Earth magnetic field EF

[Tinkerer]

The EF is quite different in different regions of the world. I have run Magnetometers in EF field strengths of 48nT to 20nT. (nano Tesla). But then there is also the difference in the field vectors.Near the poles, the Earth's magnetic field lines run near perpendicular to the surface. Near the Equator, the lines run near parallel to the surface.
Locally, the field lines can be influenced very much by permeable ore deposits and galvanic or telluric currents.
The effect of swinging the coil is therefore quite different.
How much difference do we notice?
How can we compensate for it?

[Davor]

Be more specific. In metal detectors we do not need to separate EF from other sources of offset.

[Tinkerer]

Be more specific. In metal detectors we do not need to separate EF from other sources of offset.

Moving a coil within a magnetic field, generates a current in the coil.

With a traditional PI detector, we use a late sample to measure this current and then subtract this signal from the main sample.
If the target has a long TC, there may still be a significant target signal present at the time of the late sample. This part of the target signal is also subtracted and therefore makes detection of long TC targets more difficult.

What are the characteristics of the EF signal?
What affects these characteristics?

- Only with the coil moving we have an EF response signal.
- A fast movement generates a greater response than a slow movement. With the usual coil sweep, the movement is constantly changing from near stand still to maximum speed, in fact generating a sine wave.
- The amplitude of the EF signal is affected by the EF density.
- The amplitude of the EF signal is also affected by the direction of the movement of the coil, relative to the surrounding field lines.
- A movement of the coil, parallel to the EF field lines generates little amplitude.
- A movement of the coil, perpendicular to the EF field lines generates higher amplitude.
- A fast tilt of the coil of the coil can generate quite a high amplitude signal, if the coil is lined up with the EF field lines in the right direction.

[Davor]

When you analyse 1/f noise and drift over time in opamps and other active components, you end up with true EF effects bordering with 1/f semiconductor noise. That said, EF pulse will equally eliminate EF and opamp drift. Therefore you have two unrelated effects competing for your attention, and a solution that eliminates both.

I'd say with nowadays MCUs and accelerometers it is easy to separate motion related induction from opamp drift, but so far I fail to see the point of such exercise.

[crane]

- Only with the coil moving we have an EF response signal.

We also obtain an EF response signal when the coil is stationary and something disturbs the EF, such as when a truck goes past or when swinging the pick, or a stationary metal detector mounted on a conveyor belt where mobile machinery disturbs the Earth's field. This signal occurs even when the truck or pick is at a distance where it normally can't be detected. Permanently magnetised rocks and rock outcrops also give a response and note that their fields aren't always aligned with the EF. All of the above are low frequency signals that can be cancelled with normal methods. You will also get a response if you touch two suitable hot rocks together near the RX coil (preferably in series with the EF). You can test this by using two pieces of soft ferrite instead of the rocks. The impulse in this case must occur during a sampling period otherwise it will be ignored and it is high frequency so isn't cancelled by the usual methods.