PC based Proton Precession Magnetometer?

[Ezra Teneflin]

Re: PC based Proton Precession Magnetometer?

>>mHz accuracy

Is it an asynchronous measurement, or are you triggering or phase locking to the signal?

[Willy Bayot]

Re: PC based Proton Precession Magnetometer?

It's an asynchronous measurement, no triggering nor phase locking. It would not be practical to do otherwise with the precession signal. If we could have synchronized the calculation on the signal, the things would obviously have been simpler.
We just increased the accuracy of the standard FFT calculations by proper zero filling, windowing and a specially designed interpolation algorithm.

Willy

[Ezra Teneflin]

Re: PC based Proton Precession Magnetometer?

Here is an interesting PC based model that I saw advertised in 'Physics Today'. (I have no affiliation with this company.)

http://www.exstrom.com/magnum.html

They also use a modified FFT routine:

http://www.elecdesign.com/Articles/Index.cfm?ArticleID=5575

But, for some reason, they chose to use a separate digitizing card rather than the sound card.



Click Here

[Willy Bayot]

Re: PC based Proton Precession Magnetometer?

Good remark!!
This is indeed an interesting site and laboratory instrument. Actually, the first PCM and FFT images that I posted on this forum a few weeks ago were generated from this very instrument.

Their HRFT algorithm gives a slight progress in accuracy from the pure FFT calculation. We already tried it at the beginning of our project but it still did not give enough accuracy for our specific purpose.

We shall have to see if an average PC sound card gives good results before going to make one from scratch with an audio amplifier and 16-bit ADC but I doubt that we could do better than an industrial-grade PC card.

Willy

[Ezra Teneflin]

Re: PC based Proton Precession Magnetometer?

I've been using a single ended LT1028 amplifier (Linear Technology) for the first amp, with relatively low value metal film resistors with some success. The noise level is high here, so I still have to reduce nearby (periodic) spurs with FFT noise reduction techniques just to 'see' the signal (using an analog Q of just below 150). I appear to be getting only about 50 nV from my small coil and sample, or my math is still off ... Moving the car slightly at over 100' changes the frequency, and the frequency matches the USGS model for my area, so it appears 'good', albeit the circuit is still far from being a production model.

Popular instrumentation amplifiers appear to exhibit too much noise to be used directly in the front end - at 4 nV / root Hz and higher.

What are you using for the front end amp?



Click Here

[Ezra Teneflin]

Re: PC based Proton Precession Magnetometer?

>>"Moving the car slightly at over 100' changes the frequency"

I should have said moving the car in increments out to over 100' gave nice shifts in frequency, but I could only resolve to around .1 Hz (post processing using the math packages on a high end LeCroy O'Scope). If you can really get down to milli Hz, that would indeed make a nice instrument!

Ezra

[Willy Bayot]

Re: PC based Proton Precession Magnetometer?

50nV is indeed a VERY SMALL signal. Can't you make a better coil with a higher SNR (more turns, more liquid)?
For the front-end pre-amp, we are just using a standard transistor stage.

Willy

[Willy Bayot]

Re: PC based Proton Precession Magnetometer?

.1 Hz is still not bad since it would give you about 2.5nT of resolution. That's at least ten times the resolution you can get from an average differential mag. I think you should work a bit more at the quality level of the coil in order to get a better initial signal.
You could use the coil calculator (Excel) of Jim Koehler to calculate its specs.

Willy

[Ezra Teneflin]

Re: PC based Proton Precession Magnetometer?

Work is focused on the input amplifier at present, but I do have some coil questions. (The TI INA217 and INA166 look very promising) Common mode rejection for outdoor fixed coils could be very helpful, especially with active CMRR techniques using the first amp outputs available on the 166).

Coil questions. Using more conventional solenoids (two for noise rejection), they should be normal to the field for best polarization. I notice that some applications place the coil at an angle from vertical to be normal to the DIP angle of the Earth's field. I'm wondering if resting a solenoid on its side, whilst normal to the field direction (oriented with the long direction E-W) accomplishes the same thing?

On toroids, there are certainly some advantages on noise rejection and physical placement. But, isn't any volume of sample not normal to the field direction left unpolarized? In other words, isn't a great deal of the sample volume left "unused" in a fixed sensor application?



Click Here

[Willy Bayot]

Re: PC based Proton Precession Magnetometer?

Sorry, I have no sensible answer to these questions. Actual experimentations should probably give you SOME answers.
1.2 nV/root Hz is a VERY good specs for a first amplification stage.

Willy

[Ezra Teneflin]

Re: PC based Proton Precession Magnetometer?

My question was phrased poorly. ALL of the sample (water, kerosene, etc) should be polarized in a toroid, for example, during a polarization cycle, assuming that the polarizing field is strong enough. The polarizing field is so much stronger than the ambient field that the ambient field is a non-factor during the polarization cycle.

However, only that part of the sample where the resulting polarization was normal to the Earth's field should give the strongest procession signal (as I understand this so far). Other parts of the volume should give less signal falling off to near zero where the polarization was parallel to the Earth's field.

I have observed this with my small bottle of distilled water tilted from the vertical so as to be normal to the field direction (accounting for the dip angle of the field). I didn't get a chance last round of experiments to try the coil on it's side with an E-W solenoid direction, but I did notice in previous experiments that a 'good signal' drops off fast as the bottle orientation is change from being perpendicular to the magnetic field.

[Ezra Teneflin]

Re: PC based Proton Precession Magnetometer?

On further consideration, there is a plane normal to the direction of the field (think of a vector representing the direction of the Earth's field pointing into the plane), so if the toroid was placed exactly on that plane (at an angle like the tilted solenoids on bottles), would the sample in the entire toroid process equally? Has anyone tried this?

[Willy Bayot]

Re: PC based Proton Precession Magnetometer?

Does a toroïd not correspond to an infinite solenoïd connecting its top to its bottom?
I always have understood that the best position for the measurement using a toroïd is on that very plane. At least, that's what Jim Koelher explains in his good descriptive document.

Willy

[msolc]

Still someone interested here?

[Willy Bayot]

reviving the PC-based PPM discussion

Still someone interested here?

Hi,

This discussion had started quite a while ago and much water has passed under the bridge since then.
The person who initiated this discussion stream was Steve Breke (Steve B.) with whom I have made a study project at that time on this subject. After a while, he got financial problems after having lost his job and I lost him as well. I would be grateful if anybody could help me to re-locate him.

About the use of audio card on a PC to analyze the signal, I stay with my previous remarks.
About the sensor coil configuration, we have now a more practical view of it.
After a large number of tests during the prototyping phase of our now-mature PPM system, we have concluded that a toroid does not give enough advantages compared to the solenoid and its much easier building process.
It is true that, in theory, a solenoid can give a null signal in its worst orientation while the toroid still gives half the signal in this orientation.
However, most of the world locations present a magnetic field vector whose VERTICAL component is rather large compared to its total value. It just means
that at most earth locations, it is sufficient to keep the main axis of the sensor coils horizontal to get a good enough signal in any horizontal orientation. That has been proven by actual implementations at various spots of the world.
Moreover, we believe at first that the side-by-side coil configuration was necessary to get a good noise rejection but we found out that the inline coil configuration is as good while being much easier to mechanically package than the first configuration. It is sufficient to insert the two inline coils in a PVC tube of appropriate diameter.

If you wish to continue this discussion, go on, I'll be happy to help anybody understanding the practical aspects of this technology.
By the way, our web site is there: http://users.skynet.be/fa352591/index.htm

Willy