DD and Concentric

[Aziz]

Ok, www.tripod.com causes problems sometimes. I will try attachements.

[Aziz]

Here is a high resolution of the Induction Balanced PCB coil of shown one.

http://ibcoils.tripod.com/wien2.pdf

2400 x 2400 dpi resolution, dimensions are displayed in mm on top.

Just click and look/save the file. If you want to build this PCB, you should print the mirrored in horizontal direction for the negative disposure foil.
Aziz

Hi Tinkerer,

you're welcome.

My 20 cm PCB-IB coil for instance is a double sided PCB-coil and I found the corresponding project files (they were lost in my hard disk). But double sided PCB-coil is very difficult to build (because of the balance problems). Double sided PCB-coils has more mechanical and temperature stability.

The interleaved spacing of the cu-traces is a good idea to lower the capacitance. I have not tried this yet. But on the next design, I will consider this (soon).

I made also one PCB-IB-PI coil (simulation), that only has one winding for the tx-coil and one for the canceling tx-coil. The balance is made by geometry (radius) and coil resistances (different current flow through tx-coils) by connecting an amount of thin parallel wires. This lowers the eddy currents on the (thick) cu-traces. I have not practical experiences with such coil yet. But I can make some PCB-IB-PI designs for experimenting (but have to code a little bit on my software).

If you have the possibility of computerized processing, then you will not want do PI technology. You will try the better VLF technique using multiple frequencies. So therefore my experiences are restricted on IB-VLF (1 - 48 kHz).

Many years ago, I tried the MD with a 40 MIPS DSP. But DSP coding is very difficult, time and money consuming and has many other restrictions. So I gave this up and started to code on simple personal computers (Laptop, PC) using simple Windows XP operating system. It is easy, fast and low-cost. The next generation of small Laptops or tablet-pcs are most convenient for this purpose and they have enough processing power (more than you really need). Also the Hi-Fi-sound capabilities are predeterminded to lower the noises and to increase the SNR (signal-noise-ratio). This enables up to -100 .. -110 dB rx-signal for processing (24 Bit / 96 or 192 kHz sampling rate).
The main advantage of this is, you have only simple and small detection hardware and the front end is the software on the PC. Easy update and change of processing is possible.

I just coded the software on Visual C/C++ 6 and most parts of it are available through the internet (Digital signal processing, Filters, FFT, Sound-Card-Driver sources, GUI sources, etc.).

In few weeks, I will post some PCB coils as postscript and PDF files (coil software is working now). Every of you should then able to start with the Laptop MD. I made the experiments for you and proved the realisation. It is really possible.
I will also show you some pictures about the hardware and software. This could take some weeks (next year probably).

Aziz

[Tinkerer]

DSP and IB coils

Hi Aziz,
again many thanks for the info.
Your Postscript/pdf output will make it much easier to print the traces.
Double sided PCB should also help getting more turns on the same space. Thick board and designing the traces so that the trace of the one side falls in the space between the turns on the other side should reduce the capacitance.
The fact that the balance changes with humidity , ground etc. is my biggest worry, so your info about the feeding of different current to the two coils should be a big help.
I understand that you are using the IB coil for a VLF system. I am using it for a PI system, however, the balance problems are the same.
Do you write DSP source code for PIC? I would like to use some DSP filtering, but am too old to start learning that.
Tinkerer

Hi Tinkerer,

sorry for my late answers. I started with the improvement of my coil software. I also took my Laptop-MD-Hardware from my basement and started to work on.

On the new coil software, I will support Postscript (+PDF) output for the PCB-coils. I also want to develop a much bigger PCB-Coil (30 .. 40 cm diameter). The coil software and especially my printer was former restricted to DIN A4 (210 x 290 mm) format. I want to manufacture some bigger PCB-IB-Coils next year.

Thanks Aziz for the very useful information.
I have used graphite composition coatings with a PP spacer of 5 mm. That seemed to work OK. I will try with the TX inside and RX outside.
Somehow the capacitance does not seem as critical with the IB coil as it is with the traditional PI coil.
The higher resistance of the winding will only become a problem for high power TX. So far I have not tried above 1.2 A TX current. Have you tried much more power than that?
It seems to me that the balance changes somewhat when increasing the power. Did you observe the same?
Is there a way to compensate for changes in balance?
thanks for the help
Tinkerer

I took also a high resistance graphite shield.

You should keep the TX-coil resistance as low as possible to get a good coil Q. I soldered my TX-coils on the PCB to lower the resistance. I tried with 10 - 20 W Power Amplifier to feed my TX coils and got some improvement. But on the other hand, you will have more instability, signal deformations (distortion factor), more power consumption, more noise, more unwantend resonance of coil signals etc. Of course, balance of IB-coil will change.

If the resistance of the coil can not be minimized, you could use higher voltage for signal feeds (12 .. 24 V).

I do this by feeding two signals: High-Current TX-coil feed, Low-Current balance correction feed. Balance will change on different soils and ground conditions. So, the second feed is also used as a ground canceling signal.

But you can not do this technique with simple electronics. You need a DSP or high performance number crunching computer (e.g. Laptop).

Aziz

[Tinkerer]

PCB coils

Thanks Aziz for the very useful information.
I have used graphite composition coatings with a PP spacer of 5 mm. That seemed to work OK. I will try with the TX inside and RX outside.
Somehow the capacitance does not seem as critical with the IB coil as it is with the traditional PI coil.
The higher resistance of the winding will only become a problem for high power TX. So far I have not tried above 1.2 A TX current. Have you tried much more power than that?
It seems to me that the balance changes somewhat when increasing the power. Did you observe the same?
Is there a way to compensate for changes in balance?
thanks for the help
Tinkerer

Hi Tinkerer,

Hi Aziz,
again I have no luck with the tripod website. It does not open for me.

I am sorry for that case. May be it depends on your provider or on tripod. Try the click right and save the file into your hard disk.

Here are some questions:
With a standard PI coil, the best depth/sensitivity is roughly equal to the diameter of the coil.
How does it relate to the IB coil? To the diameter of the TX coil or the RX coil?

It depends on the size of the target and the pyhsical effects of the target (intensity of the distortion of the balance). Of course, it depends on many other factors: Size of the TX coils, integrated surface area of the RX coil. The advantage of the PI coils are, that they do not need a canceling coil. Therefore the magnetic flux is higher. On IB-coils, you create a local canceling area. As you can see the magnetic field direction vectors on my pictures, you can see a change of direction (from centerpoint to first TX-coil (up/down-change)). The search coil (RX) is placed somehow to add all of the integrated area, to induce 0 Volt. That means, all added magnetic fields inside would cause zero vector. The RX-coil can only detect the disturbances inside of the rx-coil-windings. As a physical effect of magnetic fields, disturbances outside the rx-coil, will also cause disturbances inside of the rx-coil.

I wound some coils in the spiral way. I found that the capacitance can be reduced that way. However, more turns and more wire are needed to obtain the same inductance. This again increases the capacitance. I feel that somewhere there is an ideal compromise, but how do I calculate that? The number of turns are not indicative, since the small diameter turns near the center of the coil, add very little inductance, while the last turns with the largest diameter add much more inductance and capacitance.

Spiral coils are relative inefficient. The magnetic flux will be cancelled slightly from the first radius to the last radius of the coil windings. So the total result is lower than on ideal circle-coils. But how the make an easy and cheap coil on the pcb? I am restricted to spriral coils. But they have a big advantage: small coils, big coils, easy to build, balanced. Detection for different sized objects. Of course, the inductance will be low. But the resistance is high.
Spiral coils are compromise.

It is also possible, to place the rx-coil outside of the tx-coils. But difficult to realise (post processing of the signals).

Another problem is the larger surface of the shield for the spiral wound coil. It adds more capacitance too.
Tinkerer

Coils, especially thin on pcb, need an enough gap to the shield, to reduce the capacity. Otherwise, it will act as a resonant circuit. A higher resistance shield (5-10k or even higher) can be a good solution (with graphite).

[plboy]

Next thing you will post here is how i didnt understand what you saying.
It is pretty hard to understand somebody so enthusiastic like yourself.
You want to discover "hot water" again and become famous. I can understand you wish to prove yourself here and be respectfull expert. I do respect you young man. Dont guess me wrong on this.
There is no proper way to establish "all in one" calculation method for coils. Due so many complex reasons. Learn more and you will see.

concentric for PI

There is an article on this site that discusses constructing concentric coils, where the number of turns on the TX and nulling coil are assumed to be related to the area contained by each coil. In practice I found that this was not correct, and my own tests indicate that they are in fact related to the radii of the two coils.
The calculator uses a modified version of Brook's equation, since the original equation was estimating the inductance about 10% lower than the actual measured value.
Hopefully I will be ready to post the software here sometime this weekend.

OK...humor&other things aside.
What you saying here is confusing!? Radii of coils are defining area contained by each coil?
So? You are practically saying here "no 2+2 are not 4, but 2+2 are 4"????
You should read more in those patents you have. Kerr for example...

[Tinkerer]

I do not know, whether Maxwall SV is able to calculate balanced IB-coils. I can not image that. It is application specific.

But to understand how IB-coils are working, it is usefull to study the EM-field of coils. I have simulated hundreds of different coils and the result of many years is the coil what I showed you.

Here is another 3d EMF of a coil:

(real 3D magnetic fields of other coil)

(real 3D magnetic fields of other coil)

Hi Aziz,

again I have no luck with the tripod website. It does not open for me. I agree with you that a good understanding of the EM field of the IB coil is necessary to be able to design such a coil. This is why I like your pictures so much.
I am new to IB design and am amazed how different the IB coil behaves from a usual PI coil. So much to learn.
Here are some questions:
With a standard PI coil, the best depth/sensitivity is roughly equal to the diameter of the coil.
How does it relate to the IB coil? To the diameter of the TX coil or the RX coil?
I wound some coils in the spiral way. I found that the capacitance can be reduced that way. However, more turns and more wire are needed to obtain the same inductance. This again increases the capacitance. I feel that somewhere there is an ideal compromise, but how do I calculate that? The number of turns are not indicative, since the small diameter turns near the center of the coil, add very little inductance, while the last turns with the largest diameter add much more inductance and capacitance.
Another problem is the larger surface of the shield for the spiral wound coil. It adds more capacitance too.
Tinkerer

[Qiaozhi]

How do you calculate your induction balanced concentric coils?

There is an article on this site that discusses constructing concentric coils, where the number of turns on the TX and nulling coil are assumed to be related to the area contained by each coil. In practice I found that this was not correct, and my own tests indicate that they are in fact related to the radii of the two coils.
The calculator uses a modified version of Brook's equation, since the original equation was estimating the inductance about 10% lower than the actual measured value.
Hopefully I will be ready to post the software here sometime this weekend.

I do not know, whether Maxwall SV is able to calculate balanced IB-coils. I can not image that. It is application specific.

But to understand how IB-coils are working, it is usefull to study the EM-field of coils. I have simulated hundreds of different coils and the result of many years is the coil what I showed you.

Here is another 3d EMF of a coil:

(real 3D magnetic fields of other coil)

(real 3D magnetic fields of other coil)

[Tinkerer]

IB coils

Aziz,
I could not access the page http://ibcoils.tripod.com/wien2.rar but I admired your pictures of the coil magnetic fields that you posted earlier.

Here is the link for download of the free version of MAXWELL SV software by Ansoft.
I understand this software is great for calculating magnetics and for visualizing the magnetic fields.

http://www.ansoft.com/membership/login.cfm?campaign=26

I feel it is very important to have the magnetic fields of the 3 coils interacting in just the right way to obtain maximum sensitivity and depth.
Tinkerer

I will post the concentric coil calculator soon. Just give me a few days to tidy-up the code.

How do you calculate your induction balanced concentric coils?

[Qiaozhi]

what about concentric

I will post the concentric coil calculator soon. Just give me a few days to tidy-up the code.