Fresh Fizz

Ok...that seems pretty simple, so these mosFETs are swtiching a pulse at the frequency of the strings, is that the deal.

No, they are switching at the frequency of 30 kHz. Compare it with a movie projector which projects images at a rate of 50 (?) frames per second. When the thing works you're watching a movie, not a 50 Hz frequency. Exception: dither -> interference between the movement on screen and the 50 Hz thingie. Like when wheels seem to turn the wrong way around. That's one reason why the switching frequency has to be high.

Out of interest, what phase shifters work in this way?

I can't give you a brand name or schematic but I've seen a schematic with a switched resistor. Then there is a mosfet switch in series with the resistor of the RC part of the allpass filter. (Most of the time it's a fet or photocell/diode in parallel with the resistor.) So when the switched is closed 50 % of the time the resistance will be twice as big.

With the harmonic switch switched, it wants to drive the harmonics and will do on the low e, but it is obviously struggling. Perhaps it needs preamp power or something to lift up these signals. My previous versions all had pretty hefty preamps where as this uses only a buffer.

The problem is the lousy performance of our output stages. It's the combination of phase shifting and high cut. (My first and not so short contribution) Sustiainac's current drive will work much better in harmonic mode. But what you can do and what I've seen in a patent is use a high boost for harmonic mode. Which means more switching troubles.

Another option is to use the current drive principle together with the lm386.

This is what the theory says. There's not much more to it than steal a bit of figure 11 of the patent.

R26 and feedback loop to C9 into opamp "U1b" (IC U1b should read U2b). By the way figure 10 contains an error, 1001 is in the (negative) feedback loop.

Cheers

Fresh Fizz

The patent images show a discrete class-D output stage made from a comparator and a pair of N- and P-channel MOSFET drivers. At a switching frequency of many 10's of KHz the amp won't "see" the DC resistance of <3ohms (unlike the linear LM386), as the impedance at that frequency will be much higher - perhaps 10's of ohms? Depends on the inductance of the coils given the number of turns and the laminated core material.

Don't worry too much about the switching frequency. Basically it's analog in, analog out (+ distortion). The Mosfets which work like switches don't dissipate power, if they are open there's no current flowing through the mosfet, if they are closed there is no voltage across the mosfet. Now if only one of the mosfets could be closed there would only be 2 analog states, 9 volt and 0 volt. The comparator and switching signal are a way of creating the analog states in between by means of duty cycles. For instance if the top mosfet would be opened for 50% of the time (of a "pulse") the analog state of the output signal would be 6.75. It must be clear to everyone that through this process the signal becomes distorted. (50% of the time 4.5 volt, the other 50% 9 volt). That's the trade-off, more efficiency, less high fidelity.

Maybe some of you know the phase shifters built from opamp allpass filters that use mosfet switches (switched resistors) instead of fets. Same principle.

Fresh Fizz

he switching class D amp they use keeps the frequency constant at roughly 30KHz, and reproduces audio frequency variations by PWM-ing the output signal. The inductive impedance at 30KHz is also going to be a lot higher than 3 ohms. I wonder if the very nature of a class-D amp makes it better suited to evening out frequency response variations and problems in our DIY versions because the applied frequency is constant and unwavering?

I don't want to ruin this great thread, but do you know that all the information can be obtained from freepatentsonline?

It would be a pity if everybody started to copy the sustainiac. On the other hand I've seen some thingies that could be usefull to us. Like a current-source amplifier to deal with impedance and phase shift, a drive current limiter (to minimize fizz?) and the class D amp.

The patent I'm referring to has a three pickup configuration with the driver in the mid position. The switching of this version appears to be active (don't know if that's any good to you, Pete), the driver has it's own preamp to turn it into an active pickup.

If you create a free account at freepatentsonline.com you can take a look at the patent images. Taste the forbidden fruit if you wish.

Cheers

Fresh Fizz

To prove this to yourself, build a driver, then try it with and without a magnet

That was my very first sustainer project. I thought I could get sustain without a magnet. I used a single power darlington pinched off in class B, the driver coil as some sort of emitter resistance. It was a total failure. Lack of power, no decent sustain possible. Instead I got cricket-like sounds. Changing the bias of the darlington in class A didn't improve the performance. Except that my transistor and driver started to get really hot.

FF

I think I went too far with the hex things...at least to far two quickly. I think that it is easy to push 'u-beaut' designs a little too far. Sometimes you can get a little too clever and the devices get harder and harder to make. Making a dual driver would appear to be twice as hard as a single coil (two coils, cores plus you have to wire them together and stick them all together and mount them into the guitar somehow) for instance...so you'd want some kind of payoff for all this work.

The danger is that once you're start copying a part of sustainiac's device you might need to go all the way and you'll finally end up copying the complete sustainiac. That's not exactly the spirit of DIY. I'm only interested in that tiny part in the sustainiac that reduces the fizz. If the bilateral driver is the solution than I will build one, if I could only find some decent magnets. So I'm awaiting the results of UI's, Col's and Curtisa's projects. In the meantime I'm reading a bit of the boring patents hoping to find something fizz-related.

Good luck to you all

Fresh Fizz

The problem I'm seeing now is finding something that will make a nice neat bobbin out of. I was thinking of going down to the hobby shop and seeing if they had any plastic rectangular extrusions I could use. Any other suggestions?

For my driver I used some left-over 3 layer strat pickguard material. I can understand that people find it too thick, but it's easy to saw and when sanded it can be superglued easily.

It is an interesting chip...a little hard for me to get it would seem. Not sure about the battery consumption and other factors that may effect it in this application over the LM386...hmmm

Tda7231 or tda7231A 's quiescent current is 4mA at 9 volt.

Also..you appear to have fallen into the fatal flaw of position four. On a fiveway selector S1a will connect the middle and bridge pickup, and you have the middle connected to ground (shorted)...so to the bridge pickup will be shorted and so the guitar will be dead!!! Not good at all.

That was the best thing I could come up with. Makes a great kill switch though. Don't you want a stutter effect à la Gibson Les Paul?

Another idea of mine: a small resistor in stead of the wire to connect the S2a and S1b neck pup connection. When sustainer is switched on it'll give bridge pup with neck pup in series. In combination with the harmonics switch it'll give bridge with neck pup in series out of phase!

Phase stuff is relevant and has potential issues here, but phase changes with frequency so I don't quite see how we can address phase without some form of variable compensation system being the outcome...exactly as the patents describe. Similarly, we are now introducing phase differences between the locations of the pickup and driver...another variable and again influenced by frequency.

There is reason to assume that it is not as much of an issue as we might fear. Is it proposed that the phase differences are the cause of fizz...'cause I am not so convinced from practical observations of the effect. The fact seems to be that it can work and that clean performance is possible. I still work on the faith that a driver can be built that will operate satisfactorily within the range of guitar frequencies.

I think that there are two approaches. One is the hi fi type of approach. Make phase and frequency response as flat/neutral as possible. The other is the musical approach of letting the ears decide (what utopian isotope is doing with the use of an eq). I think that the hi fi approach is a good starting point, especially if you're in the process of setting up agc and/or overdrives. At the beginning there just are too many parameters to take into consideration. Now I'm arrived at the point that I'm going to experiment a bit with different eq to see if there is an improvement in sound and feel of the sustain effect.

Why don't you use a simple connector for the outboard power supply. I use a plastic female 6,3 mm connector. Then you have the option of using the power supply or a battery in a small box attached to your guitar strap.

Fresh Fizz

Is it bypassing the other pickups...and why does the neck pickup connect to the output of the circuit just before the driver? Also remember in positions 2 and 4 the middle is connected to the neck or bridge inside a 5 way selector.

How shall I put it? In my switching system the pickup selector has priority over the sustainer switch. When the Neck, Mid or combinations are selected there will be no feedback. Only when bridge pup is selected there will be feedback depending on the sustiainer on/off and power standby/off switches. These 2 switches could be combined into a single switch, but personally I think it makes more sense to have 2 switches. The quiescent current is low, around 4 mA. When using only 1 switch you have to deal with spikes, pops every time the sustainer is switched on and filter cap(s) are charged.

Unfortunately, there are still positions (#2) where the middle pickup connects to the neck and as we know, induced driver signals are created there...adding a the middle will alow them to flow into this pickup and so move them far closer to the bridge pickup to "hear"...if it doesn't scream on it's own. Even if controllable, the effect of being in position #2 will be different from #1 (neck pickup alone).

Then it could be something like this:

Fresh Fizz, I've been doing some simulation of the Fetzer Ruby, and the phase and amplitude responses are terrible. Are you using the basic F/R, or is it modified ?

Using this circuit could easily be the cause of your problems with the lower notes.

If you can post a diagram of the circuit and settings you are using (assuming they are non-standard), FET biasing voltage, pot settings etc., I'll check them out and see what the response curves are like.

Thanx Col, but I'm doing my own calculations in Excel. I haven't updated my schematic yet but the old one is not here.

The Fetzer Ruby at best (LM386 input attenuator an most attenuation) goes from 92º shift @ 80Hz down towards 0º @ ~2KHz.

With the attenuator at minimum attenuation, 80Hz gives 115º of shift !

The magnitude is equally hairy, theres a hefty drop from from 300Hz down to 80Hz of between 5 and 10db depending on settings.

That's because the input impedance of the lm386 is only 50kOhm. You need a bigger capacitor. In my schematic there is a voltage divider on that spot which takes care of that problem.

I have been looking at the lm836 national semiconductor datasheet. I saw that there is an easy way to get a 6 dB low boost with only 2 components. That's pretty cool!

(see amplifier with bass boost)

Cheers

Fresh Fizz

Very nice samples Pete.

Your sustainer device sounds definately more refined than mine. Did you use one setting per sample or have you've been adjusting settings while playing? To me it seems that you don't need more compression or output level there is more than enough feedback. And those tapped notes seem to go into feedback with ease. Maybe your driver is more efficient than mine. Or a compressed signal is more obedient than a distorted one. I presume that you were using fundamental mode all the time.

The strength of using agc is that it gives you a classy sound. But to me it seems that the fizz is more noticable when it occurs. With overdrive there is allways some background nasty-ness but it sounds more creamy. I only need to find a way to make my sustainer not to react so aggressively right after the attack.

Could you perhaps make a sample of a problem area of mine?

A single G note 3rd fret on high e-string (on my guitar I get a weak, fragile 3rd)

I've thought about your switching troubles. How about this one (what's missing) ?

Fresh Fizz

As far as getting sustain on the fundamentals.

I checked out the phase response of my circuit again (simulation), and from 80Hz to about 300Hz the phase response goes from about 50º to ~8º where it levels off until roughly 3-4KHz.

That 50 degrees phase shift is caused by the output cap. And you probably included the Zobel in your simulation

I wonder where the line is between gaining efficiency through higher inductance and losing it through losses in the circuitry that compensates for the phase response of the bigger inductance ?

I think that Pete has allready found out by trial and error. Thicker wire -> more turns of copper windings needed to get 8 ohm -> higher inductance. For low and mid frequencies this means a gain in efficiency but the highs stay where the are. It works the same as with overwound pickups. (Loosing the twang.)

Here's another possible reason why some folks have trouble with getting a fundamental note - this is particularly for folks who use a stronger drive current.

The electro-magnet should never be stronger than the permanent magnet, it should only ever help or hinder the permanent magnet.... If the electromagnet is stronger than the permanent magnet, a 'double pull' will occur causing a harmonic effect.

Explanation: When the electromagnet is in phase with the permanent magnet, you get lots of pull. As the two magnets start to oppose each other, the pull on the string reduces.

If the electromagnet is too strong for the permanent magnet, then there will also be a part (the peak) of the opposing swing of the wave cycle when the pull on the string increases again.

If this is happening, you either need a stronger permanent magnet or a reduced drive current.

I use a pretty strong ceramic magnet. But I use rather thinny blades as cores. I wonder whether if the blades get saturated they will produce harmonics. And for that reason:

1. have a tendency to build feedback on a harmonic rather than the fundamental

2. because all of us don't use a core as thick as sustainiac's could this saturation be the cause of fizz?

Fresh Fizz

For more on this subject, I found this discussion:

Adverse Tone Effect Discussion

Fortunately this is not something I have to deal with. Apart from the fizz the sustainer device doesn't seem to change my tone.

What I am getting is very much as you desire. On low level gains, I am getting a pretty clear sound, nice chord sustain and a fairly even response. Low strings ring at their fundamentals for about as long as they would without the device, but slowly morph to a harmonic (the low e to b above for instance) after this time. So, it plays fundamentals but on very low notes these are not sustained, their harmonics are...which is a really cool effect actually.

At higher gains, this response is much faster. At these gains at the expense of some non-fizz sounding distortion, you get very touch sensitive harmonics. A pick harmonic will typically continue to sustain as one, a sounding harmonic slid from fret to fret, brings them all out as harmonics in the same range. Tapping, instead of picking the strings brings out harmonics too.

Using such gains and techniques I get very strong harmonics all the way up to the 22nd fret high e (a little piercing) and even can get "evolving harmonics" where after prolonged sustain with high gain, the harmonics are revealed.

So you're saying that because of the overdrive I use my device is acting rather nervously, my sustainer is responding too soon, too intense? It's not that it's functioning badly but I want to know if there is any room for improvement. From what I'm reading from you my sustainer could well be a little one dimensional. So I'm waiting for the new soundclips from you.

the driver would actually bend the strings out of tune in operation!

COOL!

Even if I wanted to install another pickup there would be no place to do so. The circuitboard is placed in the mid pickup hole. I want to get more involved in solving problems which are specific to a more complex pickup/switching configuration but from where I am now at the moment it is a luxury problem. And you have to be prepared to answer all kinds of silly questions. OK, one for now, why is the sustainer switch positioned before the pickup selector switch?

Who am I to want 2 pickups if he only has one?

cheers

Fresh Fizz

To Col

or a pickup with a 6K resistance and an inductance of 6H (ballpark guesses), the lag at 300Hz is also ~62º

if its inductance was 3H, the phase would instead be 43.5º

I got some specs for real pickups - this one has a fairly average phase response: ( i looked at all sorts)

'DiMarzio dual sound' resistance = 13.68k inductance = 6.4H

here are the shifts at different frequencies from the guitars range (fundamentals only)

13.2º @ 80Hz.... 42.5º @ 300Hz... 61.3º @ 600Hz... 74.2º @ 1200Hz

It's nice to see that somebody finally took interest in my phase shift "thing".

(For me it's only a tool, a test to perform after having constructed a driver. There could be a lot of other things wrong with the driver but not it's electrical properties.)

What is important to realize is that with the guitar pickup it's not about power but voltage amplification. For the most efficient power transfer the load that's been connected up to the pickup would have to be equal to the pickup's impedance. But in case of voltage amplification we can afford to keep the current low. Which means the pickup sees a high impedance as load. To do the calculations right you have to take this load in series with the dc coil resistance of the pickup. If you do the calculations now you'll see that there is hardly any phase shifting.

You undoubtly know what happens if you connect your guitar to a low impedance amplifier circuit. You loose all the highs. You could look at it from a power perspective, lots of phase shift which means that a lot of the energy is backfired to the pickup, or from a voltage perspective, voltage attenuation because of high reactance of the pickups coil and low input impedance of amp.

You'll certainly end up with something closer to a sine wave pretty quickly if your system is not set up to minimize the damping effect the driver has on all but the fundamental.

Not necessarily the fundamental, it could be a harmonic. It's more a survival of the fittest, there allways seems to be one heavy favorite. It would be nice if we could create some sort of bi-state where two frequencies were equally ...favorite (?). Would two agc's or two overdrives in parallel with separate eq be a possible way of achieving this?

to Utopian Isotope (in dutch it's even a greater word play: utopische isotoop)

I assume you don't really want to kill your driver by frying it?

In my case, I still have an overpowered sustainer amp, that gives so much juice that driver starts to stink.

Because I overdrive a fet I know exactly what it's maximum signal is. That signal that's passed to my power IC is attenuated in such way that the IC never is overdriven. Meaning my IC doesn't produce distortion and never goes over the limit. (also been checked on oscilloscope). My device produces a lot of distortion and power but so far that hasn't harmed my driver.

Doing this way it's possible to take the highs out of the distorted signal before the power IC. If the IC gets overdriven there is no way to do this.

For some reason my pickup with it's ceramic magnet puts a lot of emphasis on higher harmonics especially on the plain strings. Or maybe it's the maple neck.

Too bad...

Greetings to you all

Fresh Fizz

The most important phase shift I'm talking about in relation to this is not the driver circuit or pickups electrical phase shift, it is the shift caused by the physical positioning of the driver with respect to the pickup.

What you really want/need is a method to convert the pickup signal into driver signal. Something like a computer programme that recalculates per string, per frequency the amplitude and phase shift for the occuring string movement. In fact simulating a pickup that's situated on the driver's position. The best solution would be to have both driver and pickup at the same position. Like Pete has been trying to do. Could it be done with an optical pickup and a magnetic driver?

So...the question is, what do you want an "artificially" sustained note to do over the length of it's duration?

I think that no matter what you'll allways end up with a sine wave. But it would be nice if it was possible to get a different harmonic by changing the way you pick the string. But in my case it's allways the same, 3rd harmonics low, fundamentals high on the fretboard. And I want 2nd harmonics as well!

I would like to rely more on another feedback principle: fingers to ears and vice versa. Stick a volume pedal in between pickup and F/R amp. And decide by hearing when the sustainer device has to be set into action.

I tried to get fundamentals below the 12th fret. Therefore I used a fet to boost the low. This way I was able to get fundamentals on my wound strings but it made my plain strings sound even thinner by comparison. And it didn't feel right. When playing on my g-string I felt the wound strings shaking insanely. Anyone knows how the commercial guys get good fundamentals on all positions? Is it eq, all-pass filters or is it their trick of trade?

Cheers

Fresh Fizz

The easiest way to think about this IMO is to use the Helmholtz model and think about what happens when the kink is between driver and pickup (although, I'm sure you can do the math with a superimposition of sine waves and get the same result). Basically if the phase relationship between the driver field and string for that string/fret combo is not exactly 0º, there will be active damping of the overtones - the worse the phase difference, the quicker the sound will revert to a pure sine wave. Using a square wave as a drive wave will only help to keep some nice overtones in there if the phase is well matched.

I don't think that the phase issue is a big problem. It can be split up into 3 areas: pickup, driver and electronics. For the driver I allready gave you a way to calculate phase shifting. Here is my document to calculate the phase shift of the driver.

For the pickup you could do the same calculations, but it's a waste of time. Any decent pickup well extends the range of 1100 Hz. Unless you have some kind of superdistortion very ugly metal humbucker. What electronics concerns, if the signal is only amplified without any toneshaping (eq) there won't be much phase shift.

About Helmholtz: I found this interesting link.

I think this and also this links show the real problem to solve to achieve fundamentals below the 12th fret. The java applet shows that the fundamental is around 10 dB weaker than the 3rd harmonic (pickup at some 2" from bridge). So the fundamental cannot compete with the 3rd.

If you're using high gain and no AGC, then when there is a strong guitar signal, much of the fizz component of the sound will be lost through clipping, so you tend to get just 'fresh' fizz, rather than recycled fizz build up - this will sound a bit like a poor quality distortion unit in the background - it's almost acceptable. When on the other hand you have a low level guitar note, the fizz won't be smashed by the clipping of the guitar tone and gets nicely amplified. It then is more readily transfered to the pickup directly through parasitic coupling. And then amplified some more, added to more fresh fizz and the process repeats, so the fizz not only becomes a more significant part of the sound, it also becomes less and less connected (in time) to the note you are playing... so it's becomes more like a background noise than a guitar distortion sound.

This means that it might be possible to make a significant improvement without actually removing all fizz - if we can in some way stop the fizz feedback building, or limit that process?

I fully agree with you. Something out of control, like driving off a cliff.

How about slew rate limiting? 1 , 2, 3

Btw Fresh Fizz, is your scope dual channel ?

No, it's a prehistoric version the name scope unworthy.

I bought it for my at 286 (remember ?). It won't work under Windows. I need to boot if from floppy without the himem.sys and no chance to read my harddisk partitions.

I will try and get some sound up so that it can be judged for what it is..."fizz" is a pretty loaded word and we may not all have the same kinds of sounds coming out.

Well my sustainer device produces a lot of "dirt" but I think that they are 2 different phenomena: harmonics created by overdriving the fet + clippers and the fizz.

I think Col's description of fizz comes very close.

I had forgotten this, but I wonder if it is possible that your output caps are in series resulting in a similar reduction in value...the formula is the opposite to resistors that increase with series...correct me if I am wrong. It cam to mind because your description is in line with the effects I get if I reduce out put caps and if I was running 600uF on mine, I would be getting strong fundamentals, probably at the expense of high string response. You may wish to check this.

No they are in parallel. Mind you, all the trouble to squeeze in an extra cap only to get less fundamentals?

It also reminds me, I did think about your suggestion of running the pickup coil (about 5,600 ohms) in parallel with the driver (this one about 8.2 ohms) but calculating it out (which I am not good at BTW) I got

1/8 (0.125) + 1/5600 (1.79) = 1/ (1.915) = 0.522 ohms

Question answered by Setain.

I was thinking Pete, have you ever considered of skipping the neck pu. Why don't you use the driver as pickup? What you need then is a noisefree microphone amplifier. And the switching system had to be rock solid of course. You don't want a few volts on your microphone amp. But maybe you want to keep it passive. I think that is what the commercial guys have done. Somebody mentioned a transformer, I think it is a step-up transformer for the driver. But I am not sure I have never seen a sustainiac in my life!

Cheers

Fresh Fizz

I did some tests with my Handyprobe (oscilloscope on PC).

I found out that the appearing wave forms are pretty much sine wave. So even with a lot overdrive the feedback will settle in a more or less sine wave-y tone.

I did the test only for fundamental mode. There is no harmonic mode on my guitar.

Maybe it is due to the overdrive but I only get 3rd harmonics below the 12th fret position. Above the 12th fret it's all fundamentals. Is that normal? I would have expected positions on the fretboard with a 2nd harmonic. I use a 660u (2 x 330u) output cap.

One thing about the fizz. I found out that the fizz is most anoying when you hit the string softly. Then the driver causes some sort of volume swell which sounds very fizzy.

Cheers

Fresh Fizz

fINALLY, A BREAKTHROUGH OF SORTS....

Congratulations, Pete!

A triple stacked coil. All pickup makers eat your heart out.

utopian isotope, you did some experiments with metal sheets above your driver.

If found this at the sustainiac website:

6. Metallized pickguards and guitar bodies. Also, metal pickup trim rings. These conduct electricity. The pulsating magnetic field that comes from the electromagnetic driver will produce pulsating electric currents in the pickguard or guitar body or pickup mounting rings. This will cause the bridge pickup to respond to these currents. The end result is poor sustainer performance. So, you must use a pickguard having no metal in the vicinity of the pickups (or scrape off the conductive shielding from this area).

More you find here.

Cheers

Fresh Fizz

There seems to be a lot of work right now into alternative grounding schemes.

Has anyone worked out if this is even an issue (at least for there system)?

If all this work is based on hunches and guesswork, then not only might it be pointless, it might even cause more problems by making potential solutions to real problems more difficult to achieve.

Yes, you're right. I'm only wild guessing. But everybody who is working on this sustainer project has some problem with fizz or grunge. It's the desire to make some progress that makes people come up with all sorts of hypotheses.

It shouldn't be too difficult to find this out through experimentation.

I don't think this is true. You could experience that an alternative grounding system didn't improve the performance. But what if more conditions have to be met before there is an improvement? It's better to think loud. Maybe some can make good use of it. Or eventually prove I'm wrong.

Thanks for suggestions. Fresh Fizz, as I looked your schematic, is there any buffer between guitar and sustainer?

My current setup doesn't include any dedicated buffer, only an eq.

As those have common ground, turning sustainer's volume down turns also guitars's volume down. Strange that I haven't noticed it before..

Next I have to make some decent splitter/buffer, something like Craig Anderton's "spluffer". It uses outdated ic's, so I have to re-draw the schematic.

It also uses split power supply, so we'll see what it has to offer.

No utopian isotope there is no buffer, my volume pot acts the same as yours. I personally like it this way.

By the way it wasn't my intention to draw a complete schematic but only the grounding issues.

Fresh Fizz

I hope this makes sense to somebody.

so yes...any ideas on virtual or more likely floating grounds (is that a term???) or ways to isolate the driver further from the signal?

Hi Pete, I'm going to make a drawing of how I think the mass, ground, earthing whatever works best. As soon as it's ready I'll post it.

Over the last few days I have been running the device with the secondary coil shorted...it gives a very clean response and strong fundumentals...this morning went back to a 100k load resistor...previously I had used a 2meg pot all of which work fairly well. These loads are pretty variable but even a 100k load enables a dramatically more powerful response.

I have yet to make the "breadboard" adjustable version of my circuit (been a heatwave down here again) but perhaps this will give some more scope for tweaking. For instance, the LM386 has a both positive and negative input (pins 2 and 3) yet we tend to ground the negative as a reference...perhaps there is some scope there to address the grounding issue...or to run a cancellation signal or filter back into this and cancel out interference. The forces are pretty powerful, but I wonder if I could take the signal from the secondary coil and cancel out distortion/fizz effects in this way also. Perhaps an even more powerful effect could be achieved if the electromagnetically induced voltage from the secondary coil were to somehow drive the signal also...if controllable, it would seem that power (amp and battery) could be much smaller...though distortions are likely to be a big problem...and oscillation (squeal) very hard to control...

Wouldn't it be best to connect the secondary coil (= neck pickup ?) to the driver coil? (Driver coil and neck pickup in parallel). This can be done because the neck pickup has a much higher impedance. Meaning the load the lm386 sees stays allmost the same. Induction voltages are gone. ( Technically you don't have a transformer anymore, but 2 coils in parallel.)

Fresh Fizz

I am not sure what the SCT actually is from that diagram! It look kind of like a bi-lateral driver.

STC is a dummy coil but humbucker style and made out of copper foil. I've put it in between driver and guitar pickup. The idea behind is that it creates some sort of flux shield. I favour an all humbucking configuration: humbucker driver, humbucker pickup, humbucker STC dummy coil.

I got rid of the squeal. I'm still hearing some distortion, but I'm not sure it is the same fizz thing you are talking about. It's more that the string resonates with what it receives from the driver. And because I'm using heavy overdrive it's fundamental with a third (down the neck). It's not an unpleasant distortion it has a nice zooming quality. (Zooming as in: What would you rather be or wasp?) When I palm mute the strings and pick I have an undistorted sound (so no distortion on the pick attack).

I have to focus now on another problem area which is the low part of the high e-string. If you ask me what the problem with this string is I would say that it has to do with what the guitar pickup makes of the string movement. It looks like there is too little fundamental and 2nd and 3rd harmonics. It looks like the sustainer tries to build feedback on even higher harmonics but for some reason won't make it. And falls back onto those lower harmonics. Sustaining but weak sound. Maybe that's the Helmholtz motion, Col!

I have allready tried to change the overdrive character, more even harmonics instead of odd harmonics, but it doesn't make a lot of difference.

I think it will be worthwhile trying to power it with a remote source and it is probably about time someone tried the stereo lead power idea. I can't see anyone really wanting to operate the thing with remote power with the guitar plugged directly into a wall wart and the amp...such a hassle.

Well it depends on how much use of the sustainer device you're going to make. I can remember that Les Paul even wanted to have a microphone connected to his guitar.

I'd still like to hear how others are finding the power on and off functions are working even with single pickup guitars and powersupplied amps...

I only have pop noise when I switch on. I'm pretty sure I can fix this.

Fresh Fizz

I'm in the mood for being long-winded again. I hope I'm telling something new to you. I kept a present for you at the end of the story!

I've told you a few days ago that my sustainer device had a tendency to squeal during the note attack. What I haven't told you is that in harmonic mode the bugger was out of control screaming like an emotional pig. This might well be due to the heavy overdrive in combination with the strong output level.

Both my guitar pickup as well as my driver are humbucking devices. So why is there a difference in the amount of squeal between the fundamental and harmonic mode?

The only answer I can come up with is the following:

I think the easiest way to explain this is make a comparison with guitar pickups. You all know the strat with its 3 single coil pickups which pick up hum easily. There is a way to buck the hum in the combined positions by using a middle single coil pup with a coil that either is been wound or connected the other way around (I don't know which one is clockwise/ counterclockwise) .

The middle pup is electrically out of phase with the other pups. But that doesn't mean that the sound is out of phase. That depends on the orientation of the magnets of the pups. Only if the middle pup has a reversed orientation of the magnets the pup combined positions are in phase soundwise. This is like the fundamental mode of the sustainer device. Can we make the strat's combined positions out of phase soundwise by swapping the connections of the middle pickup or by some way inserting a signal inverting amp? Yes we can but we loose the humbucking effect.

The only way to create out of phase sound and keep the humbucking effect is to reverse the magnet of the middle pup. The same goes for the sustainer device. If by switching from fundamental to harmonic mode the squealing worsens or improves it means that both configurations are not equal when it comes to bucking the hum.

And this works the other way around. If we can reason that there are 2 different configurations there is no point trying to make the bad one as good as the good one.

So are there 2 different configurations? I think there are. When we imagine the sustainer driver to the left and the guitar pickup to the right we could have these combinations.

S and N: orientation magnet

+ and -: relative phase of coils

What matters is the relationship of the driver coil and pickup coil that are closest to one another.

1.

driver_pup

SN___SN

| |___ | |

+-___ +-

Electricaly out of phase

Good for fundamental mode

2.

NS___SN

| |___ | |

+-___ -+

Electricaly in phase

Bad

3.

SN___SN

| |___ | |

+-___ -+

Electricaly in phase

Bad

4.

NS___SN

| |___ | |

+-___ +-

Electricaly out of phase

Good for harmonic mode

The best quiet configurations are 1 and 4, to go from 1 to 4 you need to reverse magnet!

This could be the reason why the commercial sustainers use a split coil for the driver. But there is an even bigger gain by using a split coil. Because of symmetry of magnetic field created by the split coils there is a lot less inductive crosstalk induced in the guitar pup.

Utopian Isotope introduced the dummy coil. I haven't tested that one but most likely it works that well because it works in an out of phase condition.

I wanted to take this dummy coil thing a step further and therefore I present to you the Siamese Twin Coil . See schematic. (The circuitry still needs to be improved.)

I have no problems with squeal anymore! Maybe worth trying.

Bye

Fresh Fizz

Once again, I found totally strange phenomenon, that saturates sustainer and/or driver, and adds even more fizz and grunge.

To test this, simply hover some non-magnetic piece of metal above driver, and you should hear the difference.

Metals that I have tested: copper sheet 0.1mm, brass sheet 0.3mm, aluminum: sheet 5mm and solid bar 50 x 50 mm, and lead sheet 6mm. Thicker material gives more dramatic effect.

I was surprised that even pure lead sheet affects so much for the driver. Since lead doesn't resonate easily, it must have something to do with capacitive forces. And maybe electrostatic too?

Seems that relationship between driver --> strings--> pickup(s) isn't purely magnetic.

Since all those elements can cause excessive behaviour(raw fuzz) in sustainer/driver, maybe they can also offer counterforce that can be used to eliminate them?

It looks like these tests can give us a new perspective of how to fight the fizz. Maybe a breakthrough? The dummy coil was indeed an eye-opener to me. Thanks utopian iso!

So the capacitive forces are against us too.

inductive emi is more likely to happen (high currents). capacitive emi is more likely to occur with high voltages. This could be the case with Pete's flat driver on top of the neck pup, where the neck pup produces high voltages.

The bad thing with guitars is that everything is so high impedance there could be a problem with capacitive emi.

Have you done or could you do these tests under the following conditions:

1. sheet not earthed

2. sheet earthed at guitar cable

3. sheet earthed somewhere more down the line like earth connection of wall plug, preamp to avoid currents through the guitar cable

I'm looking forward to hear the results from you

Cheers

Fresh Fizz

I'm amazed how rich this thread is. There are so many ideas going on at the same time! You could call ourselves the sustainer think tank.

I had an idea of sending a tiny signal through a pickup coil...perhaps say a middle pickup on a strat, perhaps fed by an op amp to avoid impedance problems....this creating a kind of magnetic shield or "loop" of many turns of wire (already in the guitar anyway)...just a thought.

Which means you're creating a second trafo-like coupling, just like your flat driver/ neck pup. This could mean more unwanted side effects.

Perhaps you have already done something like this...can you explain this two grounds a bit more?

My problem is that I'm still catching up with you guys. Maybe I shouldn't have mentioned it already. My sustainer device is still suffering from diseases that you have already solved. Like optimalizing signal levels, eq, amount of overdrive/compression to fight the fizz and squeal. But I have the idea that this might be a way to address the shared earth problem. In your guitar the sustainer device is earthed means the guitarcable. But what if there is a slight resistance (because of connectors). Let's say that resistance is 0,2 ohm. It means that there will be some of the sustainer device's output voltage on that resistance.

V_on resistor = V_sustainer * 0,2 / 8,2

You could say that 0,2 / 8,2 is feeded back to the guitar pickup (same earth). The sustainer device has an overall amplification that might well exceed 8,2 / 0,2 times. So there is allways a concern for oscillation/instability. Because of phase shifting of the combination of driver coil and zobel network there allways will be frequencies where the phase shifting (driver output to pup input) is not exactly 180 degrees.

With the 100 ohm resistor I use the feedback is 0,2 / 108,2

The idea is that the cable of my sustainer device's power supply will serve as connection to earth. So from the power supply goes a connection to the earth of my guitar preamp. At this point it's not working this way because my adapter is not earthed. But even so it should improve the problem with the shared earth. Technically the sustainer device's ground is floating from earth but the efffect should be minimal.

I remember, Pete, where you said that you got squeal when the strings touched the core of your flat driver. I think it has to do with this shared earth phenomenon. Core 's connected to earth together with eddy currents.

I remember spazzyone with the outboard amp who said not to experience any big troubles with fizz.

I am not sure about this...so to turn down the gain of the sustainer you loose guitar volume as well? The device needs to be balanced...it could be that you have sufficient power capacities to run the driver under sqeal too hard...but to dial it back, you are also loosing your guitar signal too.

I'm a compressor junk, I don't loose a lot of volume. To me it's a very natural thing that by adjusting the volume you get more sustain. Just the same like when people use their volume pot to get more overdrive of their amp.

You are using a much more powerful amplifier and an unusal application. Here is a schematic for this chip amp

tda7231 schematic

It lists the power as 1.6w into 4 ohms. We are running LM386's typically into 8 ohms...between 0.5 and 1 watt, but restrained by the capacity of batteries, with outboard power, you are giving it virtually unlimited current capacity! This project is a lot about balance and more power does not equate to better or improved performance. Also, there is little to no experience with this amplifier and the effects of running low ohm loads on the amp or such driver designs. It could be an advantageous choice of amp (any reason for choosing this?) however battery operation is probably not an option at these wattages, if you were seeking improved headroom to get a nice clean signal. Your preamp though undermines any advantage of this though. Remember, if you run a LM386 at full blast, you will automatically get some clipping, and with a battery, the power is clipped at the level to which the battery can supply it.

Reason? Maybe the same why people stick safety pins in their nose. Trying to be different.

In 8 ohms I should be able to get 1,8W. But I settle for about 1,2 W. I have outboard power so consumption is not a big deal. In understand that the combination of having lots of compression/overdrive and lots of output power could be too much but I was a bit concerned that my driver with the slim blades wouldn't be that strong. Better to start with too much than with too little! It prooved to be very strong. When my high e string wasn't yet at right pitch the feedback made it almost to woodpecker itself into the fretboard.

Another advantage of the ic is that one side of the dil package is connected to ground. So I can use a ic socket with a peace of tin soldered onto it. It makes a perfect heat sink. I sticked my sine wave generator to it and let it run full blast for a few minutes. Kept it pretty cool.

The specs talk about soft overdrive characteristics. Tubey?

Pete thanks for the advice given. I'm sure that I can use some of the ideas you mentioned.

More food for thought

Enough said

greetings,

Fizz

It sounds like you have the gain set so it is nearly squealing.

When you hit a note or chord, the extra signal level (more square signal wave shape with more harmonics?) is enough to start the build up to squeal, however, this is quickly cancelled as the fast decay of the guitar removes energy faster then the squeal can build up... so you get a small attempt to squeal during the attack of each note. ?

Does that sound about right?

Yes, I think it has to do with the dynamics of the sound. During the attack there is a tendency to squeal. Because the fundamental is masked by all those higher harmonics string feedback doesn't happen immediately. Only after the attack the string feedback starts to build. If you mean the fast decay of higher frequencies (attack) I'm your man.

I have holding back some valuable information. I use a treble booster inside my guitar (1 fet, 16 dB high boost). It serves as input signal for my sustainer. It's only after the overdrive I cut the high but overall there is still high mid boost. On my guitar there are no extra controls for the sustainer device. There is only a 6,3 mm female connector for the 12V power supply. I have 2 grounds (?) one for the pickup ,treble booster, guitar cable and another one for the sustainer device. Both are interconnected through a 100 ohm resistor. My volume pot serves as sustainer drive controll as wel. Cheap

My driver is a humbucker with 2 blades, it looks a bit like those dual rails pickups. The overall impression I have of the driver is that it's good higher up the neck sustaining with a fat sounding fundamental but a bit thinnish lower on the neck. And that is almost the same on all 6 strings. Maybe a peculiarity of my driver?

It could well be that I use too much gain. When I set the drive lower there is still good sustain.

Some tweaking is going to happen. I'll try to get rid of the overall high boost.

Fizz

Good luck...are we looking at a multi pickup installation by chance...would be interesting to see what you com up with...

Nope. A single humbucker Seymour Duncan TB5 at the bridge position of my stratocaster. The sustainer circuitry fits in the middle pup hole position. I use a frankenstein pickguard. It's been cut in 2 pieces. So we have one piece with driver and pup fitted in and a part to fit in the controls. This way it's possible to open the guitar without removing strings all the time. I only had to add to screws for the part below, that makes 13, but we're not superstitious (?).

Will the resistor from the +ve of the output cap to ground not cause problems...it is effectively tieing the +ve and -ve of the driver output around the cap...or does the other resistor nullify it. Will it cause problems with the input.

You may be right though...not sure...will build it into the testing circuit I am working on.

It's a way of keeping the voltage at the cap at 4,5 volt. The driver's load is 8 ohm so the 2 resistors (4700 ohm) won't have that much effect on the working of the ic.

I suppose that you have a main power on/off switch to which the voltage divider can be connected besides a switch that only switches off the f/r.

But to be honest I don't know if this addresses the problem.

With the testing I did I only experienced problems when switching on my sustainer circuit. (Pfffeeewww in dutch, don't know in english)

Switching off seemed to be quiet.

So I am back om my feets again.

It's a lot easier to do the theoretical blaba than to really get a sustainer system to work and be really satisfied with it.

The good thing is that I've got the sustainer working. I'm using an overdrive (fet with diode clipping) into a tda7231a chip. It has tremendous overdrive and power and I get good sustain on all of my strings and all positions. At this moment I'm only using the fundamental mode which is free of squeal.

The bad thing?

FIZZ!!!!!!!!

After closer investigation I came to following conclusions. Maybe you could tell me if you share my opinion, Pete and Col.

1. When I say that there is no squeal I notice that there is sensitivity point, a specific frequency band, where the system would like to sqeal if it only had enough energy/volts/whatever to feedback.

When I listen through a compressor in front of a distortion/overdrive into headphones I imagine I can hear that frequency band in the background. When I switch off the sustainer it's gone.

When I hit a low note on the low e string it seems to waken up this monster. Like it's thinking: "finally enough energy". But there's not. So it's like a short colorization of the sound (emphasizing that particular frequency band) and then going into string feedback.

This is a very unpleasant sound effect (not musical sounding)

2. Crosstalk of harmonics generated by the overdrive.

This is something I can live with because it doesn't sound unmusical. Surely I will try to diminish this effect as much as possible but it's like balancing off efficiency and sound quality.

I'm thinking of what is the next best step

Cheers

The Good Fizz