Hum-cancelling with a backplate coil.

[JohnH]

Has anyone here built a noise-cancelling back-plate into a guitar? I've been thinking about these recently, it seems like a really good concept. It was invented by Ilitch Electronics, who market them. I think Surh also included them under a licence. It's not particularly cheap, but it looks like a high quality product. If it works, then you can use your favourite single coil pickups with no noticeable change in tone, and no hum. That would be worth having.

Ilitch Electronics

Noise Cancelling with single coils - stacked coils

The basic idea of noise-cancelling coils is to put a coil in series with the pickup, that does not sense the string vibrations significantly, but which picks up hum in opposite phase to the main coil, and so cancels it out. So in principle, you get the sound from a single coil and the hum cancelling performance of a humbucker.

This idea is used in most stacked 'single coil' noiseless pickups, and it works. But the extra impedance of the hum-cancelling coil changes the response, adding inductance and resistance and so tending to dull the response. This seems to be a common complaint about them, from those who prefer the traditional sparkle of a classic simple single coil pickups.

'Back-plate' noise cancelling coils

The idea of the back-plate system is that it is a relatively large coil in terms of area. The signal induced in a coil by varying magnetic flux is dependent mainly on area x turns. So to pick up the same hum signal, from a larger area of coil, fewer turns are needed by the ratio of width x depth, or diameter squared. And total wire length is also reduced. With a physically wider coil of fewer turns, a thicker wire can be used, so total resistance can be reduced greatly. Add to that, a wider coil in air, with no core and fewer turns, has a much lower inductance, measured in milli-Henrys, instead of 2 to 3 Henrys as per a typical single coil pickup.

It needn't be a back-plate as such, it just needs to be a relatively large coil with its axis parallel to that of the pickups.  So in principle, it could also be under the pick-guard for example.

All of that means that we can have such a coil in series with the pickups with hopefully a negligible change to electrical (and hence tonal)characteristics, but still have it pickup up enough 'anti-hum' to quieten down the guitar.

What I'm thinking of is a couple of things:

1. Try to work out what the characteristics of such a coil needs to be, and how it might affect the tone if at all.
2. Work out ways to switch it into a 'nutzy' wiring design, with various series, parallel and blended options, without to much wiring craziness and all hum cancelling, using the pickups on their own where they make humcancelling combinations, and including the anti-hum coil when needed.

J 

[JohnH]

A bit more on this:

If we say that a typical single coil has average dimensions of about 10mm x 60mm, with about 7500 turns of 42gage wire, then, to create a larger coil of say 100mm x 140mm that generates the same amount of hum, the number of turns would be approximately:

7500 x (10 x 60) / (100 x 140) = 321

Let's say that instead of 42 gage wire, we used something thicker, say 34 gage, at around 0.86 Ohm per m, then the dc resistance will be around 0.86 x 321 x 2x (0.1 + 0.14) = 132 Ohms, ie not too much.

Using 34 gage wire, with those turns and dimensions, an online calculator gives 57mH as inductance.

Then I added those Ohms and Henrys to the typical values for a lower output single coil pickup, using 5.5 and 2.2H, using GuitarFreak, and found that there was no output differences more than about 0.2db at any frequency.

So I think these large noise cancelling coils are worth investigating further. The exact number of turns would need adjusting by tests though, to get the best cancellation

[sumgai]

Jun 29, 2014 15:28:49 GMT -5 JohnH said:

A bit more on this:The exact number of turns would need adjusting by tests though, to get the best cancellation

An added advantage to using "larger" diameter wire: you open up the possibility of "tapping" the coil with additional leads, such that you can then derive the best humcancelling position for each pickup, and not have to compromise with just the two ending leads. Selectable via the main pickup switch(es), of course.

HTH




sumgai

[newey]

Since the coil doesn't sense the strings, could it not be pretty much any shape or size and still cancel the hum? The ability to pick up the hum is a function of the length of the wire, no? So, what's it matter whether it's in a ring shape or not?

Or does the resulting field's shape/size dictate the degree of hum cancellation?

[JohnH]

Jun 29, 2014 15:55:24 GMT -5 sumgai said:

Jun 29, 2014 15:28:49 GMT -5 JohnH said:

A bit more on this:The exact number of turns would need adjusting by tests though, to get the best cancellation

An added advantage to using "larger" diameter wire: you open up the possibility of "tapping" the coil with additional leads, such that you can then derive the best humcancelling position for each pickup, and not have to compromise with just the two ending leads. Selectable via the main pickup switch(es), of course.

HTH




sumgai

Absolutely! Thinking there would need to be some tests to check the best 'ball-park' values for number of turns, then over-wind the coil and set up taps at say 0, 15% and 30% above and below the best guess. Trim pots between taps can further fine tune, and Ilitch has a couple of these in his system.

The extent to which each pickup setting is fine tuned for hum would depend how much switching is adopted and how similar the pickups are. As a base case, the ability to set to the match the M pup on a matched set, and use that for all, even though B and N are a bit higher or lower, would still give a good benefit.

[JohnH]

Jun 29, 2014 16:18:58 GMT -5 newey said:

Since the coil doesn't sense the strings, could it not be pretty much any shape or size and still cancel the hum? The ability to pick up the hum is a function of the length of the wire, no? So, what's it matter whether it's in a ring shape or not?

Or does the resulting field's shape/size dictate the degree of hum cancellation?

Yes I think shape does not matter; just area and turns. It could be circular, square or pick-guard shaped. Bigger area needs less turns, and also less total wire, since doubling diameter doubles length but gives 4x area.

[newey]

I was thinking of studiostriver's unused 9-volt battery nacelle, and all the other folks who pull active electronics out of their guitars. Could make it the size/shape of a 9V battery.

Or perhaps the switch frame for a P/P pot could be used as the bobbin, with windings around it. The whole thing could then just replace a pot.

[JohnH]

Jun 29, 2014 16:32:27 GMT -5 newey said:

I was thinking of studiostriver's unused 9-volt battery nacelle, and all the other folks who pull active electronics out of their guitars. Could make it the size/shape of a 9V battery.

Or perhaps the switch frame for a P/P pot could be used as the bobbin, with windings around it. The whole thing could then just replace a pot.

Good thought. But for this to work as well as it should, it needs to be a physically wide coil, since that is the key to getting very low resistance and inductance out of a coil that is picking up the right amount of hum. But ill run some numbers based on your ideas.

Another version that is interesting is to take a smaller coil such as you suggest, and run it through an active buffer to drop its impedance right down. Then use the output of that as a ground reference for the pickups, so that they are being fed 'anti-hum' tbrough the active circuit.

[JohnH]

I had a look at the likely properties of a smaller noise cancelling coil, wound in the shape of 9v battery, about 50mm x 30mm.

Being smaller, the turns need to go up to 2400 and it would be around 400 ohms. Most significantly though, the calculated inductance goes up from around 60mH (for the wider back-plate coil), to 1H. This in series with pickups would definately affect tone. So a wide hum-cancelling coil is definately the key.

[newey]

Aha! Well, scratch that idea . . .

[sumgai]

Shape does indeed play a part in the calculations. Specifically, cross-section is important to the inductance. A rectangular or ovoid "coil" will not have the same operating characteristics as a circular one, given the same length of wire running around the perimeter.

This is primarily true for air core coils. Adding an iron (or other metal) core can, and usually does, obviate that statement. One need only look at our pickups to get an understanding of this notion.

As usual, Google is your friend. Unfortunately, the first several results concern themselves more with using a coil as either an actuator or a generator, so some digging is required. However, the math on those pages is relatively simple (well, most of the time!), so the idea of shape still stands - no matter whether the coil is a sensor like our pickups, or the reverse, the idea of shape still plays in important part.

Oh, and BTW, if you really want to waste some time, check out a Mobius coil - really! ;D

HTH





sumgai

[santellan]

What about a multilayer circuit board in the shape of a pick-guard or inserted into a machined recessed cavity in the pick guard. The coils could then be centered around each pickup or around all 3 pickups which is what Ilitch is doing with his kit. Can your calculator come up with the trace size and winding counts? 

[JohnH]

My belief is that, for the purpose here of picking up a required amout of hum signal, based on flux through the coil, the main parameters are area and turns, and shape and wire length are secondary to that. A non cicular coil of a given area needs more wire length due to greater perimeter/area.

Given two coils of equal turns and area, inductance varies a bit. Taking the battery-sized rectangular coil considered above (2400turns, 50x30 and 1H), a circular coil of equal area (43.7mm diameter) has an inductance of 0.9H.

I was taught this stuff 35 years ago, and the theory of it was a key factor in my not becoming an electrical engineer! But, thanks to the internet, at least there is some help. There are lots of online calculators, Ive been using this one:

Inductance calculators

[JohnH]

I've been exploring how a low impedance hum-cancelling dummy coil could be wired into some typical guitar wiring designs, and some more nutzoid ones.

The basic idea is to add the dummy coil in series, and with three single coils the simplest method conceptually is like this:

3x singles, non RWRP



There would be no RWRP coil needed, since the dummy coil takes care of all that. In the diagrams, I'm using '+' to indicate positive tone phase, and 'H' to show a nominal hum direction. To get hum cancelling , the path from hot to ground needs to go through coils with opposite hum direction.

3x singles, with RWRP

But most single-coil pickup sets have an RWRP coil. With these, we would need to feed the series connection to the dummy coil in the opposite direction with M, as compared to B and N. Also, when in Strat positions 2 and 4, the pickups are already quiet and we don't want the dummy coil adding its own extra hum. This needs some more switching, but the second half of a standard Strat 5-way could do the job:



It conveniently shunts the dummy coil in positions 2 and 3.

Coil splits and series blending

This is the other main idea I want to explore. My belief (unproven in practice), is that with a diagram like this, one can gradually shunt one coil of a series combination or humbucker, and maintain humcanelling through the whole sweep. When the variable resistor is at max, the dummy is not active, and when at min, the anti-hum is still present, but the tone signal from the bypassed pickup coil is now shunted by the very low impedance of the dummy coil:



Reckon that would work?

I want to mash the last two diagrams together to make some SSS and HSS options with blending and coil shunting.

cheers

John

 

[JohnH]

Here is a schematic for a basic HSS circuit, using the low-impedance hum-cancelling coil, labelled 'HC':



If this worked as I think it might, it would address the first few items on a typical wish list for an HSS set-up:

• Able to cut the humbucker to a single coil
• A fully coil-cut bridge humbucker can sound a bit to thin, so a blend/spin-a-split control to allow in-between settings
• Position 2, B+M can be with full humbucker or the split bridge, or in-between 
• All settings and control positions noise cancelling, but with the natural tone of simple single coils available, due to the low-impedance HC coil.

Since the 5-way switch is fully used to control the HC coil, it is not available for switching tone pots. Ideally this circuit would have just a  simple master volume and tone (not shown), also since one knob is now a variable coil-cut knob.

Side note: The exact characteristics of the HC coil would need to be tuned to get optimum overall results, and would be a best compromise between balancing with the various coils.  A trim pot could make fine adjustments, or with a super-switch, possibly different trim settings for each position based on partly bypassing part of the HC coil. But I quite like the idea of keeping it simple, and I think a worthwhile benefit can be obtained even without a perfect match. For example, on my current HSS strat, I have a neck single of about 6k being able to go in series with a bridge Hb coil of about 4k. This is still able to cut down most of the hum, possibly to about 1/3 which is a -10db reduction. On the diagram above, the HC coil could be tuned to be between a single pickup and one Hb coil, for an even better result. All speculation though! 

[newey]

This is great stuff, John!

Now, I was with you up until the last diagram, where the pot is blending the dummy coil. Doesn't the diagonal arrow represent the wiper? And doesn't the wiper need to be connected to something to vary the resistance across the pot?

[JohnH]

Jul 4, 2014 20:47:50 GMT -5 newey said:

This is great stuff, John!

Now, I was with you up until the last diagram, where the pot is blending the dummy coil. Doesn't the diagonal arrow represent the wiper? And doesn't the wiper need to be connected to something to vary the resistance across the pot?

Thanks and yes, that is slight graphical laziness, to represent a pot wired as a 'variable resistor', using just two lugs being the wiper and an outer lug. But I believe it is still a valid circuit symbol for such a thing.

The next one I have in mind is a fully humcancelling version of this SSS design, which I wired up for my son.:

Strat SSM3 

It does the basic settings plus N+B combinations and blends up towards series combos by adding M to N, B or B+N. I think it can be adapted with the HC coil to do all that with low noise.

[newey]

"Graphical laziness" . . . I got it now.

Are you planning to actually build this into your son's SSM3? If so, full coil-building reports will be anticipated!

[JohnH]

Jul 4, 2014 21:35:56 GMT -5 newey said:

"Graphical laziness" . . . I got it now.

Are you planning to actually build this into your son's SSM3? If so, full coil-building reports will be anticipated!

Probably not on that specific one, since he is currently over the water working in New Zealand. Also, this version will need a three pole switch to do series/parallel while keeping control of he dummy coil. I'm thinking it would be a good candidate for a Fender S1 switch, which have 4 poles and are now a lot easier to get than they used to be.

But the first step, having now concluded that the theory has merit, would be to try to make one of these coils and test it out on a couple of guitars in a non-destructive way .I  think for a test rig, it could be wired fully outside of the guitar and connected at the jack between guitar ground and cord/amp ground. That way it could be tested with several guitars. 

[JohnH]

This is how I was thinking an SSS Strat could be wired, using the dummy coil 'HC' to get hum cancelling in all positions, including a blendable series mode and an N+B option.

It uses a standard Strat 5-way switch, and a 3-pole 2-position switch for mode changing. A 3pdt toggle could be used, or I think a Fender S1 would be a cool option for this. There is a blend control, for bending in series Mode 2, which can be preset for Mode 2 since it is not used in Mode 1.  Simple master volume and tone pots would be used (Not shown, I'd suggest 250k with a no-load tone pot)

In Mode 1, as shown, the standard Strat options are provided, fed through the HC coil in the single positions 1, 3 and 5 so they become humcancelling. In positions 2 and 4, the HC coil is bypassed since the pickups together are naturally noise free if M is RWRP.

In Mode 2, M drops down to go into series with B and N, and as drawn here, B replaces M so you get N, B+N, B, B, B. This gives the missing B+N option which is not usually offered, and it is not usually humcancelling.  In Mode 2 the variable resistor starts to add series resistance to the HC coil, allowing the M pickup to add its tone in series, so we get NxM, (B+N)xM and BxM, as a blendable progression from N, B+N, B. Hopefully, this will be approximately humcancelling at each position of the pot. With regard to hum, N and B will be each generating the same +hum voltage. HC and M will each make the same 'anti hum', and this should remain there as the pot is varied.

The circuit could equally well have B and N swapped in position, to give B, N+B, N, N, N in mode 2, blendable with M in series. Which is better will depend on which settings are favoured, and which one-switch transitions are most useful.  

John

[col]

Jun 29, 2014 15:28:49 GMT -5 JohnH said:

A bit more on this:

If we say that a typical single coil has average dimensions of about 10mm x 60mm, with about 7500 turns of 42gage wire, then, to create a larger coil of say 100mm x 140mm that generates the same amount of hum, the number of turns would be approximately:

7500 x (10 x 60) / (100 x 140) = 321

Let's say that instead of 42 gage wire, we used something thicker, say 34 gage, at around 0.86 Ohm per m, then the dc resistance will be around 0.86 x 321 x 2x (0.1 + 0.14) = 132 Ohms, ie not too much.

Using 34 gage wire, with those turns and dimensions, an online calculator gives 57mH as inductance.

Then I added those Ohms and Henrys to the typical values for a lower output single coil pickup, using 5.5 and 2.2H, using GuitarFreak, and found that there was no output differences more than about 0.2db at any frequency.

So I think these large noise cancelling coils are worth investigating further. The exact number of turns would need adjusting by tests though, to get the best cancellation

John,

What about a small dummy coil (say, about a sixth of the area of the pup coil), but using higher gauge wire, and wired in parallel. By my calculation, 50 gauge wire would be about 6.3 times more resistive than 42 gauge, and 56 gauge would be about 26 times more resistive. If the average dummy coil loop length is just a sixth that of the pup coil, the dummy coil would need to be wound about six as many times (using your figures, that's about 45,000 loops).

50 gauge: 6.3 (more resistive) x 6 (as long), results in a dummy coil which is about 38 more resistive than the pup coil (I expect, akin to having a pot parallel with the pup at lower frequencies; and no real effect at higher frequencies). For the 56 gauge wire, the dummy would be 156 times the resistance of the pup coil (must have negligible - even inaudible - effects upon tone and volume).

The problem, of course, is sourcing such coils. I imagine they would be near impossible to produce (especially using the 56 gauge wire); I can't even find anything using 50 gauge wire. But this would make for relatively small dummy coil.

Your thoughts?

[JohnH]

I like your thinking, but I dont like its chances of working. A dummy coil producing the same level of anti-hum as tbe main coil, but with many times higher impedance, would have its efforts crushed to negligibility by the main coil.

Some days I know how that feels...

[col]

Sept 9, 2016 4:10:05 GMT -5 JohnH said:

I like your thinking, but I dont like its chances of working. A dummy coil producing the same level of anti-hum as tbe main coil, but with many times higher impedance, would have its efforts crushed to negligibility by the main coil.

Some days I know how that feels...

So why is this not a problem for the dummy coil described in the opening post of this thread? Why isn't the anti-hum of the dummy coil "crushed" by the hum produced by the main coil (with its much higher impedance)?

[JohnH]

Sept 9, 2016 12:09:43 GMT -5 col said:

Sept 9, 2016 4:10:05 GMT -5 JohnH said:

I like your thinking, but I dont like its chances of working. A dummy coil producing the same level of anti-hum as tbe main coil, but with many times higher impedance, would have its efforts crushed to negligibility by the main coil.

Some days I know how that feels...

So why is this not a problem for the dummy coil described in the opening post of this thread? Why isn't the anti-hum of the dummy coil "crushed" by the hum produced by the main coil (with its much higher impedance)?

Because there, the two coils are in series, and the output from such series wiring is the sum of the two parts, independent of their impedances.

But in parallel, say with two coils where they would separately produce the same output, but one is 40x the impedance of the other, the new output will be 1/41 of one coil and 40/41 of the other coil.

[col]

Thanks for the explanation, John.

[nikogo]

Great idea.
I think would it be a good place for the hum cancelling coil to hide it under the edge binding on LP?
The area would be maximal - all front projection of a body.

[pablogilberto]

Halu!

I'm interested in doing some projects and experiments on this.

I have some questions to help me with my experimental setup:

#1
How do we measure amount of hum picked up coils?
I'm thinking of measuring them using a oscilloscope or something? 

I think that matching the amount of generated hum is helpful in identifying the dummy coil specs (wire gauge, shape, turns, etc) that will properly work(cancel)

#2
If a simple experimental setup is not possible, do you know any simulator that can do this?
Or computation by hand at the least?


#3 
I've checked with other noise cancelling coils using the same principle, what I'm thinking is that, say traditional SSS Strat (same polarity)
For Position 1, 3 and 5, the hum will be cancelled.
BUT what happens in position 2 and 4?

2 coils are active and connected in parallel.
This means that the hum generated is doubled?

Will the dummy coil still be effective?


Thanks in advance for the help!

[JohnH]

Mar 17, 2022 21:59:50 GMT -5 pablogilberto said:

Halu!

I'm interested in doing some projects and experiments on this.

I have some questions to help me with my experimental setup:

#1
How do we measure amount of hum picked up coils?
I'm thinking of measuring them using a oscilloscope or something?

I think that matching the amount of generated hum is helpful in identifying the dummy coil specs (wire gauge, shape, turns, etc) that will properly work(cancel)

#2
If a simple experimental setup is not possible, do you know any simulator that can do this?
Or computation by hand at the least?


#3
I've checked with other noise cancelling coils using the same principle, what I'm thinking is that, say traditional SSS Strat (same polarity)
For Position 1, 3 and 5, the hum will be cancelled.
BUT what happens in position 2 and 4?

2 coils are active and connected in parallel.
This means that the hum generated is doubled?

Will the dummy coil still be effective?


Thanks in advance for the help!

I think this is still a good idea. There's probably a few variables that are hard to predict though, so for development by testing, I think you could make a best guess and then adjust from there.

The real and dummy coils go in series.

The basic idea is that the amount of hum in a given environment would be based on area of coil x turns. And a bigger area for the dummy is best. So if you have a pickup of say 60 x15mm and 4000 turns, and your dummy will be say 200 x 150, it's area is 33 x greater so might need 4000 ÷ 33 = 120 turns. But what would the core of the pickup do to its affective area? Maybe it will suck in more hum. But if you made a first-pass prototype like that, you could test with the scope compared to a pickup in the same place and same orientation. Then you could estimate a ratio to change the number of turns by for a second test.

Once you get close, you could put them in series and see if you can get cancellation when phased appropriately.

And if you can set a case with slightly too much dummy, a pot across it could trim it down for a null.

[dougied]

Certainly numerous patents referenced in the Ilitch patent:

patents.google.com/patent/US7259318B2/en

As long as it's for personal use, you're free to fabricate someone else's patent.