Effect of water on high Q metal detector

[revmaster]

Hi,

Recently i build a metal detector based on high Q principle, surprisingly got very good results much much better than PI detectors which i had worked on earlier but it had a lot of problem with water. It is very unstable in water , tried sheilding also but that didnt helped .

Any ideas on improvement & why is it happening ??

[Unregistered]

?

Hi,

Recently i build a metal detector based on high Q principle, surprisingly got very good results much much better than PI detectors which i had worked on earlier but it had a lot of problem with water. It is very unstable in water , tried sheilding also but that didnt helped .

Any ideas on improvement & why is it happening ??

What is "metal detector based on high Q principle"???

[revmaster]

It is based on eddy current losses, basically you check losses in tx due to presence of metal.

[Aziz]

Hi,

It is based on eddy current losses, basically you check losses in tx due to presence of metal.

Do you mean the Off-Resonance Detector?

A simple LC-Oscillator gets very little power to keep the oscillator alive. Small energy losses will dump the peak voltage on the coil. Often, the oscillator gets totally out of work, when the target is large.

Is this what you mean with high-Q MD?

[revmaster]

Hi,



Do you mean the Off-Resonance Detector?

A simple LC-Oscillator gets very little power to keep the oscillator alive. Small energy losses will dump the peak voltage on the coil. Often, the oscillator gets totally out of work, when the target is large.

Is this what you mean with high-Q MD?

Hi Aziz,
Yes this is the principle, the oscillator is critically balanced to get max gain. But the problem is that it considers water as a target when coil is dipped in the water.
Results in free air were astonishing it was 30% ahead than the pulse induction based detectors.

May be it is inherent property of this technology...... or am i missing something ?

[Aziz]

Hi revmaster,

Hi Aziz,
Yes this is the principle, the oscillator is critically balanced to get max gain. But the problem is that it considers water as a target when coil is dipped in the water.
Results in free air were astonishing it was 30% ahead than the pulse induction based detectors.

May be it is inherent property of this technology...... or am i missing something ?

In water, you have to adjust the oscillator current drive different to air. So giving with a 10-turn pot more range to cover most reagions.

Indeed, I had similar results with this easy but very extremely sensitive type of MD. Particularly the sensitivity to small metal objects (gold nuggets) gives a new meaning to such technology.
The reason for the sensitivity lies on the following effect:
On every wave radiant of the frequency the small effects getting more into a sum with more wave radiants. So the best sensitivity is given, when you swing the search coil slow (giving more wave radiants for detection).

Of course, to avoid the eddy currents of the search coil itself, you should use a litz wire (HF wire). Every thin wire of the litz wire is isolated from each other.

Aziz

[revmaster]

Hi Aziz,

Thanks for your kind replies.
1) Effect of water reduced consideribly once i reduced the freq of operation, i was at 20khz earlier now switched to 6khz and the effect is prominent. But there is reduction in sensitivity, may be need to do more tweaking.
2) Wire used is 20 swg enamled wire. How would litz wire in this case as i dont think capacitance of wire is of concern here.

How would this detector perform in mineralized / wet soils ? I have not tested it yet, any ideas ?

Cheers,
Michael

[Aziz]

Hi Michael,

Hi Aziz,

Thanks for your kind replies.
1) Effect of water reduced consideribly once i reduced the freq of operation, i was at 20khz earlier now switched to 6khz and the effect is prominent. But there is reduction in sensitivity, may be need to do more tweaking.
2) Wire used is 20 swg enamled wire. How would litz wire in this case as i dont think capacitance of wire is of concern here.

How would this detector perform in mineralized / wet soils ? I have not tested it yet, any ideas ?

Cheers,
Michael

2) Wire capacitance:
The litz wires are connected parallel, so the total coil capacitance depends only on the lenght of the coil-wire. Anyway, the capacitance of the coil does not really affect the results due to the processing technique. Litz wires used generally on radio antenna or coil construction to increase the Q factor of the coil/antenna. Typically LW/MW antennas are constructed with such litz wires (HF wire). So this can be dramatically increase the sensitivy, if you change the coil with litz-wire-coil. Eddy currents of the coil will be reduced very much. You only will detect eddy current caused by the material nearby the search coil.

I have tried to supply the LC-oscillator with a good stabilized reference voltage controller. Indeed, this gives much more stability of the detector. The reference voltage do not need to drive much current, because the LC-oscillator is driven with critical unsufficient and therefore less current.
Another tip is to decouple all the voltage supply to the electronics parts. Decoupling the supply voltage for oscillator, ampfifier, etc. give a bit more success. Also the high resistant graphite shielding of the coil brings more stable results.

The mineralized ground will also be detected. But you can not avoid the detecting. Its a physical effect (eddy current).

Try some of my tips.
Regards,
Aziz

[revmaster]

Hi Michael,



2) Wire capacitance:
The litz wires are connected parallel, so the total coil capacitance depends only on the lenght of the coil-wire. Anyway, the capacitance of the coil does not really affect the results due to the processing technique. Litz wires used generally on radio antenna or coil construction to increase the Q factor of the coil/antenna. Typically LW/MW antennas are constructed with such litz wires (HF wire). So this can be dramatically increase the sensitivy, if you change the coil with litz-wire-coil. Eddy currents of the coil will be reduced very much. You only will detect eddy current caused by the material nearby the search coil.

I have tried to supply the LC-oscillator with a good stabilized reference voltage controller. Indeed, this gives much more stability of the detector. The reference voltage do not need to drive much current, because the LC-oscillator is driven with critical unsufficient and therefore less current.
Another tip is to decouple all the voltage supply to the electronics parts. Decoupling the supply voltage for oscillator, ampfifier, etc. give a bit more success. Also the high resistant graphite shielding of the coil brings more stable results.

The mineralized ground will also be detected. But you can not avoid the detecting. Its a physical effect (eddy current).

Try some of my tips.
Regards,
Aziz

Thank you Aziz,
I will try litz wire and update you with results for sure.
What freq are you working on right now i am at 6khz was at 20khz earlier ?
The problem now is that to achieve low freq i had to increase C which has resulted in narrowing of dynamic range of operation voltage. Now i am concerned about effect of temperature on oscillator, i have provided temperature compensation to osc circuit but dont know how would coil react with change in temperature......

Any ideas about reference controlled oscillator, currently mine is a simple LC oscillator & i adjust the drive manually using a trimpot.

Regards,
Michael

[Aziz]

Hi Michael,

Thank you Aziz,
I will try litz wire and update you with results for sure.
What freq are you working on right now i am at 6khz was at 20khz earlier ?
The problem now is that to achieve low freq i had to increase C which has resulted in narrowing of dynamic range of operation voltage. Now i am concerned about effect of temperature on oscillator, i have provided temperature compensation to osc circuit but dont know how would coil react with change in temperature......

Any ideas about reference controlled oscillator, currently mine is a simple LC oscillator & i adjust the drive manually using a trimpot.

Regards,
Michael

I do not exactly know, what the operating frequency was. It must lie between 12 and 20 kHz. So I tried on different frequency ranges (5-30 kHz). I was given up this technique due to too much sensitivity to all effects (ground, capacitance, instability of the LC oscillator, no disc, etc.)
You should not increase the C. You should increase instead the inductivity of the coil L. So you can increase the Q factor of the coil. Q factor is also dependend on frequency. The more, the more Q factor.

Look for A/D converter circuits. They need also a very stable reference voltage. This voltage is then buffered via opamp for driving more current. Some voltage references deliver enough drive current and you can spare the voltage buffering. The power supply of the LC circuit should be totally decoupled from the other parts of the electronics.

Of course, temperature changes to coil gets more into account with such technique. The resistance of the coil will change and therefore the driving coil current. You should isolate the coil good for temperatur changes.

Remember, this detector type is very sensitive eddy current detector.
Aziz