VLF metal detector

[david]

Hi, everybody,

I am a new guy in this field of hand-held hobby-type metal detectors. I have some shallow knowledge about the units. I think that VLF metal detectors is easier to understand than PI. The VLF metal detector often has two coils, which are named as TX and RX. The circuit transmits suqare wave or sinusoid wave to drive Tx coil, and form a alternating magnetic field. The coil arrangement make it reach induction balance, so if there's no target metal object, no signal should exist in Tx coil circuit in IB theory.
The discrimination function is amazing, it can "see through" the ground and tell us what is there. I know that it is prevailing to implement discrimination in software algorithm for modern metal detectors. Does anybody know something about software algorithm, or share some information about that, and I am eager to do some research on that with MATLAB.
Sorry for my poor english. Thanks.

[Qiaozhi]

Hi, everybody,

I am a new guy in this field of hand-held hobby-type metal detectors. I have some shallow knowledge about the units. I think that VLF metal detectors is easier to understand than PI. The VLF metal detector often has two coils, which are named as TX and RX. The circuit transmits suqare wave or sinusoid wave to drive Tx coil, and form a alternating magnetic field. The coil arrangement make it reach induction balance, so if there's no target metal object, no signal should exist in Tx coil circuit in IB theory.
The discrimination function is amazing, it can "see through" the ground and tell us what is there. I know that it is prevailing to implement discrimination in software algorithm for modern metal detectors. Does anybody know something about software algorithm, or share some information about that, and I am eager to do some research on that with MATLAB.
Sorry for my poor english. Thanks.

Have a look at some of the information here on Geotech ->
http://www.geotech1.com/cgi-bin/page...&file=main.dat
This may help you to understand how discrimination works.

[dfbowers]

Hi, everybody,

I am a new guy in this field of hand-held hobby-type metal detectors. I have some shallow knowledge about the units. I think that VLF metal detectors is easier to understand than PI. The VLF metal detector often has two coils, which are named as TX and RX. The circuit transmits suqare wave or sinusoid wave to drive Tx coil, and form a alternating magnetic field. The coil arrangement make it reach induction balance, so if there's no target metal object, no signal should exist in Tx coil circuit in IB theory.
The discrimination function is amazing, it can "see through" the ground and tell us what is there. I know that it is prevailing to implement discrimination in software algorithm for modern metal detectors. Does anybody know something about software algorithm, or share some information about that, and I am eager to do some research on that with MATLAB.
Sorry for my poor english. Thanks.

Hi, David.

I have been tinkering for a bit, and it wasn't that hard to figure out. Here is what I did with a simple off the shelf Arduino UNO.
http://www.geotech1.com/forums/showt...t=18958&page=2

If I can only figure out how to get the Arctan function to work with "C" programming I could get the true phase shift numbers!!

Basically, if you take multiple samples of the signal coming from the all metal channel and the disc channel simultaneously and plot them against each other on a graph, (X vs Y) you get sloped line. The more samples that are taken, the more accurate the results..
http://www.youtube.com/watch?v=m9DuMekVwic

(For high conductors, the disc channel will be stronger and for low conductors, the all metal channel will be stronger).

Then, take the arctan of that slope and you get the phase relationship.

[Davor]

There are some simple tricks that do not require arctan to extract phase. Classic phase detector uses analogue multiplication of clipped I and Q channel information to do the same. Trick is that multiplication of two square-like signals is a triangle like result, e.g. phase information is a linear function. In case of a phase shifted sinewawe (VLF MD) a product of such clipping, multiplication, averaging ... is a direct phase readout.

[david]

Have a look at some of the information here on Geotech ->
http://www.geotech1.com/cgi-bin/page...&file=main.dat
This may help you to understand how discrimination works.

Thanks, Qiaozhi. I have read your threads and you are experienced. I have read some patents in Geotech website you mentioned about discrimination before. As I know, the signal from RX coil will be demodulated into two component, which are name as R(resistance) and X(reactance), Then the ratio between R and X determine the phase shift. and the discrimination is due to this phase shift. right? I got a Bounty Hunter Platium (maybe the other name is Teknetics gamma?), I disassembled it and use it to collect some raw data of metal outside. This raw data measurement test point is after the pre-Amp, band-filter, synchronized demodulation. maybe the envelop of RX sinusoid, and I think they are just two component R and X. Please see my attachment files.

[Qiaozhi]

Thanks, Qiaozhi. I have read your threads and you are experienced. I have read some patents in Geotech website you mentioned about discrimination before. As I know, the signal from RX coil will be demodulated into two component, which are name as R(resistance) and X(reactance), Then the ratio between R and X determine the phase shift. and the discrimination is due to this phase shift. right? I got a Bounty Hunter Platium (maybe the other name is Teknetics gamma?), I disassembled it and use it to collect some raw data of metal outside. This raw data measurement test point is after the pre-Amp, band-filter, synchronized demodulation. maybe the envelop of RX sinusoid, and I think they are just two component R and X. Please see my attachment files.

Although you can see from your data that different targets can be identified, the best place to take your VDI samples is at the outputs of the R and X channels. There is an add-on VDI circuit somewhere in the Tech Forums that does exactly that. The processor is used to find the arctangent of R/X, allowing you to directly display the phase-shift produced by the different targets. For a more stable display, take several samples and find the average before doing the trig calculation.