GELLER Labs "Backyard Science"
Thoughts on a proton precession magnetometer design - a Proton Magnetometer Project
The goal of this project is a low cost high performance proton magnetometer (a digital magnetometer) kit for amateur scientists to be able to accurately measure and monitor changes in the Earth's total magnetic F field and to observe geomagnetic storms. There is a regular daily (diurnal) variation in the Earth's magnetic field. During events related to solar activity, there can be sudden changes in the field (such as a sudden impulse) as well as large excursions in the field which can be more than ten times the regular diurnal variation caused by magnetic storms.
(be sure to hit refresh to pick up our latest changes and entries)
Friday, January 1, 2010
Last week I began to investigate a failure of the various frequency determination routines at the 10 milliHz to 1 milliHz level of resolution. Close inspection of some of the field data was starting to show bistable dead zones at 1 nTesla resolution. The symptom was data riding along at one level then falling throught the "dead zone" and continuing below or above it (depending on whether the field was falling into the dead zone or rising into it). One perplexing aspect of the dead zone is that sometimes it can be repeated with a sine wave swept through it using a synthesizer and sometimes not. For example, in some set ups, the Buneman frequency estimator runs nearly perfectly with a sine wave, but then shows dead zone with the PPM expontionally decaying filtered sine wave. In one experiment, I wrote a vi to roughly apply an inverse exponential window to normalize the PPM waveform. There was still a dead zone.
I am going to next try an adaptive filter algorithm where following each initial frequency determination, a second FIR filter is tuned (set with calculated parameters) which then filters the very same data set and applies it to the Buneman frequency estimator (shot to shot). It will be interesting to see if an adaptive filtering scheme solves the dead zone problem. Here are some rough notes on some of the dead zone testing.
Monday, January 4, 2010
Still investigating the Buneman dead zone problem. Buneman tracks the signal generator to about +/- .002 Hz, so I hate to discard it without some effort into understanding the dead zone gaps. I did notice today that my zeal in setting so many taps in a FIR filter might not be necessary. Running overnight tonight on the USB-6008 with only four taps on the bandpass filter. The data appears to have a reasonable spread shot to shot of well under 1 nT. The number of samples and sample rate clearly affect the location of the dead zone. It is not clear yet if can be eliminated by optimizing those parameters along with the digital filter. Promising is the fact that there are setups with no dead zones (at milliHz resolution) when using a sine wave from the synthesizer.
I probably need to build a Helmholtz coil around the sensor for more efficient sweeping of the field. But, the front end electronics is now under about a foot of snow, so for now moving the car gives me a defined sweep on the order of 20 nT.
Also, making some progress on direct monitoring with a triggered measurement using the hp 53310A Modulation Domain analyzer. Even with an additional high order Krohn Hite analog filter in the lab and a reliable cycle trigger, hardware measurments to 0.1 Hz are surprisingly difficult compared with the DSP side. The old Canadian observatory unit I looked at from Presentry used a phased locked loop multiplier with a scaled counter clock. I see now how important the averaging (filtering) of the PLL was over the decaying expontial sine wave. In some ways, it is remarkable that low field NMR has been done for so many years before DSP tools became commonly available.
QUESTIONS/COMMENTS/notice of typos, etc. send email to joegeller @ gellerlabs dot com
COPYRIGHT © 2009 JOSEPH M. GELLER, All rights reserved.