Let's made a PC-base metal detector with usb interface !!!

Let's made a PC-base metal detector with usb interface !!!

Aziz,
Track your interesting project and I have one question for you:
Can I use the Asus eee notebook PC 4G in the MD? The advantage is low weight (0.92kg) and low price.Thanks advance to the response

[simonbaker]

Hi all,

I made a pulsed VLF with a power MOSFET last days. The occured effects were absent. But the coil assembly is quite critical. I have doubts to meet the critical specifications. PI coils are most enviable to me and of course very very easy to build.

Digital lock-in amp VLF MD is still possible. One have to reduce the sensitivity to minimize the disturbing effects. This will reduce the detector possibilities to a yet another VLF MD. That makes not sense to me to follow this project anymore.

So I tend to enter the PI technology now and will try to implement lock-in amp or boxcar averager to increase performance. This project was a good experimental work to gain more understanding in noise reduction and increasing SNR.

Aziz

I hope you continue to experiment with VLF detector. It would be good to understand those effects fully. The effects must be the same for all VLF detectors. Your digital processing has best chance to see through the effects and is a good tool for analyzing them and designing coils to physically minimize the effects also.

Cheers,

-SB

[Aziz]

Hi all,

I made a pulsed VLF with a power MOSFET last days. The occured effects were absent. But the coil assembly is quite critical. I have doubts to meet the critical specifications. PI coils are most enviable to me and of course very very easy to build.

Digital lock-in amp VLF MD is still possible. One have to reduce the sensitivity to minimize the disturbing effects. This will reduce the detector possibilities to a yet another VLF MD. That makes not sense to me to follow this project anymore.

So I tend to enter the PI technology now and will try to implement lock-in amp or boxcar averager to increase performance. This project was a good experimental work to gain more understanding in noise reduction and increasing SNR.

Aziz

[Aziz]

Hi Qiaozhi,

What is your transmit frequency, and what value capacitors do you have across the coils?

The transmit frequency is in the range of 16-21 kHz. Don't know the exact frequency because I am using lots of different coils.
Aziz

[Qiaozhi]

Hi Qiaozhi,



The above green/black coil is TX coil (center tapped). It has 2x14 windings wound on a 16 cm pot and bend to D form.
The below red/brown coil is RX coil (center tapped, connected in series). It has 2x15 windings (total 30 windings) and wound on a 16 cm pot.

Well, I tried many options: floating shielding (potential free), shielding connected to ground (0 V), etc. It does not change anything.

It seams a slow changing effect (high time constant): It looks like a typical charging time behaviour of capacitors. The effect increases much with resonant mode (LC tank). Of course there is an amplification on resonant mode.

Now trying some typical soils: Wet sand, highly mineralized sand (iron oxyd), humus soil, etc.. I am swinging some portion of soils above the coil to see the effects. Coil assembly is still experimental and I can not go field testing at the moment.

Still looking for the source of this strange effect.

Regards,
Aziz

What is your transmit frequency, and what value capacitors do you have across the coils?

[simonbaker]

Hi Qiaozhi,



The above green/black coil is TX coil (center tapped). It has 2x14 windings wound on a 16 cm pot and bend to D form.
The below red/brown coil is RX coil (center tapped, connected in series). It has 2x15 windings (total 30 windings) and wound on a 16 cm pot.

Well, I tried many options: floating shielding (potential free), shielding connected to ground (0 V), etc. It does not change anything.

It seams a slow changing effect (high time constant): It looks like a typical charging time behaviour of capacitors. The effect increases much with resonant mode (LC tank). Of course there is an amplification on resonant mode.

Now trying some typical soils: Wet sand, highly mineralized sand (iron oxyd), humus soil, etc.. I am swinging some portion of soils above the coil to see the effects. Coil assembly is still experimental and I can not go field testing at the moment.

Still looking for the source of this strange effect.

Regards,
Aziz

Can you be more detailed? For instance: if you move your hand suddenly over coil, are you saying that the effect slowly accumulates? How many seconds before it is steady?

Also, exactly what are you measuring for this "effect"? Is it the computations from your lock-in amplifier software, or do you see it on an oscilloscope?

If it is the computations, can you be more detailed what the computations are? I assume you drive the TX coil with a sine wave from the audio output, and sample the RX coil voltage with the audio input.

Then I assume you do something like generate a reference signal from the TX sinewave, maybe adding some phase shift, and multiply times the RX signal and integrate for a time constant. You do again maybe with a quadrature signal so you get two different integrals which can be used to derive the phase and amplitude of the target signal.

Perhaps it would help if you could plot the response you are seeing and show us this "capacitive-like" response you are finding. It would be specially good if you could synchronize some video of the actual target (e.g. your hand) as you move it over the coil.

By the way, what is the time constant you are using for your integration? Is it possible you have it longer than you think and that is causing a very slow response?

If you pass a metal target such as a coin over the coil, how does the response compare?

Regards,


-SB

[Aziz]

Hi Qiaozhi,

Hi Aziz,

What are the inductance values for your coils, and do you have any tuning capacitors across them?
Also - I assume that you have actually connected the shield to the 0V line. Stupid question, I know ... but I've got to ask it.

The above green/black coil is TX coil (center tapped). It has 2x14 windings wound on a 16 cm pot and bend to D form.
The below red/brown coil is RX coil (center tapped, connected in series). It has 2x15 windings (total 30 windings) and wound on a 16 cm pot.

Well, I tried many options: floating shielding (potential free), shielding connected to ground (0 V), etc. It does not change anything.

It seams a slow changing effect (high time constant): It looks like a typical charging time behaviour of capacitors. The effect increases much with resonant mode (LC tank). Of course there is an amplification on resonant mode.

Now trying some typical soils: Wet sand, highly mineralized sand (iron oxyd), humus soil, etc.. I am swinging some portion of soils above the coil to see the effects. Coil assembly is still experimental and I can not go field testing at the moment.

Still looking for the source of this strange effect.

Regards,
Aziz

[simonbaker]

Hi SB and others,

I have made the coil shieldings with a AL mesh wire. Mesh wire is cut into 1-2 cm stripes and wound around the coils. There is allways a gap in the shielding (to avoid closed loops). This will allow to lower the residual RX coil voltage much (see below). This is a practically proofen good fact to lower the residual RX coil voltage.

But:

Still have other disturbing effects (hand detection, ground effects) which tend to be a K.O. (knockout) criteria for this project. I have no idea at the moment how these disturbing effects can be cancelled. The lock-in amplifier is obviously too much sensitive for such effects.

There is no difference between hand detection and ground detection. So the reason must be somehow capacitive or other type. I am sure, the capacitive effects tend to be much higher when measuring ultra tiny signals.

I will follow the other way now: the brute force method. Increasing inductive effects to dominate over capacitive effects. I will force more current through the coil by switching it like a PI. The switching clock must be derived from the sound-card output. It will be switched in the VLF range (10 - 24 kHz).

Aziz

I never heard ground detection is capacitance effect -- I thought due to mineralization, moisture, or other inductive effects. Maybe you do not have the problem you think? Just normal ground balancing type stuff.

Regards,

-SB

hello all looking

the examiner it is a made pc metal detector usb interface. please looking the web examiner www.imagelocators.com examiner very best detector i have a one examiner .examiner geoelectron.and metal detector and magnetometer.

[Qiaozhi]

Hi SB and others,

I have made the coil shieldings with a AL mesh wire. Mesh wire is cut into 1-2 cm stripes and wound around the coils. There is allways a gap in the shielding (to avoid closed loops). This will allow to lower the residual RX coil voltage much (see below). This is a practically proofen good fact to lower the residual RX coil voltage.

But:

Still have other disturbing effects (hand detection, ground effects) which tend to be a K.O. (knockout) criteria for this project. I have no idea at the moment how these disturbing effects can be cancelled. The lock-in amplifier is obviously too much sensitive for such effects.

There is no difference between hand detection and ground detection. So the reason must be somehow capacitive or other type. I am sure, the capacitive effects tend to be much higher when measuring ultra tiny signals.

I will follow the other way now: the brute force method. Increasing inductive effects to dominate over capacitive effects. I will force more current through the coil by switching it like a PI. The switching clock must be derived from the sound-card output. It will be switched in the VLF range (10 - 24 kHz).

Aziz

Hi Aziz,

What are the inductance values for your coils, and do you have any tuning capacitors across them?
Also - I assume that you have actually connected the shield to the 0V line. Stupid question, I know ... but I've got to ask it.

[Aziz]

Hi SB and others,

I have made the coil shieldings with a AL mesh wire. Mesh wire is cut into 1-2 cm stripes and wound around the coils. There is allways a gap in the shielding (to avoid closed loops). This will allow to lower the residual RX coil voltage much (see below). This is a practically proofen good fact to lower the residual RX coil voltage.

But:

Still have other disturbing effects (hand detection, ground effects) which tend to be a K.O. (knockout) criteria for this project. I have no idea at the moment how these disturbing effects can be cancelled. The lock-in amplifier is obviously too much sensitive for such effects.

There is no difference between hand detection and ground detection. So the reason must be somehow capacitive or other type. I am sure, the capacitive effects tend to be much higher when measuring ultra tiny signals.

I will follow the other way now: the brute force method. Increasing inductive effects to dominate over capacitive effects. I will force more current through the coil by switching it like a PI. The switching clock must be derived from the sound-card output. It will be switched in the VLF range (10 - 24 kHz).

Aziz

[simonbaker]

Hi all,

I have built another low capacitance DD coil today. Without the shielding on TX & RX, the RX coil signal can be adjusted to zero volt. But the capacitance effects came again: hand detection test failed totaly.

After shielding first the RX coil, the residual RX voltage rises much. Shielding of TX coil increases the residual RX voltage also much.

Other effects arise with TX & RX shieldings: Even the coils are totally shielded with AL foil, hand detection test failed again.

I have to test the coil resistance further more (Q of coil).
Aziz

Aziz: does capacitive effect have same phase as inductive (metal target) effect, or different? Can you tell them apart? I keep wondering if you are detecting your hand as conductive target, like wet sand.

Capacitive effect depends on charged bodies. I am not clear on this, but normally shielding is to protect coil from charged bodies outside the coil, such as grass, etc, is that right? If you ground yourself and bring hand near coil, should not be much capacitive effect, should there? Where is the charge? Unless perhaps whole metal detector accumulates a charge somehow and needs to be discharged.

Of course, ends of coil have voltage difference, but so close together not enough electric field to matter at a distance, is there?

Can you be sure the "hand effect" is capacitive?

Regards,

-SB

[Aziz]

Hi all,

I have built another low capacitance DD coil today. Without the shielding on TX & RX, the RX coil signal can be adjusted to zero volt. But the capacitance effects came again: hand detection test failed totaly.

After shielding first the RX coil, the residual RX voltage rises much. Shielding of TX coil increases the residual RX voltage also much.

Other effects arise with TX & RX shieldings: Even the coils are totally shielded with AL foil, hand detection test failed again.

I have to test the coil resistance further more (Q of coil).
Aziz

[simonbaker]

Hi SB and others,



Yes, the seen parallel capacitance to the coil itself (Cpar). I think, it is called parasitic capacitance.
I was totally wrong, that the shield of the TX coil would not affect the residual RX coil voltage. But it does much. Both TX and RX coil itself should have low parallel (parasitic) capacitance. Due to electric field of TX coil, these capacitances will allow to flow electric induced currents to the RX coil. This is the reason for some residual RX coil voltage and the affection for ground effects. The coil capacitance will be of course increased due to additional faraday shielding.

I made the hand detection test without faraday shielding (TX & RX) with low capacitance coils: no detection!
Lowering the coil capacitances will reduce ground effects. This also will allow to minimize the residual RX coil voltage further and the RX signal amplification can be increased without saturating the op-amp output (= more sensitivity).


Additional capacitive coupling between TX and RX coil.




One can use an insulated wire to keep the wire distances to each other as far as possible. To minimize the proximity and eddy current effects, litz wire (tin plated only) with insulation could be used such as in PI coils. HF litz wire (each strand of wire is enamelled) makes not sense, if you use standard tin plated litz wire such as used for electrical installation and cabling (much cheaper and good available ).
Special winding techniques could be used to minimize the coil capacitance, but it doesn't make any sense for VLF detectors (the coils becoming much bigger).

PCB coils are not recommended: too much coil resistance, low Q.

I have to make some more fundamental tests yet.

Regards,

Aziz

I can't understand it really. Spacing wires apart in each coil makes better nulling and less capacitive effects to nearby objects? It boggles my brain. But very interesting if we can make use of it.

Spacing wires and reducing capacitance would raise the self-resonance frequency of a coil, not sure if that has anything to do with it.

I hope you keep experimenting with it and explaining it in detail.

Cheers,

-SB

[Aziz]

Hi SB and others,

Interesting! I had thought the residual voltage might be partly due to the capacitance between the TX and RX coils where they almost touch. You are saying it is the capacitance in each coil itself (parasitic)?

Yes, the seen parallel capacitance to the coil itself (Cpar). I think, it is called parasitic capacitance.
I was totally wrong, that the shield of the TX coil would not affect the residual RX coil voltage. But it does much. Both TX and RX coil itself should have low parallel (parasitic) capacitance. Due to electric field of TX coil, these capacitances will allow to flow electric induced currents to the RX coil. This is the reason for some residual RX coil voltage and the affection for ground effects. The coil capacitance will be of course increased due to additional faraday shielding.

I made the hand detection test without faraday shielding (TX & RX) with low capacitance coils: no detection!
Lowering the coil capacitances will reduce ground effects. This also will allow to minimize the residual RX coil voltage further and the RX signal amplification can be increased without saturating the op-amp output (= more sensitivity).

Also interesting you say more capacitance between windings means more sensitivity to nearby conductive objects. Why would that be?

Additional capacitive coupling between TX and RX coil.

It also brings up the subject of winding. If windings at one end of the coil are near windings at the other, there is more voltage between them, and thus more effect you would think. Do you have any special ideas on winding layout for minimum capacitive effect?
Regards,
-SB

One can use an insulated wire to keep the wire distances to each other as far as possible. To minimize the proximity and eddy current effects, litz wire (tin plated only) with insulation could be used such as in PI coils. HF litz wire (each strand of wire is enamelled) makes not sense, if you use standard tin plated litz wire such as used for electrical installation and cabling (much cheaper and good available ).
Special winding techniques could be used to minimize the coil capacitance, but it doesn't make any sense for VLF detectors (the coils becoming much bigger).

PCB coils are not recommended: too much coil resistance, low Q.

I have to make some more fundamental tests yet.

Regards,

Aziz