kevin@cheetah.inmos.co.uk (Kevin Cameron) (06/24/91)
I always wanted to build a Hall-effect compass - but never got round to it. The major problem seemed to be that the devices are quite temperature sensitive and are therefore difficult to get absolute values from, however it does always exhibit minimum resistance at minimum field strength. I would therefore suggest that an accurate measurement of the Earth's magnetic field can be found by placing the device inside a coil which can be magnetised to neutralise the Earth's field, and the coil current will then be proportional to the Earth's magnetic field. Two devices at right-angles can then give you a bearing. Hope somebody finds this useful. Apologies if this re-iterates anything in "Radio Electronics", which I havn't read. ------------------------------------------------------------------------------- Kevin Cameron INMOS, 1000 Aztec West, Almondsbury, Bristol BS12 4SQ, UK kevin@inmos.co.uk Tel: (UK) 0454 616 616 x364, Fax: 617 910
neufeld@aurora.physics.utoronto.ca (Christopher Neufeld) (06/25/91)
keving@cheetah.inmos.co.uk (Kevin Cameron) writes: Followup-To: Hall Effect Sensor This really fouled up my machine, it refused to post off-site, so I'm restarting the thread manually. >I always wanted to build a Hall-effect compass - but never got round to it. > [ goes on to discuss nulling field coil and Hall effect probe ] This is not the way to go. The Hall effect is not strong enough to pick up gauss-sized fields easily. It's usually used to measure fields some ten thousand times larger. At such low fields you'll have real problems with thermally induced voltages when you try to measure the voltage across the sample. Further, the errors will be large enough to make bearing readings (from knowledge of the component in two different directions) essentially useless. Now, a clever scheme which ought to be workable is a fluxgate magnetometer. I won't go into details of the construction, you can find them in any IEEE proceedings on magnetism. You need a signal generator and a way to extract harmonics of the input frequency, then some nulling and sensing coils, with a material chosen to saturate magnetically in the regime of interest. You can buy packaged fluxgate magnetometers the size of a 35mm film case, tri-axial, with analog outputs proportional to the magnetic field strength along each axis, for about $3000. That package contains everything, including the multiplexing hardware, frequency generator and analyzer, and three fluxgates. It just takes regulated supply voltages in. I offer this as an example that the thing is possible for not exorbitant amounts of money, and in a small case. >Kevin Cameron INMOS, 1000 Aztec West, Almondsbury, Bristol BS12 4SQ, UK >kevin@inmos.co.uk Tel: (UK) 0454 616 616 x364, Fax: 617 910 -- Christopher Neufeld....Just a graduate student | If ignorance is bliss neufeld@aurora.physics.utoronto.ca Ad astra | why aren't there more cneufeld@{pnet91,pro-cco}.cts.com | happy people? "Don't edit reality for the sake of simplicity" |
kevin@cheetah.inmos.co.uk (Kevin Cameron) (06/25/91)
|> keving@cheetah.inmos.co.uk (Kevin Cameron) writes: |> Followup-To: Hall Effect Sensor |> >I always wanted to build a Hall-effect compass - but never got round to it. |> |> This is not the way to go. The Hall effect is not strong enough to ^^^^^^^^^^^^^^^^^^^^^^^^^^ |>the regime of interest. You can buy packaged fluxgate magnetometers the |> the magnetic field strength along each axis, for about $3000. That ^^^^^ |> Christopher Neufeld....Just a graduate student | If ignorance is bliss I'm sure you're right about it working, but $3000 seems a bit expensive compared to an ordinary magnetic compass. You could probably buy a global positioning system for that price (Inmos have developed a pocket one based on transputers). It only needs to be accurate to ~1 degree to be competitive with a magnetic compass on performance, but it would need to cost nearer $100 in total to be worth buying instead. How strong is the Earth's magnetic field anyway? ------------------------------------------------------------------------------- Kevin Cameron INMOS, 1000 Aztec West, Almondsbury, Bristol BS12 4SQ, UK kevin@inmos.co.uk Tel: (UK) 0454 616 616 x364, Fax: 617 910
neufeld@aurora.physics.utoronto.ca (Christopher Neufeld) (06/25/91)
In article <16869@ganymede.inmos.co.uk> kevin@cheetah.inmos.co.uk (Kevin Cameron) writes: >|> keving@cheetah.inmos.co.uk (Kevin Cameron) writes: >|> >I always wanted to build a Hall-effect compass - but never got round to it. >|> >|> This is not the way to go. The Hall effect is not strong enough to > ^^^^^^^^^^^^^^^^^^^^^^^^^^ >|>the regime of interest. You can buy packaged fluxgate magnetometers the >|> the magnetic field strength along each axis, for about $3000. That > ^^^^^ >|> Christopher Neufeld....Just a graduate student | If ignorance is bliss > >I'm sure you're right about it working, but $3000 seems a bit expensive >compared to an ordinary magnetic compass. You could probably buy a global >positioning system for that price (Inmos have developed a pocket one based >on transputers). > Well, that one is more sophisticated than it has to be for your purposes. Potted in epoxy, intensities measured to high precision, you could probably use it to detect submarines. A cobbled-together fluxgate magnetometer shouldn't run $3000. I just gave that as an example. A similar Hall effect sensor would contain a variable current power supply, a precision variable voltage source and meter, and some logic to calculate the field given current and voltage values. That stuff isn't cheap if bought as commercial products. >It only needs to be accurate to ~1 degree to be competitive with a >magnetic compass on performance, but it would need to cost nearer $100 >in total to be worth buying instead. > >How strong is the Earth's magnetic field anyway? > About 0.5 to 1.0 Gauss, or under 1e-4 Tesla. Hall effect sensors which I've used have been for fields in the neighbourhood of 0.5 to 2.0 Tesla. You've got a factor of 10000 to recover there. >kevin@inmos.co.uk Tel: (UK) 0454 616 616 x364, Fax: 617 910 -- Christopher Neufeld....Just a graduate student | If ignorance is bliss neufeld@aurora.physics.utoronto.ca Ad astra | why aren't there more cneufeld@{pnet91,pro-cco}.cts.com | happy people? "Don't edit reality for the sake of simplicity" |
ieeug330@Eagle.oscs.montana.edu (Mark Wistey) (06/26/91)
I have a copy of a physics lab on magnetometry using a bar magnet, a magnetic compass, and a yardstick. It was sensitive enough to detect the magnetic field produced by an H.P. calculator sitting nearby! But what can I say, it was a lab, and we were bored... Anyone who's interested, just ask at any of the addresses below. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Mark Wistey... gmw1142%msu3.dnet@mtsunix1 (BITNet) | Bicycling, Apple II's, gmw1142%msu3.dnet@terra.oscs.montana.edu (InterNet) | Christ-like living, 513 S. 9th, Bozeman, MT 59715, USA (BikeNet) | Music, and Physics :-)
mahaun@sactoh0.sac.ca.us (Mark A. Haun) (06/26/91)
In article <1991Jun25.161850.11815@helios.physics.utoronto.ca>, neufeld@aurora.physics.utoronto.ca (Christopher Neufeld) writes: > In article <16869@ganymede.inmos.co.uk> kevin@cheetah.inmos.co.uk (Kevin Cameron) writes: > > > >I'm sure you're right about it working, but $3000 seems a bit expensive > >compared to an ordinary magnetic compass. You could probably buy a global > >positioning system for that price (Inmos have developed a pocket one based > >on transputers). > > > Well, that one is more sophisticated than it has to be for your > purposes. Potted in epoxy, intensities measured to high precision, you > could probably use it to detect submarines. A cobbled-together fluxgate > magnetometer shouldn't run $3000. I just gave that as an example. A > similar Hall effect sensor would contain a variable current power > supply, a precision variable voltage source and meter, and some logic to > calculate the field given current and voltage values. That stuff isn't > cheap if bought as commercial products. > > >It only needs to be accurate to ~1 degree to be competitive with a > >magnetic compass on performance, but it would need to cost nearer $100 > >in total to be worth buying instead. For anyone not aware, Radio Shack sells a fluxgate compass for $54.95 (Cat. # 63-641, page 50 in '91 catalog). I don't know its accuracy. Keep in mind that designing a working electronic compass is a lot easier than actually measuring changes in the Earth's field. For a compass, all you need to sense are the relative strengths of the field in two directions. I would not be suprised if Hall devices *can* be used in a crude electronic compass. Of course, if you want to actually watch the field and see changes in it, you need to attain a resolution down to at least 100 nT or so (10 nT would be more realistic). That is a tricky thing to do, even with fluxgates. -- Mark A. Haun / 3445 Del Mesa Ct. / Sacramento, CA 95821 / Phone: (916) 488-2965 UUCP: {ames | apple | sun}!pacbell!sactoh0!mahaun | Amateur Radio KJ6PC INTERNET: mahaun@sactoh0.SAC.CA.US / pacbell!sactoh0!mahaun@ames.arc.nasa.gov Amateur Pkt Radio: kj6pc@wa6nwe.#nocal.ca.usa -or- [44.2.0.56] on 144.93 MHz
jim@ghp.UUCP (Jim Stewart) (06/26/91)
Kevin Cameron writes: >|> You can buy packaged fluxgate magnetometers the >|> the magnetic field strength along each axis, for about $3000. That >I'm sure you're right about it working, but $3000 seems a bit expensive >compared to an ordinary magnetic compass. I agree... The autopilot in my boat costs about $900 CDN. It uses fluxgate technology, and with the $900 CDN you also get a waterproof control box, large servo motor, and the mechanical stuff to tie it all together. It has +/- degree controls so it must be fairly sensitive, although it just works relative to the last position, not absolute with respect to the earth. I would be curious as to where to find just the sub-assembly. -- Jim Stewart, VE3SRJ UUCP: jim%ghp@mnetor.uucp BELL: (416)862-0430