Wed, 01/20/2010

A Magnetometer is a scientific instrument used to measure the strength and/or direction of the magnetic field in the vicinity of the instrument. Magnetism varies from place to place and differences in Earth's magnetic field (the magnetosphere) can be caused by the differing nature of rocks and the interaction between charged particles from the Sun and the magnetosphere of a planet.

Magnetometers are a frequent component instrument on spacecraft that explore planets.

John purchased two high quality magnetometers (model FGM-3H from Speake & Co, Llanfapley, England). The units are very temperature sensitive (~ 100nT/degC) so, for temperature stability, one unit is now installed underground at the Wurillba Observatory site.

Blair has measured the sensitivity of the magnetometers as better than 1 nano-Tesla (nT), and has completed calibration. Testing of the custom interface electronics (based on a PIC microcontroller) is completed and is producing data comparable to professional monitoring stations anywhere in the world.

Sample of data captured on the 16th March 2010 (green) compared to the previous day 15th March 2010 (yellow)

The interface computes the period of the magnetometer (frequency) output using the onboard 4MHz clock crystal as a timing reference. Using a 24-bit counter the unit will be capable of measuring magnetic field strengths ranging from 0 to 100,000 nT with an output resolution of 0.1nT.


  • Our FGM-3H magnetometers

  • Current Status. Feb 2018
  • 2 units were manufactured in 2010, one unit was installed at our Wurillba site in early 2010, the other unit being used for software development, testing and public demonstrations.
  • The magnetometer sensor was buried at around 1 metre at the rear of the property. We now have  8 years of excellent data at a sample rate of 1 per second.
  • For the Technical minded a short discussion on the interface electronics. Sensitivity to temperature is quiet high, figures quoted are from 40 to 100nT per degree C. Both our units have a temp coeffient of around -50nT/degree C. An LM35 analogue temperature sensor is attached to the FGM-3H magnetometer sensor and temperature variations are able to be compensated for in the software. The units are usually buried under the ground for temperature stability and we did this at our remote site.
  • Double power supply regulation was implimented for the sensor as suggested by the manufacturer.
  • In simple terms, The FGM-3H sensors are of the fluxgate design and output a 'square' wave with a frequency in the range of 40KHz to 140KHz and a sensitivity of approx 1 Hz per nT (nanoTesla) at a frequency around 62KHz. The FGM-3H sensors are more linear in their response by measuring the period of the output waveform than by measuring the frequency of the output waveform so the electronics has been designed to measure the period with high accuracy.
  • The output waveform is divided by 2 to the 14th (16384) to give an output frequency with a range of approx 2.4Hz to 8.5Hz. This is used to gate a 4MHz clock producing groups of 4MHz clock pulses. Each group of clock pulses (approx 800,000) are counted by a 24 bit counter. At the end of each gate period, the number of clock pulses captured by the counter are read by a 16F877 PIC micro. The counter is then reset ready for the next gate period. With this arrangement, we end up with approx 13 counts per nT, and a very large full scale range (approx +_ 30,000nT).
  • The count value is in 3 bytes and at the same time, the signal from the LM35 temperature sensor is measured as a single byte. At an appropiate time interval (user adjustable from 1 to 240 seconds), these 4 bytes are passed onto a computer (via RS232) and are displayed in graphical form. The data is logged to a file for further analysis as required. At this stage, we are logging data every second.
  • The effects of CME's on Earth's magnetic field have been regularly recorded.Future plans. A tri-axial version and a proton precession magnetometer. There is some interest in sampling at 10 samples per second at the same sensitivity, however this will require some changes to hardware.
  • We now have 8 years of magnetometer data recorded at 1 sample per second. Just over 250 million samples!