Journal notes, Some Run Time with the New prototype
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)
Thursday, April 22, 2010:
Ran on a different Windows 7 machine overnight. I forgot to set the power settings so the USB port does not shut down automatically for power savings. I think that is why the run terminated.
Overnight run: PDF, USGS PDF, restarted PDF.
There were a couple of interesting peaks in the scalar F field this afternoon followed by an oscillation around 6pm est: PDF, TXT, USGS PDF. Interesting magnetic evening continues: PDF.
Friday, April 23, 2010:
Good run overnight: PDF, TXT, USGS PDF, sample spectra: PDF, PDF. Interesting to contrast the minor G1 storm (NOAA Space Weather Scales) to the relatively quiet period and beginning of the normal diurnal cycle this morning. The new prototype (still on the Win7 desktop) continues to run well: PDF.
Saturday, April 24, 2010:
Another good run overnight: PDF, TXT, USGS PDF, 48 hour view PDF.
Sunday, April 25, 2010:
Overnight run PDF, TXT, USGS PDF, view over 3 days PDF. The rectangular section in the afternoon is from the car being out of the garage.
I realize now that this experiment is too advanced for most schools, perhaps with the exception of some very advanced high school science students or some science high schools. On the other hand it should be a very interesting project for undergraduate engineering and physics students, perhaps even for some graduate work with an in depth study of the FDM method. I also think that most experienced amateur scientists and electronics hobbyists will be able to successfully build the experiment.
I think that the "jam jar magnetometer" remains the experiment of choice for K-12 geomagnetic demonstrations and science fairs. Here a good NASA reference page on the soda bottle magnetometer: NASA GSFC (be sure to click through the links there).
Monday, April 26, 2010:
Overnight run: PDF, TXT, USGS PDF. It was a relatively quiet magnetic day with a nice showing of what appears to be a typical local diurnal cycle. The morning record shows a significant amount of vehicle traffic, including the two monday morning garbage trucks.
All seems well with the latest prototype. Short term, I hope to look at a simple interlock that trips if the coil is powered for more than a desired time period, and at replacing the batteries with a power supply (on order). The interlock is needed since the #24 coil wire is not rated for continuous operation in the 1 to 3 ampere range (because the tightly wound overlapping layers do not cool well). So far, I have only seen this failure is when I terminated the program when the polarize suppy happened to be on. So, the main computer and/or USB 6008 cannot perform the interlock function, since the failure mode is when one or both of them is stopped. I am thinking a current source powered by the polarize gate resistor voltage (from a high side driver, see schematic from April 17, 2010) charging a film capacitor. In a production model, a comparator would change state for a capacitor voltage above a pre-determined level. For the prototype, I will use a meter relay to observe the performance. In either case, the interlock will need to open a 12 V contact on a latching or step relay.
5/11/10 update: Alpha Wire Company raised the prices and this cable at distributers is now >$200 for 100', double the price in late April, so it is no longer a practical option for an amateur observatory. Very dissapointing.
I would like to try out the new Alpha XTRA GUARD continuous flex data cable for the run to the two counter wound sensor coils. I was looking at a sister cable for another project and realized one of the samples they sent might be perfect for this observatory application. The shielding is superb, not to mention weather resistance and flex ability. # 86702CY SupraShield (Premium Foil/Braid) (sign up for an account to get the PDF, no affiliation). A more cost effective solution is possibly their 86304CY xtreme environments cable. I will look into other manufactuers as well.
I wonder if it would reduce more of the background noise pickup? (at some minimal level the matching and positioning of the two counter-wound coils probably dominates the noise floor).
Tuesday, April 27, 2010:
Overnight run: PDF, TXT, USGS PDF. Step changes are from vehicles.
afternoon: The focus is so much on electronics development that there have been few extended data runs to date. With other projects active this week and waiting for parts deliveries, I have been letting the FDM magnetometer run for a few continuous days again. The diurnal cycles are so interesting and probably packed with lots of good geophysical information.
View over a few days of the diurnal cycle: This record begins with a relatively minor G1 storm and then the mean field has fallen off by on the order of 10 nT from 4/22 to today: PDF.
This reminds me that I need to quantify the errors and variation in time and temperature of the USB 6008 digitization rate, the only variable that could cause a frequency or field error. My impression is that the digitizer sample rate error is probably at least an order of magnitude below the drift in the actual field observed in the pdf, but another thing for the to do list.
Wednesday, April 28, 2010:
Well, the latest run ended when Microsoft updated Windows 7 overnight and rebooted the machine. I suppose a working observatory program could restart on reboot by itself, but the prototype is no where near that stage yet.
I have been meaning to stop anyway, at least to log some of the latest operating paratemeters: The present timing sequence is wait for the small signal relay to close (200 ms); polarize MOSFET (1.8 s) at 1.2 Amperes, wait for coil discharge (energy dump) 10 ms; pause before observing precession signal (100 ms) (stabilize front end amplifier bias currents and establish quiescent DC values, such as across the front end resonating capacitor). Continuing with the previously described counter-wound coils where only one coil has the EFNMR fluid (presently Prestone De-Icer windshield washer fluid) and only that coil is polarized, however to read the precession signal, both coils are used in a center tapped configuration for ambient noise rejection (very important).
Waiting for the new (used) hp 6237B bench supply to replace the batteries and probably the Kepco polarize supply too. Now that the system common is at all three power supplies, a multi-output power supply should drop right in with no modifications needed. I expect the hp 6237B to be very convenient for now with its output ammeter to monitor all three rails (+6V analog, -6V analog and 12V polarize). The hp manual says it is okay to operate in the "grey area" up to 1.2A, so it should be fine for polarize as well. As presently configured, the +/- rails can be anywhere from about 6V to 15V. I used the sister hp 6236B for mixed signal circuits (+15/-15/+5) for some years, it was a nice unit to work with.
Later, an experimenter might opt for a more economical open frame supply, or a home brew linear supply, or even batteries with a sufficiently high A-H rating (e.g. used UPS batteries).
Also, I am still thinking about the safety interlock, so incase I stop the program during the polarize time and forget to turn off the polarize power supply, the coils don't run at DC.
Friday, April 30, 2010:
Overnight run: PDF, TXT, USGS PDF, sample spectra PDF, a slightly unusual overnight pattern.
Saturday, May 1, 2010:
Overnight run: PDF, TXT, USGS PDF. Towards the end of this record I had a chance to inject a GPS stabilized signal at 2290.450 Hz in place of the precession signal. This is the first calibration check done in some months. I was pleasantly surprised when, for various input amplitudes, the result was always 2290.450 Hz (a function of the USB 6008 clock stability), especially since the lab temperature is probably about 8 degrees F higher than past weeks today (we enjoyed an unusually warm spring day today). There are a number of failed measurement cycles this afternoon. I think it is a combination of car traffic (a nearby yard sale), outdoor machinery (weed wackers, etc.) and the warmer copper leads and coils (under a heavy black plastic cover) lowering the polarization current to below 1.2 A (The polarization supply is presently running in voltage mode). Readings seem better now that I turned the voltage back up to give about a 1.25 A polarization current. I wonder if there is also any temperature dependence of the sample fluid? Probably a secondary affect if at all, one more thing to look into. Perhaps I should change to a heavy white plastic cover for summer.
When I was reloading the XP computer that failed in earlier testing, I noticed there were a number of failed measurements (still over on the Windows 7 machine). I had moved the computer nearer to an ethernet connection, but left the monitor plugged into the same strip as where the polarizing power supply is plugged in. It makes sense now that there was far more ground noise since there was a ground loop closed by the monitor cable and the computer and monitor grounds. This is probably what I was seeing when I adjusted the voltage higher earlier today, not a problem with polarization current. I changed the polarization power supply to a non-grounded plug using a "cheater" adapter and ran a #12 ground wire from the power supply chassis ground and the FDM magnetometer central ground/common down to near the home ground, not far from where the cold water pipe enters the home (not far from the place where the home AC distribution box ground is as well. The FOM values are lower now. This grounding business needs still further study! At least the magnetometer ground is still one common ground at the analog and digital systems, so it should still be easy to use a multiple output bench supply.
After all that, it seems the old XP machine has a problem, probably a bad power supply. The measure waveforms on the Win7 machine became noticably quieter as soon as the old XP machine was shut off.
Project Documentation (very early stages)
QUESTIONS/COMMENTS/notice of typos, etc. send email to joegeller @ gellerlabs dot com
COPYRIGHT © 2009, 2010 JOSEPH M. GELLER, All rights reserved.