magnetometry

[red-panda]

Thanks Tom and Dan for the encouragement, here's a description so far of my experiments with magnetometry.
My interest was originally sparked by the UK tv programme, Time Team, where the geophycisists would produce spectacular plans of long forgotten structures, using magnetometry and resistivity. A quick search on the web found a gradiometer design published in EPE in July 94, a copy of which was duly purchased, using sensors from Speake (FGM1), also available from Fat Quarters in the USA, and a resistivity design in EPE. I duly built the gradiometer, which seemed to work better than I hoped. A move to Canada followed, along with the realisation that, unlike England, where you could almost pick a field at random and find something, I probably needed to find another way of playing with the sensors. Remembering that mineral exploration companies, and the military, have both found uses for airborne magnetic surveys, I decided to miniaturise the whole setup to fit on a micro (Hornet II) helicopter.
The sensors basically give a pulse train output, whose frequency depends on field strength, and is around the 75KHz region in nominal conditions. To get this down to something I could record using the rpm input on an (Eagletree) FDR the output signal first goes to a 4013 configured as a divide by 2, and is then pseudo mixed with either the second sensor output for a gradiometer, or an rc oscillator for magnetometer, and again divided to give a nicely shaped output pulse. The gradiometer is therefore giving an output of the unsigned difference between the sensors, and the magnetometer an output related to the field strength. The two sensors are mounted on a 0.5 meter length of cf tube, which will be mounted across the heli frame with adjustment on one so the pair can be nulled. The single sensor is simply mounted on a length of cf to get it clear of the frame, and can be nulled using the rc oscillator frequency.
Initial tests have revealed that the concept basically works, but could be improved. I'm currently awaiting the arrival of a coax heli, as the Hornet with me at the controls is far too twitchy (tests so far have been done with the heli strapped to a broom handle, with me holding the other end, and if you want to get strange looks in the local park, this is definitely the way to go!), and I'm also trying to find where I hid my PIC programming kit, as I see no reason why a microcontroller and a bit of i2c memory shouldn't replace the assortment of ics, and the fdr, although I would then lose the future possibility to log gps and altitude.
I'm guessing that the gradiometer will only be useful at relatively low altitude, for locating smaller anomalies, but the magnetometer should work at higher altitudes for things on the scale of iron ore deposits.
Pictures below I hope are self-explanatory, sorry about the hand drawn diagrams, I haven't got any cad software.
If anyone wants more info, please let me know.

chris

Links
www.epemag.wimborne.co.uk
www.speakesensors.com
www.fatquarterssoftware.com

[kd7ost]

I'm very intrigued by your project Chris.

Do you have any sense for how deep the magnatometer will read anomolies? I guess it would depend on the strength of the gauss field of said anomoly? (That's a question. This is new to me and I had to go read up on magnetometers before replying) ;)

Also, I'm not understanding how you will record the anomolies. Do you have a way of marking the location in mind.

Very cool project. I don't know any hobbyist that is doing what you're doing. Seems like you're breaking new ground as far as I know.

Dan

I found this on Gradiometers.

http://www.ultramag.com/gradcomp.htm

[red-panda]

Dan
Good questions, and I've had to think about the answers.
The sensitivity of the gradiometer, properly set-up, is around 1Hz per nano-tesla. As far as I can find from my references a 1m deep by 1 m wide trench in rock, backfilled with soil, and covered by about 1m of soil will give an anomaly of around 1nT at a 1m survey height, a similarly buried kiln or furnace ruin will give around 50nT, and for lumps of iron 200nT is a possible figure, however as the field drops rapidly with distance I suspect an altitude much above 5m would be no good. Luckily the way I fly model helis means an altitude much over 3m is very unlikely. For the magnetometer the sensitivity is obviously the same, but since we're looking at the overall field, and not just local minor variations, I'd guess that a decent iron ore deposit for example would show up from several tens of metres. (sorry about the metric units – I'm originally from Europe). The problem would be flying straight and level to avoid errors induced by the relatively large earth's magnetic field.
Regarding the positioning problem, full-size airborne magnetometry uses gps, but if your'e looking for a mountain of iron, 3m is a good enough accuracy. I'll therefore say its good enough for us as well, and once I've saved up a few dollars, I'll buy an Eagletree gps to go with the fdr which I'm currently using to record just the survey data. The gradiometer is a bigger problem as smaller signals need a much better accuracy. The archaeological surveyors either use standard surveying techniques to put a physical marker grid in place, or dgps to do it electronically. My first thoughts for simplicity would be a camera pointed vertically down, with an overlaid display of mag data in an analog form, or, since I'm not into ap, maybe use the Eagletree gps option, and make sure I fly over an obvious reference point or two so I can calculate an offset for gps errors.
A couple of sidenotes. I haven't yet properly nulled my gradiometer, as to do so would not only involve standing in the middle of my local park to get as far as possible away from all the magnetic influences in the house, but also removing my belt-buckle, cellphone, coins, keyring and so on. I need to build up courage to do this. Secondly the mapping software supplied with my ordinary gps is not as accurate as the gps. If I stand in the front yard with plenty of satellites in view, my accuracy is around 3m, and I show up as being on the opposite side of the street!

Chris

Link
www.eagletreesystems.com

postscript
Have just ordered a new heli platform, and contacted Eagletree about upgrading my fdr.

[Tom Harper]

Chris,

An exciting project. FYI you might check out this thread of my GPS efforts:

http://www.rcgroups.com/forums/showthread.php?t=546671

I too bought the EPE boards but never got around to building them. I want to do some soil conductivity logging. They were an excellent education. Problem is that technology changes faster than I can build projects. You might look into the TI MSP430 processors. They are 'in circuit' programmable. The starter kit includes a small applications board and it's $20 total. If you are interested I can get you started with some sample code.

I'm sure there are tons of difficulties that can be identified in using a magnetometer pole on a model heli. But, that's not the important part. The fact that you are actually doing it is great!

Looking for a fire pit in a couple of square kilometers is not the most likely application. I think the heli would be useful to survey a known ruin. Say something 50 meters square. Multiple passes over a known grid may show up anomalies even if they are small. Some of the magic would be in the data crunching.

I use power point for schematics. Cheaper than CAD.

Let's hear more.

[John O'Sullivan]

I think this thread is one which can lead to a new perspective on airborne surveying.
I am a geological engineering who has been involved in mineral exploration since 1960. During this time I originally concentrated my focus on geochemical surveying, and was involved the discovery of a number of world class Carbonate hosted Lead Zinc deposits in Ireland. These were discovered by Canadian mining companies who placed a heavy emphsais on Magnetic and Electromagnetic surveying. Unfortunately the Irish deposits responded poorly to Magnetic amd Electromagnetic surveys because of the poor response of the principal zinc mineral, sphalerite, to those geophysical techniques. Induced polarization, which was related to Chargeability effects was found to be effective however.
In recent years, diamond exploration in Canada has blossomed and the discovery rate is unsurpassed anywhere. The primary tool in these discoveries is airborne magnetic surveying. The kimberlite pipes, the host rocks for the deposits are characterized by circular magnetic high with a circular magnetic low over the central part of the pipe.

Within the past four years an off-shoot of Austin Developments, Universal Wing has developed a atonomous drone Aeromagnetic system which can carry out grid based surveying at distances up to 80 miles from take-off areas. Initially they were using Dara Aviation aircraft ( http://www.daraaviation.com/ ), but more recently have designed their own more conventional design. they have tested their models in extreme arctic conditions. Their web site ( www.universalwing.com is woefully out of date). However, I spoke with them at the Prospectors and Developers convention in Toronto in March and am highly impressed with their new design. I do not want to jump the gun in presenting their new design before they do, but I do have photos if anyone is interested I can email them the photos privately.

These guys, based on Vancouver are no fly-by-nights as one of their principlals, Buddy Doyle, was a key player in the discovery of Canada's first diamond mine.

The model carries two magnetometers placed in the wingtips of the 13 foot model. Vertical gradient magnetics is also in the works.

Red-Panda, I believe you are at least somewhat in touch with the mineral exploration potential of magnetic surveying and in particular vertical gradient mag. You are operating on a micro scale compared to the Universal Wing people, but the basic principals are similar. The Universal Wing guys are Vancouver based and I suggest you contact them. They are down-to-earth guys and your ideas may well fit in with their development plans. You can use me as an introduction.

I am not a geophysicist, and an totally confused by most of the stuff they turn out, but the results can be impressive.

John O'Sullivan

john O'Sullivan

[red-panda]

Sorry to be slow in replying, I'm semi-retired, so things like the bathroom reno. tend to interfere with the serious work!
Tom, I like the idea of the 50m grid micro-survey, makes absolute position so much easier. I'm still looking at the Eagletree gps option, but an area this small would mean relative errors are virtually non-existent.
John, thanks for the info, I could walk to those guys office in about 15 minutes. They are about 2 orders of magnitude better than me in sensitivity, but their platform is also a bit bigger than mine, which probably also applies to their budget, however given a few weeks I should be in a better position to to know if I'm actually a mad or sane experimenter. I've just ordered some more sensors to build a duplicate rig as I suddenly realised that using the rig on an electric heli, with all those magnetic fields whiirling around at high speed, could well be burying any useful data in the noise, so want to compare low hover data with an identical set-up at the same height, with or without the heli nearby.
Regarding data visualisation, since I'm using the eagletree rpm input as field strength, as far as I can tell, I should be able to generate a 3D plot of position versus "rpm" with the gps option to give some sort of meaningful display. I'm fairly happy to write a bit of PIC code to sort the sensor to data-logger interface, but anything above that comes in to the "buy something of the shelf and modify it" category

[Tom Harper]

Chris,

The EagleTree software is good, but I like to read it into Excel for more flexibility. Their files are space delimited. Excel will read the file with the EagleTree extension. You have to select 'space delimited' when it asks. Excel has good 3D viewing.

BTW: Express PCB has free CAD software for schematics.

http://www.expresspcb.com/

[Ljotham]

Hi Red Panda

very interested in your mag ideas, I am Geophysicists - academic though. I have done archaeo mag (both gradiometer and total field surveys) in the past and they are quite sensitive to upset - i.e. any magnetic material in reasonable proximity to the sensor - is there anything magnetic on your plane ? (although mag distortion from nearby noise can be corrected for). Constant vertical/horizontal attitude, instrument drift, constant elevation, diurnal variation in magnetic field etc etc etc.

However, of these the biggest one for the model plane I suspect, will be elevation and attitude if using a gradiometer (especially with a pair of single vector sensors) and just elevation for total field measurements. For total field measurements (easiest) you will need three fluxgate mags to give the three perpendicular components of the magnetic field and then calculate the resultant total field vector (absolute magnetic field strength/direction) unless the sensors you have are already total field.

As for sensitivity - well - that’s mostly down to elevation as magnetic field strength falls of as 1/ (height cubed) for a sphere (good approximation). In archaeology you often look at variations on the order of 1 nT to 10s of nT (nT is 10e-9 Tesla) from a height of approx ground level (0.5 m) double the height and the field goes down by eight - you see where this is going, not only that but what would be a rapid spike in readings at ground level rapidly broadens as you increase separation from the magnetic body making features less distinct (something to do with Laplace).

From the discussion of sensitivity it is apparent that good elevation information is going to be vital. The aeromag surveys in the past have been on much larger scales where features are of the order of km and height variation is on the order of m thus giving less distortion. The archaeo mag survey features are on the order m so elevation becomes critical.

You then need to plot your data so you will need accurate positioning.

please keep posting on your developments - there is a lot to do but it would prove very usefull for surveying costal/marshy areas (you could probably charge money).


another fluxgate mag maker to try
is Stephan Mayer instruments - he does a good magnetometer for rc models its tiny about the size of a USB memory key and weighs about 10g


Laurence

[Tom Harper]

Laurence,

What archaeological features were you able to detect? Could you identify ground disturbance as well as metal?

Tom

[Ljotham]

We mostly detect ditches due to magnetic bacteria/debris and burnt brick, flint etc due to thermoremnance - ferrous metal stands out a mile with huge spikes on the data and often has to be clipped - it’s usually bits of barbed wire/tractor and not treasure.

just read the reply concerning apparent conductivity measurement using low frequency continuous wave RF - sounds like EM31/EM34 survey - heavy TX and RX coils if I remember right although could be made lighter - measurements that are made with this set up are usually phase based i.e. the higher the quadrature component the more conductive stuff is around you/nearer it is to you. - Its not that directional (being essentially dipolar radiation) - but the Australian navy were thinking of using it for costal bathymetry

At the mo I am developing lightweight ultra wideband radar antennas/systems for engineering geophys - synthetic aperture radar from a low flying plane hmmm - dreams

seriously though - many unis/R&D dept. are looking into airborne avionics and there is a market for it - one I know of is in inaccessible costal heritage areas threatened by potential sea level rise. Large planes either cant or wont fly low enough

Laurence

[red-panda]

Sorry guys, haven't been around recently as I accidentally found a job, which has cut down on free time, but increased the cash flow, and quantity of odd electronic components arriving in the mail each day. Thanks for the info Laurence, you've brought up many points I was just getting round to considering, and answered my questions about magnetic field strength. I've got a very old Masters degree in engineering lying around somewhere, but it often seems to have forgotten what it learnt, and please give my regards to the pubs of Southampton, as I remember many happy days working in the Nuffield and Mayflower theatres in my former profession.
The test platform will be a co-ax rotor helicopter, and the presence of two brushless motors is a major concern, I'm wondering if mumetal screening may help, but am still constructing the basic set-up, so will run static tests to see how big the problem is first. I've also been sidetracked into designing an altitude hold system to give a stabilised height over ground of around 2m. For position info I'm still thinking of a combination of gps and a vertical video camera, both of which can use standard off-the shelf kit. As usual with my projects, this started as an "I wonder if", and has now developed a life of its own. I'm trying to go with the original lightweight just for fun concept, aiming for an auw around 400gm., but am simultaneously considering the bigger "lets actually make it work" possibility. The upside about the new job is that its at the University of British Columbia, and I now have access to good indoor and outdoor sites for tests, and am looking into ways of tapping into the vast pool labelled research.
I blame Tony Robinson and Time Team for it all.

chris