The Tekatch-Welch Magnetometer

Design Considerations

In building this magnetometer, we had several design goals:

E-W component of geomagnetic field with noise of +/- 1 nT
Sampling at frequencies lower than or equal to 1 Hz
First-order removal of temperature sensitivity
Precision temperature monitoring at sensor head
Digital transmission of sensor data
RS-232 serial interface
Use of low-cost components whenever possible
Orientation mode (for quick alignment to magnetic E-W)
Standalone mode (to allow computer-free operation such as alarm)

Schematic

Component Links

Code for PIC16F876

2003 Dec 31 version

Tcl/Tk Script for Viewing Data

2004 Jan 4 version

Image Gallery

Note that the pictured circuit uses a MAX233 chip!

Results January 6, 2004 - Things are working much better now! The magnetometer is currently sitting in a closed closet on the top shelf and the temperature changes are much reduced. Our plot below shows the difference signal between two FGM-3h's (red) and the temperature difference between the two LM92's (blue). A clean geomagnetic signal is seen! The approximate conversion between difference counts and nT is 6.7 difference counts per nT for two oppositely oriented FGM-3h's counted over an interval of 2*0.983 seconds. The Ottawa traces for the same interval are shown below ours. (Ottawa is 500 km northeast from our location.)