PEV Servo Data
Build a Geomagnetic Observatory ! GELLER Labs "Backyard Science"
Thoughts on a proton precession magnetometer design - a Proton Magnetometer Project. Build an Earth's field magnetometer.
The FDM MAGNETOMETER1 project is a low cost high performance proton magnetometer (a digital magnetometer) kit under development for universities and amateur scientists to be able to accurately measure and monitor changes in the Earth's total magnetic F field and to observe geomagnetic storms. Magnetic storms can cause large excursions in the field and are of concern to interests ranging from electrical power grids, radio communications, and satellite operations, to aurora watchers and amateur radio operators.
1 Filter Diagonalization Method "FDM" (harmonic inversion), see Jan 21 and Jan 23 entries, based on: Vladimir A. Mandelshtam, Howard S. Taylor, Harmonic inversion of time signals and its applications, Journal of Chemical Physics (1997), Volume 107, Issue 17, 1997, Pages 6756-6769
(be sure to hit refresh to pick up our latest changes and entries)
Project Documentation, Links and References (very early stages)
Journal Notes:
Tuesday, March 22, 2011
Overnight: PDF, TXT, Peak Envelope Voltage (PEV) servo and Tau 2 plot PDF. The temperature was relatively constant over night (~35 F). It is interesting to see how the PEV and Tau 2 curves are remaining relatively constant over a long period of nearly constant temperature on a so far, cloudy day (little direct solar illumination of the covered sensor coils) PDF.
Keep in mind that the PEV servo work and added Tau 2 parameter are all just fine tuning around the edges. The goal of the project is accurate and high resolution monitoring of the geomagnetic field. Speaking of which, there was an interesting small negative triangular shaped pulse just after 10 am (1400 UTC) this morning PDF.
I extended the digitization interval from 1.1s to 1.6s (16,000 samples about 10 kHz). This was done to accomodate longer Tau 2 times. I realized during some simple fluid testing that the Tau 2 measurements were hitting a "rail" with longer times that extended the 1/e amplitude point out beyond the shorter 1.1s digitization interval. I think the longer sample will give equal or better quality results. If not, the alternative is to measure the Tau 2 time indirectly by measuring fall time to a higher amplitude than 1/e (63.3% fall) and then inferring the fall to the 36.7% amplitude point. It might be advantagous to change the default Exp FV value with the longer record, this needs further study.
Wednesday, March 23, 2011
Overnight: PDF, TXT, PEV Servo PDF. The geomagnetic field is somewhat unsettled (well below storm levels). The PEV servo plot, including the Tau2 values (green curve) shows the Tau 2 (time constant of the fall of the precession signal) dependence on temperature. We are not recording NMR sensor fluid temperature, however outside air temperature varied from about 35 F to a low of about 26 F overnight. The servo'd polarization time (red curve) is not a direct measurement Tau 1 (the NMR spin alignment time constant), however it can be seen there is some similarity in the shape of the two curves. It may be possible later to compute a Tau 1 value from the polarization time.
Afternoon: The geomagnetic field has remained unsettled all day PDF.
Thursday, March 24, 2011
Overnight: PDF, TXT, PEV Servo PDF. The high temperature yesterday was about 33 F, the low overnight about 16 F. It does appear that the servo'd polarization time (red curve) might be related to Tau 1 by a constant, at least at any given temperature. The Tau 2 value represents a direct measurement from the filtered precession waveform envelope. The Tau 2 curve includes noise caused by ripples (noise, EMI/RFI, interfering fields, etc) on the digitized precession signal. The system is so robust, it is difficult to remember at times that the original 1.6 second digitization of the decaying exponential free induction decay (FID) signal (the precession signal) on which all these measurements are made is under ten microVolts! As expected, both Tau 1 and Tau 2 are strong functions of the temperature of the working NMR fluid (outdoor air temprature plus direct solar heating).
Carl Olsen's (Colorado, US) PEV servo plot takes us up to a sun illuminated container temperater of 76 F (air temp 54 F). Carl reports that the overnight low was 25F compared to 31F the night before. PDF TXT (The Tau 2 spike was caused by a temperature probe momentarily placed in the sensor container.)
We made a number of small changes to the LabView supervisory program today. PDF On the housekeeping side, the total number of measurements now updates at the beginning of each measurment cycle. As soon as a measurement survives auto-retry, the plotted measurement value is updated. There is a bit of a conflict between the digitization needs for FDM (the frequency estimator) versus the precession waveform filtered envelope and Tau 2 (related to fall time) calculations. FDM gives most accurate results with about 1.1 seconds worth of precession waveform data. The signal to noise ratio is best early on in the free induction decay (FID) signal (the precession signal). Also, there is less likelihood of frequency spread caused by EMI/RFI and/or moving ferrous objects with shorter time windows used for FDM. On the other hand, a longer window gives a better diagnostic look at FID decay curve. Also, at warmer temperatures with liquids having longer Tau 2 times, it would be desirable to have a window as long as 2.5 seconds. So, in another change, we now digitize for a relatively long time (e.g. 2.5 seconds) and then extract the first 1.1 seconds worth of data for the FDM executable. The precession waveform envelope waveform has been delayed to begin with the peak value (excluding the analog rise to peak associated with transferring over to the narrow band low noise amplifier each measurement cycle). Regarding the Tau 2 measurement, we added an exponential fit (thin black line) to the graph of the filtered precession waveform envelope. The Tau 2 measurement is taken from the fitted line as opposed the actual filtered envelope. Since the filtered envelope waveform typically has some ripple by the time it has decayed to 1/e (~63% decay to the 37% amplitude value) there should be less noise in the Tau 2 result using the fitted curve. Finally, the PEV Servo is now servoing on the peak value of the fitted curve.
Friday, March 25, 2011
Overnight: PDF, TXT, PEV Servo PDF.
Saturday, March 26, 2011
Overnight: PDF, TXT. The geomagnetic field was very quiet overnight. The outdoor air temperature varied from a high yesterday of about 33 F to a low this morning of about 17 F, PEV Servo PDF. We servo the polarization time to control the amplitude of the free induction decay (FID) signal. The feedback signal for the servo is based an exponential fit of data from the filtered envelope computation. The filtered envelope is calculated directly from the FID signal. A view of the FDM peak frequency spectral line (our Larmor line) shows a slight deviation of FDM amplitude compared to the servo'd peak envelope amplitude (presently 2 V) PDF. The FDM figure of merit (FOM) plot looks good with the statistical mode or most common FOM value this run (700 points on a 2 minute measurement cycle) at 5e-7 PDF.
Sunday, March 27, 2011
Overnight: PDF, TXT, PEV Servo PDF, FDM Amplitude PDF, FDM FOM PDF. Beta tester Carl Olsen and I have been discussing the addition of a temperature sensor near the fluid sample. Carl suggested the LM34/35 family of temperature sensors as having no detectable ferrous content. I tried a LM34 since that was the one that I had on hand. I used the bottom of page 7 PDF Two-Wire Remote Temperature Sensor (Ground Sensor) configuration JPG1, JPG2. Note that in the diagram, the arrows to the straight lines suggest where the length of the twisted pair goes to the remote location (they should have used the twisted wire symbol) and the output is represented by the + and - symbols across the 499 ohm resistor. I used a small 6 V lead acid batter (5 AmpHours) to power the sensor (via a series diode). I destroyed the first one by accidently reverse connecting the power. I used a lightly shielded cable (loose weave braid) with the shield tied only at one side to the instrument singl point "star" common Earth ground. The circuit draws less than 2 mA, so it should run indefinitely on that battery, and for now I am concerned about creating a noise path (and/or ground loop) to the analog inputs on the USB 6008. I used another 6008 differential input across the sense resistor of figure 3 to read the temperature. The 6008 takes the average of ten samples at 100 Hz at the beginning of each measurement cycle. Slope and offset values were included as simple multiplication and addition operations in the program. (rought calibration was done with an ice bath and at room temperature). As can be seen in the graph, with the sensor just inside the coil form in air, there might be some lag between the sensor temperature and the fluid temperature. Also, the stand passes in and out of shade from parts of a tree, so that might explain some dips in the temperature. The circuit is not ideal, since it is rated from +3 F to 100 F, and the LM34D that I had on hand (optimized for 32 to 212 F) is also less than optimal. However, it is good to get a first look at temperature directly measured in the sensor container near the NMR fluid sample. Despite the apparent lag between temperature sensor tempurature and fluid temperature, it should be possible to obtain some meaningful temperature to Tau 2 and polarization time (related to Tau 1) data using periods of relatively steady temperature (where the fluid temp catches up with the sensor temp) PDF. Not surprisingly, the temperature in the transparent coil container, in the coil form, under a heavy plastic bag cover, is some 20 to 25 F above the ambient air temperature.
Afternoon: Good progress. By moving the temperature data in time (delay), I was able to determine that the thermal delay between the sensor air temperature and the fluid temperature is about tens of points for my current temperature sensor (LM34) position. Once the temperature points were aligned properly in time, I was able to realize a curve fit for both the polarization time (for a 2 V peak envelope voltage) and Tau 2 versus temperature. It turns out to be a very simple equation to first order: y=c1*e^(c2*temp). Here are the curve fits (for Prestone DeIcer) with a vertical axis log scale in time PDF and with a linear vertical axis in time PDF.
This raises another possibility for a simple FID amplitude controller. Since the Tau 2 values have been measured with relatively less scatter than the peak envelope value, and we can know the exponential equations for any given fluid, Perhaps, the way to go is to read a few measured Tau 2 values into a moving average and then calculate the predicted appropriate polarization time from the Tau 2 value! Note that this is not a servo, but rather an open loop "feed foward" calculation based on a Tau 2 measurement, to set a polarization time for any desired peak envelope voltage.
Early evening: PEV Servo plot PDF Late evening: Running overnight on a Tau 2 feed forward open loop scheme to dynamically set the polarization time each measurement cycle. Also, added a display of fluid temperature based on the Tau 2 measurement PDF. Here is the beginning of a new performance plot. The light blue curve is the new fluid temperature variable PDF.
Note to new readers: The main purpose of this equipment is to measure the Earth's geomagnetic field. If you can solder together a printed circuit board and wind coils on PVC forms you build, and can build a simple wood sensor stand, that is all you need to build the amateur geomagnetic observatory. The rest of a lot of the discussion of the past few days is "deep in the guts of the instrument" design work that might not interest many users. However, for others, this system can be used as a proton magnetometer development system, including the ability to rigorously characterize the NMR properties a working NMR fluid.
Project Documentation, Links and References (very early stages)
QUESTIONS/COMMENTS/notice of typos, etc. send email to joegeller @ gellerlabs dot com
COPYRIGHT © 2009, 2010, 2011 JOSEPH M. GELLER, All rights reserved.