Homemade search coils

[johnyp]

Hello to all. Γεια σε όλους.

I would like to know when a shield is needed in a coil. Does it have to do with the frequency, the operation mode ( pi,bfo,ib ) or with the ground characteristics ?

Thank you. Ευχαριστώ.

PS. Σήμερα πήρα χαμπάρι πόσοι Έλληνες είναι μέλη εδώ !

[Carl-NC]

Always, due to ground effects. Also helps with RF noise.

[johnyp]

You mean that regardless if it is a pi machine or an ib one i must use shield? Doesn't shield drops receivers sensitivity? Thank you.

[plumabob]

Actually my experience is that it can actually increase sensitivity. I have witnessed this with my PI coils. I can run a lower delay once shielding is applied. My last 3 coils did exactly this.

Bob

[Carl-NC]

You mean that regardless if it is a pi machine or an ib one i must use shield? Doesn't shield drops receivers sensitivity? Thank you.

Yes, all detector types need shielding. A well-done shield does not affect sensitivity, and without it ground noise will kill sensitivity.

[plumabob]

So then Carl is it true then that with shielding the coil is quietier so now the smaller targets can be heard? or if you please Carl, How you explain my ability to drop my delay after shielding is applied?

Bob

[johnyp]

Ok. Thank you all. I will try different materials as shield using midediver4 and tracker-fm and post my results. Thank you

[maikl]

Yes, all detector types need shielding. A well-done shield does not affect sensitivity, and without it ground noise will kill sensitivity.

Hi Carl. Most hobbyists to build PI detector avoids to put a shield. They say it is slow and reduces the detector sensitivity. Where do they make a mistake?

[mikebg]

Hi Carl. Most hobbyists to build PI detector avoids to put a shield. They say it is slow and reduces the detector sensitivity. Where do they make a mistake?

Hi Maikl,
This equivalent network explains why we need shielding of search coil. Please ask me if you need explanation with words.

[Halo]

One question I have is about how to apply the shielding (scotch24 in my case).
Is it better to apply it horizontally around the circumference or to wind it around and around (with a slight angle)?

[maikl]

Hi Maikl,
This equivalent network explains why we need shielding of search coil. Please ask me if you need explanation with words.

Hi mikebg. Please explain to me your scheme. Which is how I see relates to the IB and not the PI design. Shield coil for IB detectors is necessary. IB detector can not be used if the coil does not shield. Some IB detectors require a metal box (connected to the negative pole) for electronics. While the PI detector quite tolerant on this issue and can be used in the field without the shield coils (more or less successfully). Most hobbyists have a problem (with PI detectors) when you put the shield on the coil. It seems that everyone has the same problem. What is it what it mistaken? Do you mistaken by putting "coaxial" cable from the box to the coil (as for the IB design), or by placing a shield need to change something on the PCB, or poorly connected cable with the shield coil, or...?

[Jose]

One question I have is about how to apply the shielding (scotch24 in my case).
Is it better to apply it horizontally around the circumference or to wind it around and around (with a slight angle)?

See this link, gave me good results.
Jose

http://www.geotech1.com/forums/attac...1&d=1294842196

[mikebg]

Explanation of equivalent network shown in posting #9

TX(i) is a current source pumping with energy TX tank.

Rab is an absorption resistance formed mainly by energy loss in soil. It is variable because depends on distance Soil - TX loop. It can not be eliminated by shielding. Only a TX "dismodulating" circuit can suppress its bad action.
See for example old Garrett TX circuits.

C1 is equivalent capacitance of TX tank circuit. For wideband metal detectors (PI) we need minimum capacitance. For examle resonance frequency of TX tank should be more than 25kHz (more than 100kHz for gold prospecting). For TX efficiency of narrow band metal detectors, we need large capacitance, for example 1uF (see old TX circuits of White's).

L1 and r1 are parameters of TX coil.

Ctx is capacitance between TX tank circuit and real ground (soil or water), noted here as GND. The capacitance is variable because depends on distance.

COM is an irreal ground, formed by metal housing of control box and shielding. It is noted as GND in sevice shematics, however this is not correct.

r & L represent parameters of a simple target, for example very thin bracelet.

Cm is the constant part of mutual capacitance between TX and RX tank. Note that there is a variable part of mutual capacitance formed by serial connection of Ctx and Crx.

M represents symbolic lossless ferromagnetic properties of environment. It is variable because depends of coil movement and distance. It can not be avoided by shielding.

Crx is capacitance between RX tank circuit and real ground (soil or water), noted here as GND. The capacitance is variable because depends on distance.

"e" is EMV (electromotive voltage) induced in RX coil. The RX coil has parameters L2 and r2.

C2 is equivalent capacitance of RX tank circuit. For wideband metal detectors (PI) we need minimum capacitance C2 (see above notes for C1). For narrow band metal detectors we also need minimal capacitance C2 in order to increase turns number of RX winding (to obtain stronger signal "e").

Rd is similar to Rab. Here it operates as variable damping resistance. To avoid its bad action (amplitude and angle modulation), we should connect in parallel to it a resistor (to decrease Q-factor of RX tank).

U is an ideal preamp unit with infinity input resistance. That means we need a damping resistor for decreasing Q-factor of RX tank circuit.

C_com-gnd is a capacitance between real ground and COM surfaces of detector circuit. It is variable because depends on distance. In parallel to this capacitance appear variable capacitances formed by operator OP and noted as C_com-op and C_op-gnd. We can avoid these capacitances using metal handle connected to COM surfaces and if the operator wears conductive shoes.

What makes the shield? It forms capacitances Ctx-com and Crx-com (not shown) instead Ctx-gnd and Crx-gnd (noted as Ctx and Crx in the equivalent network).

[simonbaker]

Explanation of equivalent network shown in posting #9

TX(i) is a current source pumping with energy TX tank.

Rab is an absorption resistance formed mainly by energy loss in soil. It is variable because depends on distance Soil - TX loop. It can not be eliminated by shielding. Only a TX "dismodulating" circuit can suppress its bad action.
See for example old Garrett TX circuits.

C1 is equivalent capacitance of TX tank circuit. For wideband metal detectors (PI) we need minimum capacitance. For examle resonance frequency of TX tank should be more than 25kHz (more than 100kHz for gold prospecting). For TX efficiency of narrow band metal detectors, we need large capacitance, for example 1uF (see old TX circuits of White's).

L1 and r1 are parameters of TX coil.

Ctx is capacitance between TX tank circuit and real ground (soil or water), noted here as GND. The capacitance is variable because depends on distance.

COM is an irreal ground, formed by metal housing of control box and shielding. It is noted as GND in sevice shematics, however this is not correct.

r & L represent parameters of a simple target, for example very thin bracelet.

Cm is the constant part of mutual capacitance between TX and RX tank. Note that there is a variable part of mutual capacitance formed by serial connection of Ctx and Crx.

M represents symbolic lossless ferromagnetic properties of environment. It is variable because depends of coil movement and distance. It can not be avoided by shielding.

Crx is capacitance between RX tank circuit and real ground (soil or water), noted here as GND. The capacitance is variable because depends on distance.

"e" is EMV (electromotive voltage) induced in RX coil. The RX coil has parameters L2 and r2.

C2 is equivalent capacitance of RX tank circuit. For wideband metal detectors (PI) we need minimum capacitance C2 (see above notes for C1). For narrow band metal detectors we also need minimal capacitance C2 in order to increase turns number of RX winding (to obtain stronger signal "e").

Rd is similar to Rab. Here it operates as variable damping resistance. To avoid its bad action (amplitude and angle modulation), we should connect in parallel to it a resistor (to decrease Q-factor of RX tank).

U is an ideal preamp unit with infinity input resistance. That means we need a damping resistor for decreasing Q-factor of RX tank circuit.

C_com-gnd is a capacitance between real ground and COM surfaces of detector circuit. It is variable because depends on distance. In parallel to this capacitance appear variable capacitances formed by operator OP and noted as C_com-op and C_op-gnd. We can avoid these capacitances using metal handle connected to COM surfaces and if the operator wears conductive shoes.

What makes the shield? It forms capacitances Ctx-com and Crx-com (not shown) instead Ctx-gnd and Crx-gnd (noted as Ctx and Crx in the equivalent network).

This is worth thinking about for me, because I also am not clear on how shields work.

When we think of coil voltage, are we not referring to the voltage across the coil? In other words, a voltage difference from one end to the other? When we wind the coil, the windings all fall together and basically of overlap. Where is the voltage that is relative to earth ground? Isn't the coil basically electrically neutral relative to earth ground because the plus and minus ends of the coil are very close together?

Also, your diagram represents the earth ground as a perfect conductor. Perhaps that is a good model, but I think of earth ground as being pretty high resistance. Also, is it realistic to model the earth ground as electrically the same as the PCB ground?

I can see the shield diffusing electrostatic charge from grass and such. I'm not clear on the ground "capacitive" effect.

Also, does the shield really increase the signal-to-noise ratio regarding RF noise in the frequency range of the TX signal? Wouldn't it reduce both equally?

-SB

[mikebg]

Hi mikebg. Please explain to me your scheme. Which is how I see relates to the IB and not the PI design. Shield coil for IB detectors is necessary. IB detector can not be used if the coil does not shield. Some IB detectors require a metal box (connected to the negative pole) for electronics. While the PI detector quite tolerant on this issue and can be used in the field without the shield coils (more or less successfully). Most hobbyists have a problem (with PI detectors) when you put the shield on the coil. It seems that everyone has the same problem. What is it what it mistaken? Do you mistaken by putting "coaxial" cable from the box to the coil (as for the IB design), or by placing a shield need to change something on the PCB, or poorly connected cable with the shield coil, or...?

Maikl,
The equivalent network in posting #9 is valid for all kind coils for metal detectors. When we use monocoil, we should short the capacitor Cm.
The equivalent network for monocoil at PI seems as shown below.
According theory of electricity, the self-inductance (monocoil) is specific case of mutual inductance (separated TX and RX windings). At self-inductance, the coefficient of coupling between TX and RX windings is maximal and mutual inductance is always positive and maximal. For metal detecting it means that monocoil is the worst sensor because generates maximal AIR signal.
In this case the target signal provokes minimal modulation index of source "e" (because AIR signal is maximal) and preamp U should have minimal gain (to avoid saturation by large signal "e").