Ebow Exposé - Part III

[Paul Marossy]

I started an Ebow thread here back in 2007 called " Ebow Expose Part II - Ebow Anatomy Inside" (https://www.diystompboxes.com/smfforum/index.php?topic=57817.0). Sorry the pictures linked there don't exist anymore since I shut down my website a while ago now. This was when I acquired a PCB from a hacked apart Ebow for some hands-on evaluation but at the time I wasn't able to get as much info from it as I thought I would. Anyway, I am considering this new thread Part III on the topic, but this time I have a lot more info to share.

So... regarding Part III, I was playing around with LTSpice simulating various circuits over the last week or so. I thought it would be interesting to do an Ebow just for fun and discovered some very interesting things! I was also able to get some readings on the coils and made some measurements of the physical size of the magnets, steel cores, etc. In the course of trying to verify a couple of components on the PCB I have discovered how it's actually wired - the schematics currently on the web are not correct. I believe I have also learned how that regular/harmomic mode switch actually works too. All of this came out of me just trying to verify the values of three components on the PCB!

UPDATE: Here is a Part II that I had to make to correct a couple of mistakes and update a couple of things due to new information I discovered.

[Ripthorn]

Nice! I recently documented a full tear down of a sustainiac sustainer, including its harmonic mode. Very cool stuff.

[Paul Marossy]

Nice! I recently documented a full tear down of a sustainiac sustainer, including its harmonic mode. Very cool stuff.

That's another interesting one. I downloaded the patent on that some time ago and gave it a good read. In the case of the Ebow, the patent docs do show something similar to the schematics floating around on the web, but apparently they had tweaked the original design at some point along the way and it morphed into something different than what the patent documents show. I'm just happy to have figured out what that mode switch actually does. In my thread from 2007 I thought maybe it was an output coil with two windings or something, but the way they did it is a lot simpler than that.

[anotherjim]

I noticed that it's impossible to simply drive a sustainer coil from an existing magnetic pickup without getting the wrong kind of feedback. It has to be a closed system picking up the string and driving it. If the guitar uses a piezo or some other non-magnetic pickup, you only need the drive coil but then it also needs a cable from the pickup output to the sustainer drive amplifier.

[Ripthorn]

I noticed that it's impossible to simply drive a sustainer coil from an existing magnetic pickup without getting the wrong kind of feedback. It has to be a closed system picking up the string and driving it. If the guitar uses a piezo or some other non-magnetic pickup, you only need the drive coil but then it also needs a cable from the pickup output to the sustainer drive amplifier.

An existing pickup is built all wrong for driving in a sustainer system. The impedance is way too high for the kind of power available from a 9V battery circuit. Most sustainer drivers are in the 4-8 Ohm range.

[anotherjim]

No I don't mean driving the string with an existing pickup, I mean using the guitar's pickup to feed the sustainer the signal.
As I said...

...to simply drive a sustainer coil from an existing magnetic pickup

[Eb7+9]

Cool video Paul,

Hey, not to be an ass but are the scope inputs set to AC or DC coupling ?!
Just checking ...

[Rob Strand]

Good work Paul.

I had a look at the coils and the input coil resistance doesn't seem to add-up with the wire used.

From what I can see the *coil* dimensions are,

OD = 7.9mm
ID = 4.5 mm
axial length = 4.3mm

The OD I'm getting from the pics.   The OD of the coil doesn't wind out to the full ID of the ring (10.75mm).
The ID I'm getting from the 4mm diameter pole piece dimension, with some allowance for the bobbin.
The 4.3mm I'm getting from the minimum of the ring height dimension and the pole-piece height dimension; could be less if there is a bobbin and it has thick sides.

When I match that to the winding resistances you gave:

                            Input            Output
Target ohms            11               8
Turns                      346            295
AWG                      37               36

Inductances:
L_air                      490uH          350uH                 ; that's the coil outside of the magnet assembly
L                           2700uH        2000uH                ; rough estimate of the coil in the assembly
                             to 4000uH    to 3000uH            ;  upper estimate far more likely than the lower estimate.

From the pics it sure looks like both coils are wound more or less to the same coil OD.
So the thing that is questionable is the resistance of the input coil, maybe you can check it again?

If the resistance is actually 11 ohm then there's something weird going on because if the coil
is wound out to that OD you would expect a much higher resistance (and number of turns) for the
small wire you are indicating.    You see from my estimates going from 8 ohm to 11 ohm doesn't
need to reduce the wire diameter by much at all.

Sometimes when people make these types of sensor structures the *pole piece* is the magnet
and the back-plate and the outer ring are ferrous material but not magnets.   A guitar
pickup is like this.

[Paul Marossy]

Cool video Paul,

Hey, not to be an ass but are the scope inputs set to AC or DC coupling ?!
Just checking ...

Yeah mean actual scope, or virtual scope?

[Paul Marossy]

Good work Paul.

I had a look at the coils and the input coil resistance doesn't seem to add-up with the wire used.

From what I can see the *coil* dimensions are,

OD = 7.9mm
ID = 4.5 mm
axial length = 4.3mm

The OD I'm getting from the pics.   The OD of the coil doesn't wind out to the full ID of the ring (10.75mm).
The ID I'm getting from the 4mm diameter pole piece dimension, with some allowance for the bobbin.
The 4.3mm I'm getting from the minimum of the ring height dimension and the pole-piece height dimension; could be less if there is a bobbin and it has thick sides.

When I match that to the winding resistances you gave:

                            Input            Output
Target ohms            11               8
Turns                      346            295
AWG                      37               36

Inductances:
L_air                      490uH          350uH                 ; that's the coil outside of the magnet assembly
L                           2700uH        2000uH                ; rough estimate of the coil in the assembly
                             to 4000uH    to 3000uH            ;  upper estimate far more likely than the lower estimate.

From the pics it sure looks like both coils are wound more or less to the same coil OD.
So the thing that is questionable is the resistance of the input coil, maybe you can check it again?

If the resistance is actually 11 ohm then there's something weird going on because if the coil
is wound out to that OD you would expect a much higher resistance (and number of turns) for the
small wire you are indicating.    You see from my estimates going from 8 ohm to 11 ohm doesn't
need to reduce the wire diameter by much at all.

Sometimes when people make these types of sensor structures the *pole piece* is the magnet
and the back-plate and the outer ring are ferrous material but not magnets.   A guitar
pickup is like this.

First, holy crap! I just got an email notification! It's been like ten years since I got one even though I had everything set to up to receive them all along.

Anyway, I took a measurement from a different place in the windings of the teeny tiny wire and came up with 20.1 ohms. Maybe that is closer to actual? Seems like it would be. It's very hard to get readings off that winding as it sustained worse damage when they poked the magnet out, and those wires break ridiculously easy if you handle them too much. 

I got out the digital caliper and measured the actual spools of wire. The thicker wire was 9.35mm x 5mm tall and the teeny tiny wire measured 9.26mm x 5 mm high. It's not quite as large as the ferrite ring that surrounds the windings. That goop it's potted in makes up for the difference in size.

The pole piece is definitely just a steel slug. The ferrite ring is attracted to a magnet but is not itself magnetic. The two magnets are most definitely magnets! 

Just for fun I was thinking about setting up a test rig for my scope to see at what frequency those ferrite rings start to saturate.

[Rob Strand]

First, holy crap! I just got an email notification! It's been like ten years since I got one even though I had everything set to up to receive them all along.

Never knew it was even possible.

Anyway, I took a measurement from a different place in the windings of the teeny tiny wire and came up with 20.1 ohms. Maybe that is closer to actual? Seems like it would be. It's very hard to get readings off that winding as it sustained worse damage when they poked the magnet out, and those wires break ridiculously easy if you handle them too much. 

Understand 100%, I've had to measure that stuff many times.

I got out the digital caliper and measured the actual spools of wire. The thicker wire was 9.35mm x 5mm tall and the teeny tiny wire measured 9.26mm x 5 mm high. It's not quite as large as the ferrite ring that surrounds the windings. That goop it's potted in makes up for the difference in size.

Well I tried to match the new dimensions with the new resistance.  The output coil looks fine.  The wire size for the input coil is 37AWG (0.12mm diameter).  Thinner than before (as expected) but it's not an enormous difference and not anywhere near 0.05mm.

So what I did instead was see what resistances ended-up with 40AWG, 42AWG, 44AWG wires.   As you can see the resistances are very much larger than 20.1 ohm.   The wire is worked out by looking at how many turns are required to fill-up the measured winding cross-section.

I've done this type of matching many times before.  For accurate results it takes a lot of effort getting all the dimensions right.  Measuring the wire diameter is always a little tricky due to the enameled coating and kinks.  On top of that you still have to make a stab at how well the winding is packed.   Normally I expect to be within 1 or two wire gauges and I often kick the estimate using the resistance and inductance measurements.   It's also easy to screw-up the calculations.   I did verify the calculation against some other software and the agreement was very reasonable, so I'm think the calculations are correct.

Code Select Expand


Matching Measurements Thin Wire
Coil Input Output Input Input Input Comment

do [mm] 9.26 9.35 9.26 9.26 9.26 OD of coil
di [mm] 4.5 4.5 4.5 4.5 4.5 ID of coil
b [mm] 5 5 5 5 5 axial length of coil
c [mm] 2.38 2.425 2.38 2.38 2.38 winding build of coil

N 566.5 359.6 1187.5 1888.3 3002.5 number of turns
dw [mm] 0.116 0.147 0.080 0.063 0.050 bare wire diameter
AWG 36.8 34.8 40.0 42.0 44.0 wire AWG

R [ohm] 20.1 8.0 88.3 223.3 564.6 Coil resistance

La [uH] 1337.5 542.4 5876.7 14858.2 37565.9 Air-cored inductance, no core
Lm [mH] 7.22 2.91 31.8 80.3 203 Inductance with core - lower limit (1)
Lm [mH] 10.8 4.37 47.6 120 304 Inductance with core - rough estimate (2)

Notes:
(1) Lower limit unlikely
(2) Wouldn't be surprised if 1.3 times this


The pole piece is definitely just a steel slug. The ferrite ring is attracted to a magnet but is not itself magnetic. The two magnets are most definitely magnets! 

I don't doubt it, it's just when I read the patent I was thinking the pole-piece was the magnet.

FWIW, the description of the amplifier connections in the patent are unclear, almost deceiving.  The schematic lines aren't even clear.   A generic amplifier doesn't have so many connections, so it's specific, without explaining what the connections do.

Just for fun I was thinking about setting up a test rig for my scope to see at what frequency those ferrite rings start to saturate.

Inductor saturation will depend on both the frequency and the current.  I can actually calculate some saturation levels but the thing that screws it up is the permanent magnet.   If the permanent magnet is strong it could already be saturating the ferrite ring - that's assuming it is ferrite and not some ferrous magnetic material with a coating.   Since there is a large airgap at the front of the assembly, between the pole-piece and the edge of the ring, it might be large enough to prevent saturation.   Without the fine details of the magnet it's hard to know.

The magnet may have some more quirks:   The common button magnets have the north/south poles on the flat faces but just banging the north pole (say) at the back of that ring and pole-face means both faces are north.   You will get a field from between the pole and the edge of the ring but it's like a weak incidental field.   Maybe it's done deliberately to prevent the ferrite ring saturating!   An alternative is the button magnet is magnetized radially, that will force the pole-piece to be say north and the edge of the ring to be south, making the field at the front of the transducer much stronger.  (FYI, do a search on this button magnet and you might see the differences in the field patterns they are offered in.

Plenty of details to sort out really!

typo's fixed.

[Paul Marossy]

Well I tried to match the new dimensions with the new resistance.  The output coil looks fine.  The wire size for the input coil is 37AWG (0.12mm diameter).  Thinner than before (as expected) but it's not a enormous difference and not anywhere near 0.05mm.

So what I did instead was see what resistances ended-up with 40AWG, 42AWG, 44AWG wires.   As you can see the resistances are very much larger than 20.1 ohm.   The wire is worked out by looking at how many turns are required to fill-up the measured winding cross-section.

I've done this type of matching many times before.  For accurate results it takes a lot of effort getting all the dimensions right.  Measuring the wire diameter is always a little tricky due to the enameled coating and kinks.  On top of that you still have to make a stab at how well the winding is packed.   Normally I expect to be within 1 or two wire gauges and I often kick the estimate using the resistance and inductance measurements.   It's also easy to screw-up the calculations.   I did verify the calculation against some other software and the agreement was very reasonable, so I'm think the calculations are correct.

You are probably correct in your assessment, Rob. I'll defer to your expertise here. That wire is extremely small. Makes me wonder how you can wind something like that without the wire breaking!

Inductor saturation will depend on both the frequency and the current.  I can actually calculate some saturation levels but the thing that screws it up is the permanent magnet.   If the permanent magnet is strong it could already be saturating the ferrite ring - that's assuming it is ferrite and not some ferrous magnetic material with a coating.   Since there is a large airgap at the front of the assembly, between the pole-piece and the edge of the ring, it might be large enough to prevent saturation.   Without the fine details of the magnet it's hard to know.

Since the coils are too hacked up to test them as a unit, like I have done with a 500mH inductor for example, I tested just the "ferrite" ring itself using an old school method I saw in an oscilloscope book I have (Handbook Of Oscilloscopes - Theory And Application). I couldn't get it to show any signs of saturation or hysteresis, going thru a very wide range of frequencies. They seem to be very neutral as far as anything an Ebow would encounter in normal use is concerned. I'm wondering what purpose they actually serve...

I don't doubt it, it's just when I read the patent I was thinking the pole-piece was the magnet.

FWIW, the description of the amplifier connections in the patent are unclear, almost deceiving.  The schematic lines aren't even clear.   A generic amplifier doesn't have so many connections, so it's specific, with explaining what the connections do.

Yeah, that schematic in the patent is whacked. There's a few things that don't make much sense when you study it. The pole pieces are in a similar arrangement to how the Fender MIM single coil pickups are made - the six pole pieces are not magnets, they are just steel slugs with two bar magnets on two sides of them at the bottom of the pickup.

The magnet may have some more quirks:   The common button magnets have the north/south poles on the flat faces but just banging the north pole (say) at the back of that ring and pole-face means both faces are north.   You will get a field from between the pole and the edge of the ring but it's like a weak incidental field.   Maybe it's done deliberately to prevent the ferrite ring saturating!   An alternative is the button magnet is magnetized radially, that will force the pole-piece to be say north and the edge of the ring to be south, making the field at the front of the transducer much stronger.  (FYI, do a search on this button magnet and you might see the differences in the field patterns they are offered in.

Huh I wasn't aware you could magnetize a magnet radially. They seem to be a little strong for their size. They can pick up a 9v battery (I weighed it, it's 1.6 ounces).

[Eb7+9]

yeah, real scope ...
wondering if waveshape you're showing is affected by input (scope) cap when set to "AC coupled" mode ..

Yeah mean actual scope, or virtual scope?

[Rob Strand]

You are probably correct in your assessment, Rob. I'll defer to your expertise here. That wire is extremely small. Makes me wonder how you can wind something like that without the wire breaking!

Have many turns of the thin wire doesn't seem unreasonable. While have some confidence in the calculations, the fact it doesn't match measured resistance means something is wrong!   Being out by such a large factor might mean the coil has been damaged and there is a short within the winding.   That would explain a lot.   In order to lower the resistance so much it would have to be a short from the leadout wires to one of the coil layers.  A short between layers won't explain it.   When a short is present the inductance drops quite a lot - because the shorted section of the coil acts like a shorted transformer.   Also, the short means the the coil won't act as very good sensor.

Yes, that thin wire is nuts.  Once you get thinner than 38AWG you need machines.   I've repaired many headphones and the wire in those is insanely fine.   When I look at $3 headphones they use *very* thin wire on an equally very thin plastic bobbin and it makes you think about how much knowledge under the hood there is to make those things for $3.

Since the coils are too hacked up to test them as a unit, like I have done with a 500mH inductor for example, I tested just the "ferrite" ring itself using an old school method I saw in an oscilloscope book I have (Handbook Of Oscilloscopes - Theory And Application). I couldn't get it to show any signs of saturation or hysteresis, going thru a very wide range of frequencies. They seem to be very neutral as far as anything an Ebow would encounter in normal use is concerned. I'm wondering what purpose they actually serve...

When you test for saturation it needs to be in the structure it is used and with the same coil.   If you have a coil wound on a ferrite core with 1 turn it might handle 1A before saturation but when you have 1000 turns it will only handle 1mA.   Then if you add an air-gap to the magnetic path the inductance of the 1000 turn coil will drop but the current before saturation increase to say 10mA to 100mA depending on the size of the air-gap.

I did some rough estimates and the output coil will saturate at around 1.5A.  You don't want to push it that hard though.  That drive capacity is probably on the money for the ebow.   An interesting thing showed up when I did this, the small diameter pole-piece needs to be an iron material in order to prevent the saturation.  If it was ferrite the coil would saturate at a much lower current (the other way to say this is if the material was ferrite you would need a larger diameter pole piece).

Yeah, that schematic in the patent is whacked. There's a few things that don't make much sense when you study it. The pole pieces are in a similar arrangement to how the Fender MIM single coil pickups are made - the six pole pieces are not magnets, they are just steel slugs with two bar magnets on two sides of them at the bottom of the pickup.

Something else about that patent is the comments about harmonic mode.   When I read the patent I read harmonic mode to mean they would flip the polarity of the output coil, or perhaps wire the output coil to the +ve rail instead of ground.

If you imagine under normal circumstances the output coil is driven with the same polarity as the input coil and that creates a positive feedback loop.   When the string vibrates at the fundamental the middle part of the string vibrates entirely up/down/left right.   If you reversed the direction of the output coil it would not create such a positive feedback loop.    However strings a tricky because they can vibrate at harmonics which means different positions of the string can be up and down,



So if you drive the string with the "wrong" polarity you can still find a harmonic with positive feedback.   The wavelength  will the twice the spacing between the sense coil and the drive coil - so that coil space is important for this mode.   Given the close spacing of the coils that's going to be a high harmonic, and maybe that's where the problems come in.

However the way the harmonic mode is implemented in the circuit seems to differ from the implementation in the patent.   Maybe they had problems getting the harmonic mode (as described in the patent) to work well in practice.   

Huh I wasn't aware you could magnetize a magnet radially. They seem to be a little strong for their size. They can pick up a 9v battery (I weighed it, it's 1.6 ounces).

It could very well be a ferrite magnet.   Yes, in the old days, before you buy 50 magnets for $10, there were *many* options for magnet polarization (See Philips permanent magnet handbooks).   You could really build things for a specific job.    These days the main type is axially polarized: north-front, south-back.   However you can find some cheap button magnets with diametrically polarized magnets (left-right).   The diametrically polarized magnets are a headache for production since if you cared the field was along the strings or across the strings  then you would have to place the magnets with the correct orientation.

FWIW, I'm not saying using axially polarized magnet is wrong or bad.  It could even work out better in practice.   All I'm saying is from a back-engineering perspective there's other options.

[anotherjim]

Is it possible the sense coil has a short between turns, after all, it was from a faulty unit.

Is the ferrite there to aid immunity to RF?

Oh, and I saw the Tek scope inputs were AC coupled in the video.

[Rob Strand]

Is the ferrite there to aid immunity to RF?

I started looking at this thing with the aim to try to back-engineer some of the unknowns through devious means.  I wrote a whole heap of analysis and thoughts but I haven't got back to it.  It keeps growing.  Initially I just wanted to give Paul some estimates to cross-check the coil resistances but it's blown out to bigger issues!

To summarize a few thoughts:

That shielding would only make sense for the receive coil.   Maybe hum prevention.  However since we aren't listening to the receive coil I suspect you could handle quite a bit of hum.

The shorter magnetic path around the coil (due to the outer ring) sets the inductance.   This may be the driving factor for the ring, especially say resonating with the 33nF cap to improve receive gain.   When I looked at this I noticed the Q is very high.  The 13k resistor doesn't look bad but it could be lower.    Not having a reliable resistance or inductance for the receive coil put a spanner in the works - sort of concluded you would need to build something and tweak it.   Then make sure it worked in normal and harmonic mode.

I haven't finished looking at the effect of using an simple axially polarized magnet *and* having the drive/receive coils with outer rings.   What I was doing is looking at the magnet as creating a magnetic field through the string then that moving "magnet" is inside the gap field between the pole-piece and the outer ring.   The "magnet" in the string points point in one direction to left of the coil center and in the other direction to the right of the coil center.   I don't even know it that will add anything as a motor/sensor - if not the outer ring would *only* be used to set the inductance.

[Jarno]

That shielding would only make sense for the receive coil.   Maybe hum prevention.  However since we aren't listening to the receive coil I suspect you could handle quite a bit of hum.

Interesting thread this!
Was going to mention the site by Brian on the Sustainer, but he stepped in himself

On the shielding, I think it does make sense to design the drive coil in such a way, at least for the sustainer, to make sure the driving field is aimed at the strings and not too much to the sides or the receive pickups will pick it up. And for that, some kind of mu metal shielding might make sense, to direct and shape the field. So it is not shielding in the traditional sense, to prevent disturbances from coming in, but rather from preventing them from going in directions you do not want them to go.

[Paul Marossy]

Have many turns of the thin wire doesn't seem unreasonable. While have some confidence in the calculations, the fact it doesn't match measured resistance means something is wrong!   Being out by such a large factor might mean the coil has been damaged and there is a short within the winding.   That would explain a lot.   In order to lower the resistance so much it would have to be a short from the leadout wires to one of the coil layers.  A short between layers won't explain it.   When a short is present the inductance drops quite a lot - because the shorted section of the coil acts like a shorted transformer.   Also, the short means the the coil won't act as very good sensor.

Yes, that thin wire is nuts.  Once you get thinner than 38AWG you need machines.   I've repaired many headphones and the wire in those is insanely fine.   When I look at $3 headphones they use *very* thin wire on an equally very thin plastic bobbin and it makes you think about how much knowledge under the hood there is to make those things for $3.

Yeah that will be my next headache... rewinding the input coil. I rewound the output coil by hand yesterday using wire from a trashed inductor. I initially was just going to look at it with the curve tracer "octopus" I made for such things. When I was done with my third or fourth attempt I finally got it wound without breaking the wire. I measured it and it was exactly 8 ohms before I put it back on that mangled PCB! Pure luck!  Still couldn't get it to show any signs of hysteresis or saturation in the audio frequency realm.

I also pulled off the broken diode and the believed to be 13K resistor and put sockets in their place. So it appears that now I will be attempting to do a resurrection of this thing to see if I actually got everything right. My main obstacle will be rewinding that input coil. 

When you test for saturation it needs to be in the structure it is used and with the same coil.   If you have a coil wound on a ferrite core with 1 turn it might handle 1A before saturation but when you have 1000 turns it will only handle 1mA.   Then if you add an air-gap to the magnetic path the inductance of the 1000 turn coil will drop but the current before saturation increase to say 10mA to 100mA depending on the size of the air-gap.

I did some rough estimates and the output coil will saturate at around 1.5A.  You don't want to push it that hard though.  That drive capacity is probably on the money for the ebow.   An interesting thing showed up when I did this, the small diameter pole-piece needs to be an iron material in order to prevent the saturation.  If it was ferrite the coil would saturate at a much lower current (the other way to say this is if the material was ferrite you would need a larger diameter pole piece).

This book shows a method for testing just a piece of ferromagnetic material using a signal generator, two metal plates and a capacitor. Doesn't really tell you what kind of voltage you should be testing it with but I will revisit this in a day or two.

Something else about that patent is the comments about harmonic mode.   When I read the patent I read harmonic mode to mean they would flip the polarity of the output coil, or perhaps wire the output coil to the +ve rail instead of ground.

That was what I originally thought it would be doing... but it's far stranger than that.

If you imagine under normal circumstances the output coil is driven with the same polarity as the input coil and that creates a positive feedback loop.   When the string vibrates at the fundamental the middle part of the string vibrates entirely up/down/left right.   If you reversed the direction of the output coil it would not create such a positive feedback loop.    However strings a tricky because they can vibrate at harmonics which means different positions of the string can be up and down,

So if you drive the string with the "wrong" polarity you can still find a harmonic with positive feedback.   The wavelength  will the twice the spacing between the sense coil and the drive coil - so that coil space is important for this mode.   Given the close spacing of the coils that's going to be a high harmonic, and maybe that's where the problems come in.

However the way the harmonic mode is implemented in the circuit seems to differ from the implementation in the patent.   Maybe they had problems getting the harmonic mode (as described in the patent) to work well in practice.

That could very well be the case. Maybe it was also something that was hard to make happen consistently due to variances in parts, etc.

It could very well be a ferrite magnet.   Yes, in the old days, before you buy 50 magnets for $10, there were *many* options for magnet polarization (See Philips permanent magnet handbooks).   You could really build things for a specific job.    These days the main type is axially polarized: north-front, south-back.   However you can find some cheap button magnets with diametrically polarized magnets (left-right).   The diametrically polarized magnets are a headache for production since if you cared the field was along the strings or across the strings  then you would have to place the magnets with the correct orientation.

FWIW, I'm not saying using axially polarized magnet is wrong or bad.  It could even work out better in practice.   All I'm saying is from a back-engineering perspective there's other options.

If that's a ferrite magnet it is very weak. Does not appear to be attracted to anything ferrous whatsoever.


AND NOW FOR AN UPDATE: I have discovered a quirk in the LM386 SPICE model I am using. I was looking at the voltages at Pins 3 & 5, and with 500mV at the input the program was simulating something like 140+ volts at the output!  I don't know if this is the right tactic to get around that problem but I put a voltage divider on the output to bring that down to 9V-ish. Now when I simulate it using 110 ohm/4mH input coil & 8 ohm/3mH output coil I see waveforms that really start to make much more sense. I also tried to simulate couple of the circuits shown in the LM386 application notes, and some won't even do anything - like the Weinbridge Oscillator, it does nothing (I think I need to create a light bulb model and try again). THEN, I saw in the LM386 application notes that there is also a square wave oscillator! I did not know this. Guess what that has in common with the Ebow circuit - Pins 3 & 5 connected together via a 10K resistor! I modeled this too and it made beautiful square waves. This made me think that maybe this Ebow is some kind of adaptation of that square wave oscillator. Kind of makes sense in my mind as I would think the output coil would be seeing a square wave-ish waveform to keep the string vibrating vigorously. Seems to also make sense that those little pads underneath the 220uF cap would be for testing in a test fixture. Probably looking for a certain waveform to be there at all times?

[Paul Marossy]

On the shielding, I think it does make sense to design the drive coil in such a way, at least for the sustainer, to make sure the driving field is aimed at the strings and not too much to the sides or the receive pickups will pick it up. And for that, some kind of mu metal shielding might make sense, to direct and shape the field. So it is not shielding in the traditional sense, to prevent disturbances from coming in, but rather from preventing them from going in directions you do not want them to go.

I was thinking the same thing this morning. Maybe it's to "focus" the energy towards the string, preventing "cross talk" with other nearby things. And may also have a secondary effect of providing shielding from EMI/EMF.

[Ben N]

Your sustained effort is appreciated, Paul.