turned on and off...if it's not connected, how much power would the circuit use? I'd certainly try it and may save on switching.

"A high impedance buffer will probably still be required to avoid loading the pickups' It absolutely does...I'd like someone to devise a simple opamp design or something to remove the need for the transistor biasing trim pot in the fetzer circuit. I used a two transistor, filtered preamp and though small, I'm sure could be simplified and shrunk with an opamp. You probably want a bit of gain though...not just a buffer, at least with a LM386 circuit.

Valuable contributions thanks. I think you are both really on to something with the always on solution...any ideas on the effect of battery consumption if the thing was idling...and whether other pickups could be engaged unless the drive coil was disconnected with this senario?

[Jaicen_solo]

If I was to build one of these, which I hopefully will when I get the time, I'd use a Mosfet buffer, and make my guitar active. A single mosfet booster can be biased to provide two buffered outputs which are out of phase. It's then a case of reversing the driver coil, or using an inverted circuit for the LM386. A bootstrapped NPN bipolar transistor could also be used.
To be honest, I don't understand why you're getting crosstalk when you have both ends of the coil grounded, there should be no induction whatsoever unless it's coming from your leads. Another idea for a sustainer 'bypass' is to simply ground the input and outputs of the 386, that way you'd get no signal coming in or out, though you may still have self noise from the amp in the coil.
It shouldn't be a problem letting the LM386 idle when not in use, they draw 4mA idling. I guess you could include a power switch as well.
You are of course right about pins 1&8 
I should have mentioned in my first post that I actually have investigated the Fernandes patent to see how their design works. The important part of what I can remember is that the driver circuit is class B, which allows more dynamic range. They also drive the strings with a square wave which has been filtered to reduce the upper harmonics. You don't really need those anyway, since the string will generate them itself simply through vibrating. The driver coil itself is also magnetically shielded.

One thing that nobody ever mentions with these threads is the use of a Pulse Width modulator to drive the coil.
Using one of these, it is possible to output a pulse from rail to rail. Now, instead of the amplifier pushing and pulling the string by equal amounts, it can be set to pull and release the strings. The Pulse width can also be tuned to compensate for low frequencies.

[psw]

  Looks like you might get the sustainer bug if you don't watch out Jaicen_solo...lol 

I haven't looked at the patents for a while...there are a lot. Sustainiac uses class D to save on power. The ebow the old LM386. I'm sure there are better circuits that could be fairly easily devised. Fernandes also uses a preamp to allow the driver to work as a pickup (possibly with an auxillary coil). Although many have proposed magnetic shielding...it looks as if Fernandes has abandoned that on current models. Fernandes also used to place the driver coil/s on their side! Sustainiac also used to use dual, phase cancelling coils but now appear not, Sustainiac still do. The floyd Rose patent is more specific about how they have implimented phase compensation circuitry. It is a fact that much of what is patented has never seen production. If anything, the devices are getting simpler!

I too thought that the guitar should may as well be active, or at least the power be continuously connected while plugged in. With the idling idea you could well find that a single 3 way switch could activate the two modes (driver phase reverse) and off (driver disconnect) which would be a neat installation. In future I would not incorporate the sensitivity knob at all (replace with a trimmer) and i note that Fernandes now do the same. Many people don't want an active guitar and the power issue did seem to be of more importance. A device like this will use up batteries, it's not just a signal processor...but it's onboard status makes it impractical to power with anything else. Fernandes alway used to use two 9 volts and was difficult to build in. Presently they claim 72 hour operation BTW.

I experimented a lot with magnetic shielding with limited success. It certainly is an option though making it could be tricky. I incorporated driver coils in U shaped channels, encapsulated them in ferris epoxy...all kinds of things. Reorienting the coil and magnetic fields is another idea. Many of the hex drivers actually had cylindrical coils and balanced magnetic fields. These coils pulled the string from side to side. A dual coil system is very attractive too as it would put out equal amounts of phase reversed EMI.

Now, instead of the amplifier pushing and pulling the string by equal amounts, it can be set to pull and release the strings.

I also did some work and there has been a fair amount of discussion over the pulse width modulation. One idea was to simply cut negative waves from the signal. It worked but...no better. Theoretically, the coil being always driven into one magnetic polarity will have difficulty letting go that state between waves and will shortly magnetise in that direction. The negative waves, along with the permanent magnet, help to pull the driver's state through neutral between pulses. In short, without the opposing wave the device has trouble pulling and letting go! This is vital with higher frequencies as the driver can only respond so fast. I tried to address this with different core materials and found that a core material like ferite was excellent (used in hex drivers a lot) in terms of speed as it transmits but doesn't hold a magnetic state (as steel does for instance...hence it's use in inductors).

I'd love to see some ideas on mosfet buffers and other circuit concepts. There are a lot more modern and efficient chips than the old LM386. The driver's are still evolving. By people making there own variations the thing will evolve faster than one guy (like me) working on their own. My proposal of the "thin coil driver" has shown to be novel and a very effective, and practical solution. It's small and compacts it's energy right up close to the strings...it's size and shape also effects it's response time. As far as EMI issues it really has not been installed enough for the variations and switching problems to be fully explored.

With my working guitar it is perhaps a little more of a special case. Highly suseptable true single coils. My driver is actually built on top of the neck pickup and shares it's magnet...so it's configuration is very much like a transformer. It also has a mid pickup. All the same, other experiments do suggest that non-used close by pickups do need to be completely lifted (much to my surprise and annoiance I can assure you).

In some recent stuff on fernandes (posted on the thread recently) I see they run everything to the circuit and back out. It could be that they use electronic switching (what I would like to have done) to run everything. I think I counted at least 24 connections running in and out of their circuit to various components inside the guitar.

As other pickups are off when the device is operating anyway, I'd prefer to see a single pickup guitar, with neck driver...and if you are going to leave the thing on anyway, make it active and incorporate some nifty active filters to compensate for the loss of other pickups (if you must).

Great contributions and I hope to see you sustaining away in the near future and contributing even more to the evolution of this kind of device. For sure, the involvement in making something like this will puzzle and delight you, besides getting you a new kind of expression from the guitar...welcome aboard, all those who are coming aboard...psw

[psw]

Randy...the maker of the above driver has made a great circuit as we have been describing...here's a pic...

And here's a pic of the circuit and driver on his project guitar...

He has reported mixed results and I suspect a couple of problems but it may be the wide core is in part explains it...
Therre are a couple of new posts if people are interested...that little circuit is sure impresive...basically it is a fetzer ruby with all pots as trimmers....very neat....psw

[psw]

me again...

For comleteness sake, here is the reverse of PG member, Randy (RV2)'s Fetzer/Ruby sustainer circuit pictured in the last post. Some inspiration for anyone making a circuit up and wanting to keep it small...

From what we have been discussing it could be made smaller still by removing two of the trim pots leaving only the one for biasing the preamp transistor!

As for driver's, I mentioned some others...this one is my favorite and works well apparently...made by PG member Tim (onelastgoodbye). The core is a magnet and the whole thing is set in epoxy. The magnet held steel plates for winding on which were removed once the epoxy set...result, a bobbinless coil with an internal magnet...solid as a rock!


As you can see I'm obsessed with this project and take every oportunity to promote it and to encourage others to do the work of taking it further.... . Also, Seasons Greetings to all and take care...psw

[Jaicen_solo]

Ok that's it, i'm going to have to build one now!
I stole a pair of JRC386's from uni about three years ago to build a sustainer and i've still not got round to it!
In one of my guitars at the minute, I have a handmade PAF copy with alnico II magnets in the bridge. This sounds great, unfortunately the neck pickup is stock and sounds, not so great
That's now going to get ripped out and replaced with my custom sustainer & SC pickup.
Since I have some Humbucker bobbins lying around, what I'm going to do is cut the top and bottoms off, and add some small steel spacers and the screw polepieces so that it looks stock from above. This will allow me to wind directly onto the polepieces for maximum flux density. I figure one of the high T ceramic magnets I have lying around should be more than enough to drive this.
I'm going to wind a 7k single coil onto the second bobbin, perhaps using the same method for a more fender like tone. That would be just about the perfect guitar for me then!!!

[psw]

That's Great J.S.

Somewhere on that bigg ass sustainer thread  i discussed doing something just like this. What I proposed/suggested was to block off the lower half of the driver coil rather than cut it up...

Basically take the pickup apart. Now strip the wire from the one that will be closest to the neck. Take this bobbin and pack up all but the top 3mm (cardboard would do). Then wind the driver coil onto this 3mm top section of the bobbin. Put it all back together and it's done.

Now you could get even clever-er . It is vital that the coil is potted as you wind. I used PVA wood glue as it is safe and works well. But ideally you could use epoxy as Tim has done. He described how it was done using kitchen plastic wrap and I brought that forward to the current page of the PG thread. Now if you were to do the same...you could take out the blocking and wind another partial coil below the driver coil on the same bobbin. It might not be completely humbucking but it would be a lot quieter than a single coil...just a thought.

Also...you could consider running the neck pickup (or the whole guitar) as an active pickup. Then you'd get all the power you need. This is what fernandes and sustainiac seem to be doing. Even the driver can "pickup" signals if sufficiently preamped...just like a low impedance pickup...just another thought. If you are really clever you could use the sustainer circuits existing preamp to achieve this effect.

Be aware that I did need to do some complicated switching to get my strat's two extra pickups (neck and mid) to be bypassed and the bridge selected. Your's would be simpler but could well need to have both ends of the neck pickup coil removed from the guitar's circuit when the sustainer is on. I did prove that it can be done so don't let that put you off!!!

So...I'm glad to see I've infected another with the sustainer bug. It really is a nifty little device and a lot of fun to play too! I hope to see your adventures posted both here and over at PG in the new year. BTW there is a new post tonight on magnets that might be of interest and I hear of a few more getting into making the thing. Each time they seem to get a little better and I'm sure that they are evolving...psw

[psw]

Hi all....

Happy New Year. Here's a soundclip of the sustainer in action with some interesting use of effects to create a fake organ sound at one point....

http://users.pandora.be/onelastgoodbye/sounds/sirensea.mp3

Also...to show there is nothing new under the sun...here's a sustainer patent from 1892!!!

[Paul Marossy]

It's been a while since I've followed this sustainer thread, but WOW!, these ones made recently look very professional. That's a pretty cool thing to see all these people building one now.

On a different note, that must be a record for the longest, largest, most visited forum topic ever. 

[psw]

Hi there...another random update on things DIY sustainer...

check out this 3D circuit realization by a PG member of the fetzer/rupy sustainer circuit...top and bottom




The area in the middle is reserved for two switches and he's even got an LED on the board...pretty neat!!

He said he made it with Eagle and POVray...anyone know about these tools? Anyway...sure "looks" good but I wonder how much a double sided board like this would set you back. The circuit previously posted was way small and yet was done on perf board. Anyway, the sustainer thread is still going strong and more people tell me they will be making them this year...pete

[Dave_B]

He said he made it with Eagle and POVray...anyone know about these tools?

Eagle is a software package for designing PCB's.  POV-Ray is a 3D package.  Both are available for free. 

I sure like the layout for the switches.  My attempts to make a 3PDT in Eagle don't look nearly that good.