IB-PI coil

[Tinkerer]

We will start with the coil.

Any suggestions for improvements are welcome.

David Emery has written a good tutorial about IB coil building. See the following link:
http://www.geotech1.com/cgi-bin/page...anar/index.dat

I started with that and then adapted a few things. I have to do with the means at hand, if you have better materials yourself, the results should be better?

We have chosen the following parameters:

TX coil:
Outer diameter max 195mm, including shielding spacer and shielding itself, so that the coil fits into a standard 8" coil shell.
Inner diameter will depend a bit on the thickness of the wire insulation. For my own coil I use 175mm
17 meters of wire should give 29 turns and about 300uH.
The exact amount of inductance will depend on the fill factor, or how tight you wind the coil.
Tighter winding gives more inductance and more inter wire capacitance. The capacitance should be minimized whenever possible.

For the TINKERERS coil we want low DC resistance, less than 1 Ohm.
I have some 0.5mm diameter magnet wire, I will use 4 strands in parallel.

Tinkerer

[Tinkerer]

8" concentric IB coil trial

Tx coil, 29 turns 4x0.5mm magnet wire with PP spacer.
OD=187mm, 350uH

RX coil 2x20 turns, center tapped, 1x0.5mm magnet wire with PP spacer.
OD=110mm, 325uH

Bucking coil 15 turns 0.9mm magnet wire. I noticed immediately the high number of turns needed. as expected, this high number of turns reduced the sensitivity of the coil assembly considerably.

For best performance, the number of turns of the bucking coil should be kept to a minimum.

My former coil assembly with a TX coil of 285mm OD and a RX coil of 155mm OD was much more sensitive. It had a bucking coil of 9 turns.

The center tapped RX coil worked fine with the differential input preamplifier.

Next I will try a DD coil assembly.

Tinkerer

[hobbes_lives]

If the 8" coil configuration is not very sensitive, and you are able to get good results with a DD, then Hays has lots of housings that will work well. I've used A for an 11" monocoil, but it will work for a DD coil as well. There's also C, 14"x10", and H, 11"x8.5". Then there's K, 18"x14", if you wanna get crazy.

Also, aren't your measurements a little smaller than 8"? How did you choose the RX coil size in relation to the TX? Just curious.

[Tinkerer]

If the 8" coil configuration is not very sensitive, and you are able to get good results with a DD, then Hays has lots of housings that will work well. I've used A for an 11" monocoil, but it will work for a DD coil as well. There's also C, 14"x10", and H, 11"x8.5". Then there's K, 18"x14", if you wanna get crazy.

Also, aren't your measurements a little smaller than 8"? How did you choose the RX coil size in relation to the TX? Just curious.

Hobbes,

thanks for the feedback. There is nothing wrong with the 8" coil configuration.
What came out not so good, is my implementation of it.
Lets look at the reasons so others can learn from my mistakes. (Since I am making a lot of mistakes, others can learn a lot from me?)
I have an 8" coil housing (gift from Dave Emery). For the TX coil to fit into it, with coil to shield spacer and all, the TX coil needs to be a bit smaller.
Now, when I looked at the space left for the RX coil, it looked rather like a small RX coil, so I took advantage and experimented with an RX coil size that is larger than half the TX coil diameter.
Now, what happens with a smaller coil diameter, conserving the same TX power, the B field is more intense than for a larger coil diameter.
To balance the RX and TX coils, we add a Bucking coil BU that has the current running in the opposite direction (or 180 degrees) than the RX coil.
The problem with that is that the target response currents captured by the RX and the Bucking coil BU will also run in opposite direction and thus cancel each other if they are of equal strength.
Therefore, we must try to make the bucking coil with the minimum amount of turns.
With the TX and RX coil a far distance apart, the k factor or coupling factor between the 2 coils is low. The BU coil needs to cancel only the small fraction of coil current that was coupled to the RX coil.
The closer we move the RX and TX coils together and the more intense the B field is, the more turns the BU coil needs.
So it is a matter of designing the best proportions for a coil assembly.

Tinkerer

[Tinkerer]

DD coil experiment

DD coils are quite common nowadays, but I never used one in the field myself.
Time to do some tinkering with DD's.
I had some old coil windings of 280mm OD, lying around, so I bent them into DD shapes and started balancing them.
They don't like to stay put just like that, so I squeezed them between two plastic disks to stabilize the assembly.

Once they were more or less balanced I checked for sensitivity.

Here are the screen shots, with the usual targets:

#1 all targets
#2 No target
#3 Steel ring, as you can see the steel ring is very similar in diameter to the cold ring, but thicker.
#4 gold ring, 0.5 grams
#5 silver ring, 0.55 grams, but larger diameter.
The targets were placed at the center of the coil on top of the plastic disk.

The scope is set at 500mV/div. TX 100uS, 12V

The pictures show the output of the differential preamp.
Note the discrimination capabilities.

I am quite happy with the results and feel it is worth investing further time into the DD coil project.

Tinkerer

[Tinkerer]

DD coil building

DD coils are easy to build and are well known.
One little problem when building a DD coil, is the fixing of the 2 coils in their final balanced position without shifting the balance.
A solution to that is to add a Bucking coil. Just 2 turns of wire are sufficient.
with the TX coil and the RX coil already fixed in their permanent position. Moving the bucking coil slightly, the coil assembly's induction balance can be fine tuned, then the Bucking coil is fixed permanently too.

Now what happens if we use the coil in conducting sea water? the balance is again upset, but the coils are now cast in Epoxy and can not be moved anymore.
The same can happen if we build the coil for moderately mineralized soil and then move to highly mineralized magnetic soil. The induction balance is very much upset and the coil will work marginally.

A solution to these problems is to control the current in the Bucking coil.
With a current control circuit we can restore the balance of the DD coil in the resistive or the reactive balance.

Attached are pictures of the building of an experimental DD coil. The coil assembly is clamped together with a plastic screw and shielded with a full disk shield on top and bottom for testing purpose.

Pictures:
#1 TX and RX coils overlapping, placed on the bottom shield, balanced and held in place with tape.
#2 A small 2 turn Bucking coil is added and taped in the balanced position.
#3 the top shield is added and the assembly is clamped with a plastic screw.
Coil test results:
#4 No target. The sample pulse is shown at 3uS.
#5 a lead Minie ball at the center of the coil.
#6 a steel lug of similar size at the center of the coil.
Note how the signal goes positive for the lead and negative for the steel.

The gain is 100 and the scope is set at 2V division

Tinkerer

[Tinkerer]

DD COIL INDUCTION BALANCE ADJUSTMENT

Here are 2 pictures showing the DD coil induction balance adjustments.
Note that the adjustment is made to the actual induction balance of the coil, by regulating the Bucking coil current, or rather the proportion between the TX coil current and the Bucking coil current.
The adjustment shown is about 500mV positive or negative.
this adjustment is made manually, with a pot. However, it could very well also be made by a MCU, using a digital pot or even a D/A.

On the pictures is also seen the ground balance sample pulse, to serve as a reference of the Voltage level.

Tinkerer

[Tinkerer]

DD coil assembly

Having tested the functionality of the DD coil, it is time to assemble the coils in a permanent way.
I chose Styrofoam as a base and glued the TX coil and the RX coil embedded, so that they are fixed in a rigid matrix.
Picture 1 shows the TX coil.
The RX coil was embedded the same way and then the two coils overlapped and glued in the position where they were perfectly balanced.
Then a small single loop was added (see annexed circuit) This makes it possible to make adjustments after the coil is put in use and in the field. The balance of the coils can be adjusted for very high conductive soil, like salt produces, or, the opposite way, for high magnetic mineralization.

Picture 2 shows the single loop final adjustment added, as well as the shield cover disk. A similar shield goes on the bottom.

In this instance I have used a graphite shield with a measured resistance of about 1k Ohm per inch. I have not left any gap in the disks, but there will be small gaps on the side.
The top, bottom and side shields have drain wires embedded, which are soldered to the cable shield.
I will test this coil assembly and if needed I will sand a small gap in the shields.

Tinkerer

[nick_f]

Note that the adjustment is made to the actual induction balance of the coil, by regulating the Bucking coil current, or rather the proportion between the TX coil current and the Bucking coil current.
The adjustment shown is about 500mV positive or negative.
this adjustment is made manually, with a pot. However, it could very well also be made by a MCU, using a digital pot or even a D/A.
Tinkerer

Hi Tinkerer,
This is a great job, the electronic balancing of the coil using the bucking coil and the mosfet is simply wonderful. I was also considering adding a bucking coil to a DD Coil (after I discovered how critical is the balancing of the coils during construction phase). I didn't know how I could control the current through the bucking coil, and your solution with MOSFET is wonderful. I also like the idea of a digital pot, set it and forget it, until the balance is upset.

Best regards,
Nicolae

[Tinkerer]

Single loop control

Hi Tinkerer,
This is a great job, the electronic balancing of the coil using the bucking coil and the mosfet is simply wonderful. I was also considering adding a bucking coil to a DD Coil (after I discovered how critical is the balancing of the coils during construction phase). I didn't know how I could control the current through the bucking coil, and your solution with MOSFET is wonderful. I also like the idea of a digital pot, set it and forget it, until the balance is upset.

Best regards,
Nicolae

Hi Nicolae,

It is good you mention the single loop control. I had a look at it and saw that there are mistakes in the circuit.
I will post a correct circuit as soon as possible.
Tinkerer

[ZED]

G`day Tinkere

Now those signals i can relate to,....your second pic (DD no target),well i see a signal,can you tell me what im seeing please ?

Zed

[Tinkerer]

G`day Tinkere

Now those signals i can relate to,....your second pic (DD no target),well i see a signal,can you tell me what im seeing please ?

Zed

Hi Zed,

are you referring to post #5? If so, I will look in my notes what the setup was. I also see that the picture of the steel ring is missing. It is the one that shows the difference between steel and gold, so I will try to find it and post it again.

Tinkerer

[Tinkerer]

Hi Zed,

are you referring to post #5? If so, I will look in my notes what the setup was. I also see that the picture of the steel ring is missing. It is the one that shows the difference between steel and gold, so I will try to find it and post it again.

Tinkerer

Well, this is funny, I wanted to add the picture of the steel ring, but now it is already there. Must have gotten hung-up in the clouds.
Anyway, these were kind of ad hoc tests, just trying to see if it was worth wile to pursue the DD coil setup.
I often set the coil on top of some red clay brick to simulate magnetic ground.
The bricks give some response, that the Ground Balance has to cope with. So far the slight magnetic background has been no problem.

Now, this setup and circuit are very different from your ML detectors. My intention is not to copy and mod ML detectors, but to look for different ways of making a PI that can tell the difference between FE and gold and give some more information about the target before it is dug up.

The pictures show the whole of the TX pulse, the Flyback and some time after the Flyback.
For FE, one side goes up, the other side goes down.
For gold and silver, both sides go the same way.
Presenting the steel ring in it's flat position (as above) does not show a very big difference, but at any angle less than 180 degrees the signal shows the FE characteristic better.
As the one side of the signal goes more positive and the other side of the signal more negative for FE, the sum of the 2 samples cancel each other, reducing the overall response of FE.
There are some notorious difficult FE targets that are very difficult to differentiate, like this rusty pieces of cans or even Crown Corks.
After a great deal of experimenting, I found a way to differentiate these also.
I do it in Analog, but it can be done much easier with specific programming of a MCU. The program can be written such that one can choose the settings to differentiate or eliminate specific targets.

Tinkerer

[ZED]

I see what you mean by both sides going up with non FE targets and one side up and one side down with FE targets.

Changing the phase of the GB sample is a good method of ground balancing,i have used it often as have others.In the designs i used a 10 turn pot was required but to do it digitly i would think a resolution or steps of 100ns would be in order.Whats the resolution on your P.I.?

Zed

[Tinkerer]

Concentric IB coil for TMPI

A top notch detector needs a top notch coil for best performance.

It takes a few tries to make the perfect coil. The attached picture shows the beginning of the coil # 1
I am now at coil # 5. The coils are getting more sensitive and more precisely designed for the specs of the TMPI. (TINKERER/MOODZ PULSE INDUCTION)

These coils are made for a differential input preamp.
They also have an electronic balance adjustment. Up to now this adjustment was made with a potentiometer, however, with the FPGA, it can be made automatic.
One feature of these coils is the multi wire cable.
For coil #4 I used CAT 5 cable to connect the coils to the board. It worked surprisingly well as long as the power was not too high.

For high power, I will have to build a specific cable, since these types of cables are difficult to find in retail.

For a production run, off course one can buy the cable by the 1000ft roll.

Tinkerer