Advanced PI coils

[Davor]

Sure Davor, when I have more time. I already analysed such a coil configuration in the past and it didn't give the benefit. It isn't easy to build too.
The flux surface area of A(RX+) - 2*A(RX-) = 0 works only, when the magnetic field in it is homogenous. But it isn't homogenous and therefore you would need much more winding turns at the inner RX coil (RX-).
We could make the TX much larger and focus to the center region, where the magnetic field is nearly homogenous. It could then work a bit but not as perfect as our tophat coil.

The tophat coil is one of the best IB coils ever.
Aziz

I'm aware of all this, but please do the simulation when you are up to it and see what unbalanced residual percentage of the excitation is present in the Rx coil. In case it is not too big there may be several methods of compensating the residual unbalance. For some odd reason I believe it could do just fine.

[Aziz]

I'm aware of all this, but please do the simulation when you are up to it and see what unbalanced residual percentage of the excitation is present in the Rx coil. In case it is not too big there may be several methods of compensating the residual unbalance. For some odd reason I believe it could do just fine.

Ok, an IB case just calculated:
TX: 1 unit diameter, turns count any
RX+: 0.8*TX unit diameter, 50 turns
RX-: RX+/sqrt(2) unit diameter, 125 turns for IB (optimal for AI = 100 turns)

We have to optimize for AI condition:
N(RX+)*A(RX+) = 2* N(RX-)*A(RX-),
where N() = turns count of the coils,
A() = flux surface area for single turn of the coils

If we slightly make the inner RX- coil bigger, we could achieve the IB balance + AI feature.
But this coil is difficult to build. Nevertheless, a resonable EMI rejection is there.
It doesn't achieve the same depth performance as the tophat coil however.
Aziz

[Davor]

Thanks!

I am actually very much interested in Rx+ : Tx = 1 and with coupling near 1. With this setup I really wish to know what is the residual unbalance, so that if I make the smaller coil big enough and squeezable I can easily achieve a perfect balance.

My reasoning is that this coil, being coplanar, has all the wires in practically same proximity to the ground, and may achieve very good ground suppression.

[Ferric Toes]

Just a picture of my latest 25in noise cancelling coil arrangement. Bottom coil TX.RX1, top coil RX2. Quiet as a mouse, but if you are in an area with little emi, the top coil can be removed.

Eric.

[Aziz]

Just a picture of my latest 25in noise cancelling coil arrangement. Bottom coil TX.RX1, top coil RX2. Quiet as a mouse, but if you are in an area with little emi, the top coil can be removed.

Eric.

[ATTACH]22319[/ATTACH]

Nice coils Eric. Thanks for sharing your coil picture.
Aziz

[Aziz]

Thanks!

I am actually very much interested in Rx+ : Tx = 1 and with coupling near 1. With this setup I really wish to know what is the residual unbalance, so that if I make the smaller coil big enough and squeezable I can easily achieve a perfect balance.

My reasoning is that this coil, being coplanar, has all the wires in practically same proximity to the ground, and may achieve very good ground suppression.

Ok, with RX+ being TX size, there is an easy (but rough) IB balancing feature: the slight co-axial (vertical) distance variation (coil coupling variation) of RX+ and TX coil. But the RX- coil requires to be large enough. A significant amount of target response will be cancelled therefore.
I'll see, what my software says. Stay tuned..
Aziz

PS: The Result:
Nope! Forget it. You can't make the RX+ coil as big as the TX coil. The magnetic flux is very strong near the TX coil. We have to make the RX+ coil smaller (at least 0.8 times TX size) or have to rise it much higher (co-axial height).

[Aziz]

Ok guys, I have tried different configurations.

Now the ultimate result (Davor's + Sergey's proposal):
Not possible to meet the AI and IB condition at the same time. You're always screwing one of the criterias.
Aziz

[Davor]

How much off-balance is it with perfect AI?

[Aziz]

How much off-balance is it with perfect AI?

Hi Davor,

in the example above with with RX+ N=50, RX- N=100 (125 optimal for IB):
TX: 300 µH, 1A coil current, f=10 kHz
Induced voltages:
RX+: -13.399 V
RX-: +10.707 V
Sum: -2.692 V
Too much for an IB coil IMHO.
Aziz

[ppodg]

Elliptical Tx coil?

Hi Aziz & Davor,

And how about elliptical Tx coil, where a>r(Rx1)>b ?
This way you can obtain ANY degree of balance between the Tx and external Rx coils.
Tx and Rx1 can be interchanged; Tx (Rx) can be square etc. as well.

Regards,

[ppodg]

Compensating coil?

Another approach: let's add a compensatig/decompensating tx or rx coil...

[Davor]

A decompensating coil is in effect the same thing as ruining the ground response balance by other means. Unfortunately

Hi Davor,

in the example above with with RX+ N=50, RX- N=100 (125 optimal for IB):
TX: 300 µH, 1A coil current, f=10 kHz
Induced voltages:
RX+: -13.399 V
RX-: +10.707 V
Sum: -2.692 V
Too much for an IB coil IMHO.
Aziz

Excellent! Thank you.
I simply needed a measure to it. So roughly (+ here - there) I have some 10% mismatch or -20dB in balance. Or if balanced I'd have -20dB in interference attenuation. That's something to start with.

Since you must have the model already at hand, what is the mismatch of Tx : Rx+ = 1:0.8 ?

...
My final goal will be a real differential coil in back to back DD configuration and a 4 quadrant Rx, but in meantime a coil with ~40 dB interference and ground suppression seem a very nice idea.

[Aziz]

..
Since you must have the model already at hand, what is the mismatch of Tx : Rx+ = 1:0.8 ?
..

Help me please. I have absolutely no idea, what you mean by that.
Aziz

[Sergey_P]

You do the calculations of induced emf in a plane (2D)
I did the calculation of sensitivity to the ground in volume field (3D, approximation of - cone)

[Aziz]

You do the calculations of induced emf in a plane (2D)
I did the calculation of sensitivity to the ground in volume field (3D, approximation of - cone)

3D math. Coils can be arranged arbitrary.
Aziz