Decoding Mag Signals

[Mark Thorley]

This one is possibly a little out there, but has anyone ever considered looking at the detail in the waveform in the sine wave signal.

In the presence of different ferrous materials I wonder if the resulting signal has any usable information ?

Mark

[Willy Bayot]

Re: Decoding Mag Signals

The exponentially decaying sine wave signal generated by the proton precession in a PPM is JUST a pure sine wave whose frequency (only its frequency, not its shape nor even its ampitude) varies according to the total magnetic field sensed by the sensor of the mag. This is completely independent from the origin of the field variation, whether diurnal or micro variations due to solar influence or magnetic storm. Magnetized material/stone/ground or any ferro-magnetic object can also disturbe the natural earth field as sensed by the system but the SHAPE of the signal is still a pure sine wave (with more or less background noise).
The mag does not generate any significant eddy current in the ground and in its contained objects as a PI detector does, thus, there is nothing interesting to find there.

Sorry to deceive you,

Willy

[Mark Thorley]

Re: Decoding Mag Signals

Thank you Willy for a succinct and VERY interesting response to my question. Such a wonderful answer leads me to the obvious next question. If I wanted a system that DID have some intelligence on the resulting signal that is proportional to a target, what would be a good choice? As you correctly say, a PI detector does induce sufficient eddy current to affect the objects in its field.

My intention is to experiment with the resulting signals and attempt to implement a neural net to discriminate between different target types. The system I envision is only for discrimination, not for discovery. Traditional/commercial units are VERY good at finding a ferrous target, but I want to play with the concept of defining the target.

Mark

[Willy Bayot]

Re: Decoding Mag Signals

Hi Mark,

I am afraid that the best you can achieve in this direction is to discriminate between metallic targets through their electrical conductivity only. In certain ways, you could grossly distinguish between round and long target shapes but not distinguish between a large, deep target and a shallow, small target made of the same metal.
VLF systems have the best, most accurate discrimination capabilities but it is also possible to achieve discrimination with PI systems, we have proven that fact by analyzing the shape of the pulse decay (see my posts on the metal detector forum fo Carl).
Do you REALLY want to search for ferrous targets? Usually, people want to discriminate them OUT.

Willy