Large PI Coils ????

[ElectroNovice]

Hi Mr Woody au
Your coil configuration will give great sensitivity but poor depth because the depth is controlled by the Rx separation which needs to be at least 2 x Rx coil diameter.
See US patent 6,853,194 by Nelson et al.

The other problem is that turns spacing profoundly affects proximity losses (Q). The winding method you are using does not produce the lowest inter -winding capacitance or the highest Q or the highest coil resonant frequency.

Also the inductive relationship between Tx and Rx (the ratio of their diameters) is critical in obtaining optimal performance. Also to get the optimal depth capability the coil needs to be in resonant tank circuit where the Tx is seen as a purely resistive load, the inductive and capacitive reactance’s canceling out. Then you can get coil magnetic moments of 9-10 fold higher than with our Pis’ we have at present. You also need to take into account the relationship between skin depth and transmit frequency which profoundly effects depth and also the frequency dependent component of magnetic susceptibility in high ferromagnetic terrains.
ElectroNovice

[Woody.au]

It is good to get these sorts of responses as it stops one covering old ground and wasting ones time. Fine on the winding but my own testing gave a Q of 5.6 and gave the best air testing distance when used as a Mono. I tried a square form factor coil and I lost about 6" of depth on a coke can sized object.

I had a look at the patent and thought that maybe it would be worth using the middle coil for TX and the inner and outer for RX? Also has anybody tried the flat spider wound coils in a P.I ? I did many years ago in a coplanar VLF coil but never compared it to a stock coil.

There is some great information on this site and the amount of knowledge is vast and impressive.

Has anybody got any more ideas that I can experiment with? I am putting aside 1 month just to do coil experiments and build prototypes and test them in the Red Hot Mineralized Gold fields of Central Victoria Australia.

Test machine is a Minelab GP-3000

Peter.

[Unregistered]

Hi Mr Woody au
Your coil configuration will give great sensitivity but poor depth because the depth is controlled by the Rx separation which needs to be at least 2 x Rx coil diameter.
See US patent 6,853,194 by Nelson et al.

The other problem is that turns spacing profoundly affects proximity losses (Q). The winding method you are using does not produce the lowest inter -winding capacitance or the highest Q or the highest coil resonant frequency.

Also the inductive relationship between Tx and Rx (the ratio of their diameters) is critical in obtaining optimal performance. Also to get the optimal depth capability the coil needs to be in resonant tank circuit where the Tx is seen as a purely resistive load, the inductive and capacitive reactance’s canceling out. Then you can get coil magnetic moments of 9-10 fold higher than with our Pis’ we have at present. You also need to take into account the relationship between skin depth and transmit frequency which profoundly effects depth and also the frequency dependent component of magnetic susceptibility in high ferromagnetic terrains.
ElectroNovice

EN, what is the best configuration for coil design acording to your words ?

thanks

[ElectroNovice]

It is good to get these sorts of responses as it stops one covering old ground and wasting ones time. Fine on the winding but my own testing gave a Q of 5.6 and gave the best air testing distance when used as a Mono. I tried a square form factor coil and I lost about 6" of depth on a coke can sized object.

I had a look at the patent and thought that maybe it would be worth using the middle coil for TX and the inner and outer for RX? Also has anybody tried the flat spider wound coils in a P.I ? I did many years ago in a coplanar VLF coil but never compared it to a stock coil.

There is some great information on this site and the amount of knowledge is vast and impressive.

Has anybody got any more ideas that I can experiment with? I am putting aside 1 month just to do coil experiments and build prototypes and test them in the Red Hot Mineralized Gold fields of Central Victoria Australia.

Test machine is a Minelab GP-3000

Peter.

Hi Woody au,
Here are some refs that may help you develop the ultimate Pi and coil combinations for getting big objects at great depth and also give you an insight into the problems posed by bad ground. If you do nothing else see: The ultimate report on coils and Pi detectors! later in my post.

An experimental study of the time-domain electromagnetic response of a buried conductive plate
Geophysics, Volume 70, Issue 1, pp. G1-G7 (January-February 2005)
Mark E. Everett, Alfonso Benavides, and Carl J. Pierce
Texas A&M University, Department of Geology and Geophysics, College Station, Texas

The influence of magnetic soil on electromagnetic sensors
Stephen D. Billings Leonard R. Pasion
UBC-Geophysical Inversion Facility University of British Columbia

A Fast 4-D TEM System for UXO Characterization
http://www.estcp.org/documents/techdocs/UX_0105.pdf


UXO DISCRIMINATION USING TIME DOMAIN ELECTROMAGNETIC INDUCTION
Leonard R. Pasion, Stephen D. Billings, and Douglas W. Oldenburg
UBC - Geophysical Inversion Facility
Department of Earth and Ocean Sciences, University of British Columbia
http://www.eos.ubc.ca/research/ubcgi..._detection.pdf

Evaluating the effects of magnetic soils on TEM measurements for UXO detection
Leonard R. Pasion_, Stephen D. Billings, and Douglas W. Oldenburg
UBC - Geophysical Inversion Facility, Department of Earth and Ocean Sciences, University of British Columbia

GEOPHYSICS, VOL. 69, NO. 6 (NOVEMBER-DECEMBER 2004); P. 1387–1393, 12
Electromagnetic detection of buried metallic objects using quad–quad conductivity
Haoping Huang and I. J. Won


UXO DETECTION IMPROVEMENTS UXO DETECTION IMPROVEMENTS USING EM63 SYNTHETIC MULTI USING EM63 SYNTHETIC MULTI- RECEIVER ARRAY GEOMETRIES RECEIVER ARRAY GEOMETRIES
Carl J. Pierce, Jr., Alfonso Benavides, Mark Everett and Jack Stalnaker
https://www.seg.org/publications/arc...2003/NSG33.pdf

EMPLOYING MULTIPLE GEOPHYSICAL SENSOR SYSTEMS TO ENHANCE BURIED UXO “TARGET RECOGNITION” CAPABILITY
Dwain K. Butler*
Alion Science and Technology Corp.Vicksburg, Mississippi 39180-6199
Donald E. Yule and Hollis H. Bennett, Jr.
U.S. Army Engineer Research and Development Center
http://el.erdc.usace.army.mil/uxo/pdfs/uxo04-butler.pdf

Fractal properties of background noise and target signal enhancement using CSEM data
Alfonso Benavides, Mark E. Everett, Carl Pierce, and Cam Nguyen
http://spiedl.aip.org/getabs/servlet...cvips&gifs=yes



The ultimate report on coils and Pi detectors!
“The most significant finding are detailed in Section 2 of this report where we use the principal dipole moments model to evaluate the improvements in system performance that can be obtained with use of multiple receivers and/or transmitters. The other most significant result here is that the depth of detection can be doubled compared to a conventional coaxial transmitter – receiver system if one uses a 3 component transmitter as well as a 3 component receiver.”

DETECTION AND CLASSIFICATION OF BURIED METALLIC OBJECTS
UX-1225
Frank Morrison, Torquil Smith, Alex Becker,and Erika Gasperikova
Lawrence Berkeley National Laboratory
http://www-library.lbl.gov/docs/LBNL...3962_Final.pdf

Also see the Geophex site where they are also now using multiple receive coils for the FDEM Gem3
http://www.geophex.com/Publications/publications.htm
Gem 3d with orthogonal coils
A new UWB Handheld Vector EMI Sensor System that Exploits Precise 3-D Positioning
“Additional, orthogonal sets of coils have been added to the Geophex Ltd
GEM-3 device. Being parallel to the transmitted field, these receive only
the signal produced by the target.”
http://el.erdc.usace.army.mil/uxo/pd...m04-oneill.pdf

ElectroNovice

[ElectroNovice]

EN, what is the best configuration for coil design acording to your words ?

thanks

The best configuration and winding method I cannot reveal as they are commercially sensitive. The biggest improvements in depth on large objects comes about by using different pulse forms eg triangle pulses and filtering so that only the fundamental is transmitted to the Tx at its resonant frequency or by using 1/2 sine wave pulses again at the TX resonant frequency. A combination of all these factors and and optimal coils can give depth improvements of reportedly up to 40% or in one of the refs in my previous post 50% The biggest hurdle is still highly ferromagnetic ground. Eliminating or dramatically reducing its effects will result in a big improvement in s/n and a lot more depth and sensitivity. The research in the mine detection field and geophysical prospecting will i believe eventually filter down to recreational Pi detectors. Pi has not yet reached its zenith!
ElectroNovice

[Woody.au]

ElectroNovice , you should change your name to ElectroKnowalot...lol


Ok here is some experimental data on coil winding tests.

A hard wound square form factor coil gives me greatest TX current density but slowest reset times, highest interwinding capacitance and wait for it " less receive signal amplitude from a lump of Yellow". A spider would coil in which
the windings are wound in a conical shape give greatest received signal level but the lowest output. (far field) when used as a TX winding.

Next experiment is to wind a conical TX basket weave and use a separate Conincal TX coil and see what if any gains can be had.

If anybody has done any of this before please speak up before I waste another stack of hours when I could be trying something else.

[treasurediver]

Your coil winding information is very interesting. Could you also give info on the pps,Tx, delay and Rx times of that detector? These are obviously very important factors linked to the coil speed.
Thanks
Treasurediver

[Largesarge]

Woody:


You might get up with Gary (chemelec.com) he's had some experience with conicals, he even has an inductance calculator for them.

[Tinkerer]

different coil windings

ElectroNovice , you should change your name to ElectroKnowalot...lol


Ok here is some experimental data on coil winding tests.

A hard wound square form factor coil gives me greatest TX current density but slowest reset times, highest interwinding capacitance and wait for it " less receive signal amplitude from a lump of Yellow". A spider would coil in which
the windings are wound in a conical shape give greatest received signal level but the lowest output. (far field) when used as a TX winding.

Next experiment is to wind a conical TX basket weave and use a separate Conincal TX coil and see what if any gains can be had.

If anybody has done any of this before please speak up before I waste another stack of hours when I could be trying something else.

Your experiments are very interesting. I have been working along the same lines, basket wound Tx and spider wound Rx. At present I am waiting for Litz wire to reduce the delay caused by the coils picking up themselves.
Elongated coils seem to give a response according to the narrow with for the depth but cover more area per sweep. I dont understand it yet totally. Because of the reduced area I would have thought that the field strenght should be intensified. (I have not done field tests yet)
Next I want to try a small rx coil, at some distance of the other 2 coils, to pick up the ground signal separately and then subtract it from the total signal.
What are the tx, rx and delay times of your detector?
Tinkerer

[bbsailor]

Please post the following hard data about your coils.

1. Coil Diameter
2. Wire size
3. Number of turns
4. Inductance
5 Resistance
6. Coil style: describe the coil style the best you can (how it is wound)
7. Self-resonant frequency
8. Shielding added, self-resonant frequency
9. Coax cable type and capacitance
10 Self-resonant frequency measured through coax cable connected to coil

With this data, or as much as you have, we can all compare what you are doing to our own coil creations.

Thanks

bbsailor

[Woody.au]

I will post a lot of data once I finish off some designs and field test them.
Currently doing 3 designs and one is looking extremely good.

[Tinkerer]

Kynar x Teflon

The short answer.

AWG 30 Kynar, 18 turns, 348uH, 672KHz, 161 pF
AWG 30 Teflon, 18 turns, 295uH, 1.1Mhz, 71pF

Measured with a 10:1 scope probe and a 1 megohm resistor between signal generator and coil. Add 30 KHz for unloaded measurement values.

bbsailor

Thanks and thanks again for the wealth of informationyou bring to the forum.
Question: are the AWG wires above equal: how many strands ? or solid? What is the outer diameter of each wire? If it is not the same, could we deduce a formula: Insulation X copper density od the coil bundle.

It would also be very interesting if somebody would share info about different filler - resin mixes or polyurethane foam used for casting, fixing the coils in the forms and the amount of capacitance each method resulted in.
Tinkerer

[bbsailor]

Tinkerer and All,

Answers to Your Questions.

The Kynar AWG 30 is single stranded with a .019 inch outside diameter.
The Teflon AWG is single stranded with a.024 inch outside diameter.

Wire that has thinner insulation will have the windings closer together and thus will have a higher inductance of a similar coil with thicker insulation.

Now, to add one more level of complexity, the dielectric constant of the insulation plays an important part in how two similar coils, made of different insulation will exhibit different self-resonances. PVC has a higher dielectric constant of 3.8 to 8 so it will have a higher turn to turn capacitance and will result in a lower self-resonance. Teflon has a dielectric of about 2 and offers the best dielectric constant for common insulations. Dry air is the best dielectric having a value of 1, but may not be practical for making coils that need to remain stable under use or be compact and light enough to fit into a coil housing.

This dielectric effect on coils can be seen by placing a single spiral wrap layer on you coil, then measure it's self resonance. Then add another layer of spiralwrap and measure it again. If you use PVC spiralwrap the effect will be greater than if you use Polyethylene spiralwrap but is still measurable. The self resonance will be slightly lower with the second layer indicating that the spiral wrap does also influence coil performance. Teflon spiral wrap is very expensive but Polyethylene has a dielectric constant of 2.2 to 2.4 and is a better, less expensive choice than PVC spiralwrap and close to the performance of Teflon.

I just wound a 4.5" diameter coil to use on my workbench.

Diameter: 4.5"
Wire: AWG30 Teflon insulated single strand
Turns: 32
Inductance of the coil before adding shield: 303uH
Self resonance of coil before adding shield: 1.44 Mhz
Coil to shield capacitance: 73pf (over one layer of spiralwrap)
3 ft. coax capaitance 76pf
Self resonance measured through coax cable: 800 Khz
Final inductance of the coil measured at the end of coax cable: 340uH

My conclusion from my coil winding research suggests that Teflon wire makes higher Q coils, meaning lower capacitance coils, than PVC insulation made coils. A good target for coil to shield capacitance for about an 11" coil should be near 40 pf per foot of coil circumference. I have wound many coils and this seems to be about the best range of results I have consistantly achieved so far.

Use a foam with a lot of air so it will have a low dielectric constant. Also, try to keep material or chemicals that have a higher dielectric away from the coil windings. This is to keep the coil Q high. Get in the habit of looking at the dielectric constant of things that are near the coil. This one recommendation will pay off and you will see measureable results.

Questions for Tinkerer:

What is the PI PPS frequency range of you PI machine and the coil peak current?

What is the total resistance of your coil also include total resistance of coil plus any series resistor and MOSFET when it is turned on?

What is the minimum delay you are seeking to achieve?

What targets are you seeking?

Keep on winding.

bbsailor

[Tinkerer]

Coil resonance

Hi bbsailor,
thanks for the very interesting coil information. With the coil resonating at 800khz, what is the minimum delay time you get?
About my design: I want to look for small gold nuggets. Therefore I aim at a delay of around 10us and a pps of around 10khz. the initial design will have a variable frequency mostly depending on the "speed" of the coil. My old scope has died, so, until I get the new one I spend my time learning as much as I can about the intrinsics of PI.
My experiments of conductive paint for shielding have ended with a coating consisting of 90% graphite and 10 % epoxy. 2 coats give about 80 ohm resistance at 1" distance.
Many years ago when I was building detectors for divers, I made the coil hosing with polyester fiberglass and cast the coil in polyester resin. With the 40us delay time this worked OK until the diver spent several folowing days on the bottom of the sea. Then the coil capacitance increased and the 40us were not long enough. then we changed the coil to a dry one that worked fine for a few days. Investigating the problem I found that polyester that is continously immersed, absorbs water. So do many plastics, nylon for example absorbs 2% water, quells and changes its electrical qualities considerably. Just a little bit of info for the divers...
Tinkerer

[bbsailor]

Tinkerer,


I sent my coil out to have it independently measured by someone with very accurate equipment. As soon as I get the results I will post them. I suspect that it is below 10us possibly 8 uS as I am using it on a modified Cscope CS6PI. I doubled the main frequency range from 2.2Kpps - 6.6K pps to a new range of 4.6 Kpps to 13.8Kpps by reducing the main timing cap from 1000 pf to 320 pf. This also speeds up all the timing. Before the fastest delay was about 16 or 17uS, now it is down to about 8 uS maybe a little less.

Some of my early coils just stopped working as I raised the frequency, which also reduces the delay and pulse width. By the coil not responding to a target, I reached the lower limit of that coil's speed at about 12 or 13 uS. Then I started to be very careful about minimizing the capacitance as was able to operate the coil at the highest frequency and also at the lowest delay which I believe is about 7.5 to 8 uS. If I had a faster PI machine, I could see if it goes lower. That is why I sent the coil out to be measured.

I only use 32 inches of coax between the coil and the PI control box.

The size of your damping resistor is very telling about the potential of your coil's speed.

What value damping resistor are you using?

My damping resistors on the CS6PI started out at about 850 ohms and now I am working at about 1K ohm. On some of my new coils I am trying to get to about 1100 to 1200 ohm damping. That that is really pushing it. I need to push up the speed of my Hammerhead near 10K pps, modify the minimum pulse delay by reducing the resistor value in series with the pulse delay control by about 500 ohms and limit the peak coil current to about 100 ma. Once you start putting a resistor in series with the coil to limit peak coil current, (to sample faster) you can also locate MOSFETS that have the lowest output capacitance and not care about the on-state MOSFET resistance.

What I did to make changing MOSFETS faster is to cut up some 8 pin DIP IC sockets. Use only the cheap sockets with flat pins as the round pin sockets will not fit the MOSFET leads. The DIP pins line up with the MOSFET spacing and make changing different MOSFETS just a matter of unplugging and plugging in a new one.

For beach hunting 10 uS is about the pastest speed that is practical as I start picking up the wet sand and salt water at these faster speeds. Although I mainly do beach hunting, I want to see how far below 10 uS I can go just to see the design limits.

bbsailor