detection depth

[Funfinder]

Still wondering about best coil size. Example: an eight inch coil can detect a target at sixteen inches, what diameter coil would detect the target the greatest distance and what distance would that be? Everything the same(inductance, current amplitude and profile, PI timings, noise level, etc.)except coil diameter. I charting some data with a 200mm and 300mm coil vs output from Hyperphysics. Entered twice the distance in Hyperphysics to get it to make sense. Still not close to my measured data. Think I'm doing something wrong with Hyperphysics. Wanted to chart distance vs amplitude with Hyperphysics for a 200, 300, 400 and 500mm coil to see if that would help determine best coil size. Any ideas where I'm going wrong(maybe the wrong law in Hyperphysics) or how to calculate best coil size for the example?

The main factor which really counts is the distance from one coil-end to the other.
Because for those low frequencies the "antennas" are far too tiny anyway.
Good AM radio antennas are usually over 50 meters in size!
No matter if IB, TR or PI - what counts is that the "inner field" will be disturbed good enough by metal objects.

And for "PI-reflection" a smaller coil would be even better if it comes to larger targets,
same as the sun gets better reflected by larger mirrors.

But smaller coils need a huge amount of sweeping which becomes a pretty "lousy" procedure with the slow PI pulses anyway.

btw. you can detect already super small stuff with a 40cm coil if the sensitivity, coil power and signal-noise ratio is good enough!

Principally seen there are limits in signal-noise so far which create some objects-ranges for different coils:

Coil-diameter divided with factor 100 gives smallest detectable object:
a 50cm coil still detects a 5mm nail

For most possible highest depth or distance its coil-diameter multiplied by 10

Which means that per instance the Jeohunter in air still detects with its 50cm coil a huge metal object from 5 meters distance
which I have tested. But the higher the mineralization, the stronger the whole signal-noise factor becomes blurred or weakened!

But as I wrote already:
It is not the coil-size itself but just how far is the distance from one (or more) coil ends to the other(s).

With a TR-setup you can place one small coil 10 meters away from another small coil and can find out if theres a huge deep metal object inbetween
as long as the mineralization is not too strong and as long as the radiation power or the receiver itself is not too weak.

[green]

I have lots of target strength vs distance charts plotted out (VLF machine, 11" x 7" bi-axial DD coil) . So I laboriously plotted your 200mm Hyperphys data onto paper to compare. And you've definitely got something wrong. All my small targets dropped away much faster than your HP figures. The only target that resembled your HP data was a 3" long ferrite rod, end-on to the coil. It was actually a good match (when close to the coil) , but I don't think there's anything to learn from the match. The rod is not a small target - when it's at 0", some of it is 3" away.
It looks like your HP model isn't correct when targets are close to the coil - there just isn't the steep fall-off in signal strength as distance increases.

But seeing as I don't know how you've modelled your coil & target, I've no idea what's gone wrong.

Probably doing it all wrong. Entered 1amp, .15 R for 300mm coil, .3m for .15meter distance(out and back), charted Gauss*E5(amplitude) at 150mm distance. Charted distance vs amplitude(entered 2 x charted distance)for the other data points.

[green]

Probably doing it all wrong. Entered 1amp, .15 R for 300mm coil, .3m for .15meter distance(out and back), charted Gauss*E5(amplitude) at 150mm distance. Charted distance vs amplitude(entered 2 x charted distance)for the other data points.

Want to add calculated distance vs amplitude data to the measured chart. The data was measured with two targets, 10grain nugget and 1x1inch aluminum can side. The measured amplitude was multiplied by 300uH/coil inductance for equal coil inductance, coil current in a control loop for equal current amplitude and profile. Measured data: coil diameter effected slope, target effected amplitude. Could someone suggest a method to calculate target distance vs amplitude data to add to the chart? Don't know if target matters, trying to match measured slope. What I tried with Hyperphysics didn't work probably because I don't have a clue, just guessing.

[Skippy]

You're definitely doing the simulation wrong, but I've tried doing it 'right' and still don't get good-looking results.
For the benefit of everyone else, this is the hyperphysics page Mr. Green is using for his maths:
http://hyperphysics.phy-astr.gsu.edu...ic/curloo.html

What I suggest is:
Use a current that gives a field strength of 1.00, makes the sums easier. So enter I = 23.874 for the 0.15 radius coil. Then at Z=0, field = 1.00 Gauss. Now at target z = 0.30 m, field = 0.0894 G. Now square this, to account for the 'round trip' back to the receive coil, and you get ReturnSignal = 8 x 10E-3 ( = 1/125 )
Sadly this is nothing like my measured DD coil figures. Going from (an estimated) Z=0.00 out to Z = 0.30 m for a small target like a US 1 cent coin produces a signal that drops off from 3000 to 0.5, ie. 1/6000. A whole lot of difference.

However... this is a DD, not a monocoil. And a coin isn't a perfect target, it's thin and flat, and not tiny in size. I've never tested any spherical targets, and obviously monocoils don't fit IB VLF's.

I've never actually tried modelling the coil/target behaviour, as for an elliptic DD, the maths is complicated.

[Tinkerer]

As a rule of thumb, for target depth/distance, you can use the width of a TX coil as diameter. So for a DD coil the radius counts for the diameter.
An important factor is the relationship of target surface area presented to the coil magnetic field and the surface area of the coil. If we imagine the magnetic field as field lines, the target response is relative to the amount of field lines that hit the target. A spherical target is hit by fewer field lines than a flat target of similar mass. Experimenting with lead spheres and pounding them flat, teaches a lot.
When the target is thinner than the skin effect, there is non-linearity in the target response.
Simple wire rings are the best targets.

[Skippy]

I can see that a wire ring would be an easier target to mathematically model than any solid shape. But I would expect it to have properties similar to a coin, particularly when oriented 'on-edge'.
I would prefer to avoid using custom-made test targets like rings, as they are hard to reproduce. I have access to a lathe, so could machine loops out of sheet metal, or coins, but most other folks would not be able to replicate them.
Spherical items make good targets, but are also hard to obtain, unless they are steel ball-bearings, and steel/iron is not generally good for testing.

[green]

You're definitely doing the simulation wrong, but I've tried doing it 'right' and still don't get good-looking results.
For the benefit of everyone else, this is the hyperphysics page Mr. Green is using for his maths:
http://hyperphysics.phy-astr.gsu.edu...ic/curloo.html

What I suggest is:
Use a current that gives a field strength of 1.00, makes the sums easier. So enter I = 23.874 for the 0.15 radius coil. Then at Z=0, field = 1.00 Gauss. Now at target z = 0.30 m, field = 0.0894 G. Now square this, to account for the 'round trip' back to the receive coil, and you get ReturnSignal = 8 x 10E-3 ( = 1/125 )
Sadly this is nothing like my measured DD coil figures. Going from (an estimated) Z=0.00 out to Z = 0.30 m for a small target like a US 1 cent coin produces a signal that drops off from 3000 to 0.5, ie. 1/6000. A whole lot of difference.

However... this is a DD, not a monocoil. And a coin isn't a perfect target, it's thin and flat, and not tiny in size. I've never tested any spherical targets, and obviously monocoils don't fit IB VLF's.

I've never actually tried modelling the coil/target behaviour, as for an elliptic DD, the maths is complicated.

Many Thanks, Tried your method with a multiplier to match recorded data at 25mm. Looks good, might need to retest 300mm coil at longer distances.

[Skippy]

That's better. It might be worth plotting the hyperphys data for Z = 0, and Z slightly beyond your max real measured data, next time.

I've just been trying to work out an 'equivalent' mono coil size that matches the measured data for my Fisher/Teknetics 11" x 7" elliptic DD coil. It provisionally appears to be about 160mm diameter, which accounts for the major difference between it's performance and Green's 300mm diam coil figures. 160mm is 6.3" in Imperial, quite small, yet it's performance isn't lacking.
I'll do a bit more experimenting, and also see what the 5" round DD coil behaves like.

[green]

Still wondering about best coil size. Example: an eight inch coil can detect a target at sixteen inches, what diameter coil would detect the target the greatest distance and what distance would that be? Everything the same(inductance, current amplitude and profile, PI timings, noise level, etc.)except coil diameter. I charting some data with a 200mm and 300mm coil vs output from Hyperphysics. Entered twice the distance in Hyperphysics to get it to make sense. Still not close to my measured data. Think I'm doing something wrong with Hyperphysics. Wanted to chart distance vs amplitude with Hyperphysics for a 200, 300, 400 and 500mm coil to see if that would help determine best coil size. Any ideas where I'm going wrong(maybe the wrong law in Hyperphysics) or how to calculate best coil size for the example?

Thanks to Skippy an attempt at charting the coils. Example: a 200mm coil can detect a target at 400mm, what diameter coil would detect the target the greatest distance? The 500mm coil has the same signal amplitude at almost 600mm distance. Would need to chart some larger coils to answer the question assuming the chart is correct and noise level didn't increase.

A target just detectable at 100mm with the 200mm coil wouldn't be detected with the 400 or 500mm diameter coils at any distance.

[Dutchie]

Maybe Biot-Savart is not the correct way because it is for constant current and not moving.
probably the following set of formulas is for our purpose.

https://en.wikipedia.org/wiki/Jefimenko%27s_equations

[green]

Coil noise test. Coils not shielded, connected to a differential amplifier(gain about 450) transmitter disabled. Scope connected to last stage of amplification after integrator, 10usec target and EF sample , 1000 samples/second. Noise increases with coil diameter on the bench(scope is the only thing turned on near the coil, about 4feet). Is it possible to shield a coil from EMI to reduce the noise level? Thinking might try to EMI shield the flat spiral coil(suggestions appreciated). Should I expect a similar increase in noise with increasing coil size with a detector on the ground? Is the noise trace shape similar to other PI detectors, shape not amplitude? Maybe someone could post some scope pictures(.5 or 1sec/div time scale) of the last stage of amplification before audio at 2 or 3 div p-p amplitude.

[Tinkerer]

Thanks to Skippy an attempt at charting the coils. Example: a 200mm coil can detect a target at 400mm, what diameter coil would detect the target the greatest distance? The 500mm coil has the same signal amplitude at almost 600mm distance. Would need to chart some larger coils to answer the question assuming the chart is correct and noise level didn't increase.

A target just detectable at 100mm with the 200mm coil wouldn't be detected with the 400 or 500mm diameter coils at any distance.

Try dividing the coil field strength by the coil surface area and the coil surface area by the target surface area. Compare the ratios with different target/coil ratios.

[Tinkerer]

Coil noise test. Coils not shielded, connected to a differential amplifier(gain about 450) transmitter disabled. Scope connected to last stage of amplification after integrator, 10usec target and EF sample , 1000 samples/second. Noise increases with coil diameter on the bench(scope is the only thing turned on near the coil, about 4feet). Is it possible to shield a coil from EMI to reduce the noise level? Thinking might try to EMI shield the flat spiral coil(suggestions appreciated). Should I expect a similar increase in noise with increasing coil size with a detector on the ground? Is the noise trace shape similar to other PI detectors, shape not amplitude? Maybe someone could post some scope pictures(.5 or 1sec/div time scale) of the last stage of amplification before audio at 2 or 3 div p-p amplitude.

To avoid loading the coil with too much coil wire to shield capacitance, place a spacer of about 1/8 to 3/16 between the coil wire and the shield.

[green]

Try dividing the coil field strength by the coil surface area and the coil surface area by the target surface area. Compare the ratios with different target/coil ratios.

Example: a 200mm coil can detect a target at 400mm, what diameter coil would detect the target the greatest distance?

Trying to understand how to use your reply to answer my question or how you use it. Can you give an example? [same coil inductance and current profile]200mm coil(23turns, 1amp peak) 400mm coil(15 turns, 1amp peak)[used coil calculator, both 325uH]. Target signal not proportional to target area, recorded integrator out for US nickels and quarters, nickel(smaller diameter) has a higher signal than quarter. Signal for three coins probably 3 times 1 coin signal(meter resolution), Use 30mm x 30mm for target that can be detected at 400mm with the 200mm coil. What distance would the 400mm coil give the same signal strength or how would you look at it? Just trying to learn or understand. I have seen coin detection distances reported in other threads that would be hard for me to detect, trying to understand the variables that I would need to change to get there.

[green]

Charted distance vs signal for the figure eight coil I've been using. Two Rx 200mm Rx coils covered by a oval Tx. The two 200mm Rx coils for the figure eight coil same as 200mm mono coil used in above thread charts. The slopes were close, amplitude was different because of different circuits(gain, current and sample times).