[TangerineSDR] Should have been: Updated RM3100 reslts from Julius Madey

Thu Jul 30 08:56:02 EDT 2020

Hi Jules and Dave,

Thank you for working on this. The temperature issue seems to be a very plausible explanation for the problem, and I could see how this is in measurement regime that maybe most people using these sensors don’t worry about. I’m interested to see PNI’s comments, and if there is any way we can compensate for the errors.

Talk to you soon and 73,
Nathaniel

From: David Witten <wittend at wwrinc.com>
Sent: Wednesday, July 29, 2020 7:51 PM
To: Hyomin Kim <hmkim at njit.edu>; Dave Larsen <kv0s.dave at gmail.com>; David McGaw <david.g.mcgaw at dartmouth.edu>; Tom McDermott <tom.n5eg at gmail.com>; Dr. Nathaniel A. Frissell Ph.D. <nathaniel.frissell at scranton.edu>; Noh, SungJun <sjnoh at njit.edu>; Gil Jeffer <gil.jeffer at gmail.com>; John Gibbons <jcg66 at case.edu>; Scotty Cowling <scotty at tonks.com>; tangerinesdr at lists.tapr.org
Subject: Should have been: Updated RM3100 reslts from Julius Madey

Dave.
Here's the first couple of pieces on temp sensitivity.  Considering the full range of the RM3100, the observed effect could be considered relatively small. However, it's in the region we need to observe.  Although the basic idea of taking the differences in duty cycle of the LR oscillator with forward and reverse bias eliminates some major temperature contributions, it apparently does not totally eliminate them.

One is of a GTJ run on 07072020 and the other a one hour calibration run I did yesterday 07272020.  Note the reversed sense of the temperature change effect.  I have a second RM3100 which I will also run in the next few days for comparison.  I used your original adapter board.  The setup in the wood lot at the end of 400 feet of CAT5E I believe uses the same adapter board you supplied to Hyomin.  Reducing the 24 hour data in Open Office Calc took a god-awful time to complete.

Feel free to distribute these ...  I'm going to send them to PNI for their comment.  I did not attempt to measure the drift, if any, of the 3.3 volt regulator output on the SparkFun i2c differential extender at the test end.  I also have data which was taken with a delay between the start command and looking at the Status register to keep the i2c bus quiet during measurement and the trends are the same.  I don't think PC trace currents created that much local field but that's another check for the future.

Jules

Julius Madey
Tue, Jul 28, 11:15 AM (21 hours ago)

to me

Dave,
I have no problem with sharing that my tests so far have shown that, at least on the y axis which was tested,  the RM3100 is well within 5% accuracy with Helmholtz coil excitation with a long period (30 minutes to 1 hour) sine wave of +/-50, +/- 524 and +/- 2893 nT.   And I would be happy to provide graphs to the repository.

My problem at this point is incomplete data on another aspect, which is why I haven't released the stuff I mentioned yesterday.

PNI claims that the basic measurement technique employed in the RM3100 eliminates temperature effects.  See the attached note.

However, I am clearly seeing what appear to be temperature related effects in the measurements, at least at lower levels like 30 to 150nT variations of the sort seen at the Jenny Jump tests.  If you know the tempco, it's possible to compensate in the data reduction ..... however......the disconcerting thing is that I'm seeing different trends vs temp in the GTJ results and the runs I did in the last couple of days.  Tempco sign differences, which suggests magnetometers may have to be calibrated before deployment ...  temperature variations of 15 C or more

If I look at the temperature related effects of the GTJ measurements (tedious work due to the large data set) it appears that the significant deviation which Hyomin noted between the GTJ and JJ data is entirely due to temperature variations over the 24 hour runs  !!

I have an inquiry in to PMI.

Jules
[https://ssl.gstatic.com/ui/v1/icons/mail/images/cleardot.gif]
Attachments area

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/tangerinesdr_lists.tapr.org/attachments/20200730/6f58f80a/attachment-0001.html>