Advanced PI coils

[Ferric Toes]

Wire frame pic:
[ATTACH]22691[/ATTACH]

Spiral pcb coil 1970's style. How the artwork person did this at the time is a mystery. No computer programs then; it was all down to adhesive black tape. This is probably the first PI pcb coil attempted anywhere.

Eric.

[Aziz]

1 Million Dollar Question Quiz

Hi all,

you know, I like quiz games.
Here is a 1 Million $ question quiz for you.

What is better regards to mono PI coil performance and SNR:
1. a more compact coil bundle with less copper and less turns count or
2. a wide & thick coil bundle with more copper and more turns count?

Both coils (1. + 2.) have the same inductivity.
2. coil has more resistance and obviously less interwire capacitance. But filled with a lot of "air". Coil looking like a donut coil.

This is a consequent question towards the UCT due to the last finding. What do you think?

Aziz

PS: Damn it, but I don't have 1 Million $ to pay the winner.

[Ferric Toes]

Hi Bob,

But you have a much more current drive.

Current is inductance limited for full coil, so for 1/4 inductance it will be resistance limited (depending on pulse width and total coil circuit resistance). However, if the current doubles in half of the inductance you are back to square one i.e. same as a normal mono.

Because the centre tap at 1/2L is not going to be a 1/2Turns, will this not give an unbalance to the common mode response. I expect a balance point for the CT could be calculated.

Ground noise is a big issue on Oz ground. And this gets more critical with half coil driving.

No reason for this that I can see.

Eric.

[Aziz]

Wasn't Tesla the one that lodged the patent on a spiral coil?

Yes, Nikola Tesla was a great person & inventor. He is one of my favourite scientists.

I think its going to be very difficult to convince a patent examiner to give his Ok for a new spiral coil patent.
*LOL*
(Oh yes, I forgot: it's only a question of "Fakelaki"/"Bakshish" (money) today.)
Aziz

[Aziz]

Hey Aziz

How about a centre tapped spiral coil into a full differential amp

would that help !!!

Bob..

Hi Bob,

it is really nice, that spiral coils can be made center-tapped using the bifilar coil wire. The only advantage is, that you can drive it at L/4 (=coil half) to push more current and damp and receive at both coil halves. There is no benefit regards to EMI noise induction. When you sample at both coil halves, you have the twice target response and the twice EMI noise induction as well. But you have a much more current drive.

Ground noise is a big issue on Oz ground. And this gets more critical with half coil driving.

Cheers,
Aziz

[Aziz]

Hi all,

BTW, I have to mention, that the spiral coil response calculations were done at the z-axis and the calculated figures in the first few inches does not contain the maximum target response. Only at the center axis. This is the reason, why you get strange relation figure at 2 inch detection depth. The mono coil response does contain the maximum response on its edge and the spiral coil doesn't yet (would require very time consuming coil calculations). Just ignore the strange relation in the first two response depth.

Some notes on the spiral coil:
It is obvious, that it isn't optimal for highly mineralized ground due to stronger magnetic field strengh on the center region. You will get more ground noise of course (don't confuse with EMI noise). A good compromise would be a round TX (as usual or semi spiral) and a separate spiral RX, which is smaller than TX. So not all magnetic flux of the TX is linked to the RX spiral coil. The RX spiral beginning could even be smaller than 0.5 diameter units of the TX coil.

Let's continue with the WBCT ... to find the ultimate coil technology (UCT).
Anyone want to oppose now?

Cheers,
Aziz

[techo_bob]

Hey Aziz

How about a centre tapped spiral coil into a full differential amp

would that help !!!

Bob..

[Davor]

Wasn't Tesla the one that lodged the patent on a spiral coil?

[dfbowers]

Nice work Aziz. I enjoy reading your daily posts very much.

Back in the day of BFO detectors, I built a small spiral coil that outperformed just about anything else I could come up with. They work great on Telsa coils too!

Don.

[Aziz]

The Spiral PI Coil Pics

Wire frame pic:


Magnetic field strength pics:




Cheers,
Aziz

[Aziz]

Surprise! Surprise!

Ok guys,

another magic is happening here. Fortunately, this spiral coil is prior art and nobody can lock it with evil patents anymore. Pity, but I didn't invent it. And it works pretty well and is even outperforming the mono coil. As I said before, never give it up! How is that now?


Spiral coil:
Beginning from 0.5 unit diameter (=5 inch) and ending at 1.0 unit diameter (=10 inch). 35 loop turns (but I'm using internally in my coil software N=34.7).

Unfortunately, this flat spiral PI coil is picking up approx. 9.7% more EMI noise. So the total coil flux area is 9.7% more compared to the mono coil. On the other hand, it is performing exceptionally well. In the relation comparison below, the 9.7% more noise is divided out (by multiplying the relation by 0.912). So we see the SNR corrected relation and is a good comparison to the mono coil regards to SNR criteria.

Coil Comparison (Calculated Response Ratio to Reference Coil)
© 2012 by Aziz Ögüt. All Rights Reserved. 10-Jan-2013
Reference Coil: 300 µH Round Mono Loop Coil
		10" Spiral PI Coil (0.5 DU .. 1.0 DU)
Target Depth [inch]	Target Depth [cm]	Relation to Reference SNR compensated (9.7 % more EMI noise)
1	2,5	3,007
2	5,1	3,949
3	7,6	3,085
4	10,2	2,433
5	12,7	2,034
6	15,2	1,783
7	17,8	1,618
8	20,3	1,504
9	22,9	1,424
10	25,4	1,365
11	27,9	1,320
12	30,5	1,286
13	33,0	1,259
14	35,6	1,237
15	38,1	1,219
16	40,6	1,205
17	43,2	1,193
18	45,7	1,183
19	48,3	1,174
20	50,8	1,167
21	53,3	1,161
22	55,9	1,155
23	58,4	1,150
24	61,0	1,146
25	63,5	1,142
26	66,0	1,139
27	68,6	1,136
28	71,1	1,133
29	73,7	1,131
30	76,2	1,129

Now look at this magic stuff. I'm really impressed. There is a coil now, which is outperforming the mono coil with ease. One have to tame the shielding capacitance somehow.


Cheers,
Aziz

[Orbit]

excellent Demonn!!!

[Aziz]

...
Practically, this type of coil performed better than standard, dual field or any other configuration so far, tested with standard detectors, or some weired configurations custom built to utilize it.
..

You're dead right according to my first coil simulation! But I have to check the SNR yet.
(Hey, the result was an incredible surprise to me, that I've began to check my coil software and coil model .. But there isn't any mistake.)

Aziz

[deemon]

Deemon, how should be controlled the balance circuit - automatic or manually?
How the balance circuit with MONOCOIL will suppress the GND signal?

If we use the coil in a classic VLF detector ( with X-R discrimination ) - we can use manual balance , why not ? It can be stable enough , especially when we need to balance only one parameter ( inductance ) - for example , like I did in my "Faraday bridge" circuit , described here - http://www.geotech1.com/forums/showt...3-mono-coil-IB . Or if we don't want to tune it manually - it can be done electronically , during the turn-on procedure , it's not a difficult task .

But here is even more interesting idea - full automatic balance during operation time . Of course , it needs more complicated electronics , but I found very interesting solution , and now finishing the circuit based on it . Of course , there are some problems ... but I hope to solve them all successfully .

P.S. And what about GND signal suppression - I believe that this task must be solved only by electronics , utilizing the different behaviour of magnetization ( and demagnetization ) of magnetic minerals and metal targets . Why I think so - because all methods of ground balance using coil construction depends on such parameters as a distance between the coil and the ground , or maybe tilt of the coil ( with some 8-type coils ) , ground gradient , and other disturbing factors .... so I don't suppose these methods to be reliable enough . And it's also true when we speaking about RF interference cancellation ...

[Tepco]

This flat spiral stuff is interesting. I tried many different flat spiral configurations before, similar to what you going to simulate, very interesting coil configuration ( i even posted some design here time ago). I never made any computer sim. on it but just based on experience, something similar to your model, roughly 300uH, 30 turns, 2/1 OD/ID ratio works perfectly good, made of two wire speaker cable. Main reason for this physical configuration is increased inter winding distance between turns, filled with insulation, reduced total self capacitance and faster sampling. However, more interesting ways to use similar coils are obvious.


Similar to bifilar center tapped , 2wire flat spiral is also combination of two, physically and electrically identical, tightly coupled coils, but electrically insulated. Very interesting designs can be developed (i tried ) using this configuration. Say, center tapped bipolar pulsing (some issues here with shielded coil), or multi period pulsing, using constant current drive and two different voltage power supplies, or floating TX from RX part, or differential for noise reduction, or... whatever.


Practically, this type of coil performed better than standard, dual field or any other configuration so far, tested with standard detectors, or some weired configurations custom built to utilize it. Biffilar coil can be similar, enabling lot of testing and experimentation to be done. Interesting.