Build your own gaussmeter

Build your own Gaussmeter

Have you everwanted to find out how strong a magnet really was, or how thestrength of the magnetic field varied as you changed the distancefrom the magnet or the temperature of the magnet, or how well ashield placed in front of the magnet worked?  Voltmeters arefairly inexpensive and easy to find, but where do you purchase aGaussmeter (also known as a magnetometer).  I built ahand-held Gaussmeter for measuring the polarity and strength of amagnetic field.  It uses a linearHall effect deviceand some op-amps and resistors and thingsfrom RadioShack. 

I will first describe avery simple, inexpensive Hall effect device Gaussmeter you canbuild for as little as $6.  Then I will describe a Gaussmeterwith a few more bells and whistles.

An inexpensive Hall effectGaussmeter

Here is a parts list forthe low-cost Gaussmeter:

First, you need a 9vbattery.  You can get them most anywhere.

Next is a battery clip toconnect to the top of the battery.  You get a package of 5 for$1.39.

The 7805 is a +5v regulatorwhich takes the +9v from the battery and reduces it to +5v whichthe Hall effect device will need.  It only costs about$1.49.

You have a couple ofchoices for the Hall effect device.  If you go with acalibrated unit, it will cost a lot more, about $60.  Withthis, though, you get the device and a calibration chart, whichtells you exactly what the output voltage is going to be when acertain magnetic field strength is present.  These photos showyou what you get:

On the leftis the Hall effect device, an Allegro A3516LUA.  On the rightis the calibration chart, showing output voltage from the Halldevice vs magnetic field, plotted every 100Gauss from 800Gaussnorth to 800Gauss south, at three different supplyvoltages.

Another choice is topurchase an uncalibrated Hall device, take a good guess at thecalibration, but still use it for accurate comparisons from onetest to another.   It just wouldn't have an absoluteaccuracy.  To obtain this, There are a couple of easychoices. 
  1.  Purchase a Radio Shack RSU 12035713 for $4.79.  This is an Allegro A3515EU.  It has a sensitivity of 5.0 mV/G, and does not have a calibration chart.  (This is great for weak magnetic fields, but may saturate when measuring strong NIB magnets up close.  To use this with the stronger magnets, you will need to keep the magnet about an inch away from the Hall device.  The device will not be damaged if a very strong magnet is placed against it, the only thing that will happen is that the output of the device will reach a certain voltage limit when the magnet is, say, a half inch away, and the voltage will not change as the magnet gets closer, since its amplifier is saturated.  The voltage will again drop as the magnet is moved further away again.)
  2.  Purchase an Allegro A3516LUA, but without the calibration chart, from Arrow Electronics, for about $2.01.  It has a sensitivity of 2.5 mV/G. 
Allegro can be reached directlyat: 
Allegro
115 Northeast Cutoff
Worcester, MA 01615
Phone: 508-850-3325
Fax: 508-853-7895

You will need something tomount these parts onto, so here again are two possibilities. Use an inexpensive perf board and solder the parts to it, or usethe breadboard and just plug the parts in - no soldering! Unless you've built electronic things before, I would recommend thebreadboard since it is easy to use, easy to change, and can be usedfor other projects in the future.  So that would cost$7.99.

You need a voltmeter forall the projects you're going to work on anyway, so I won't addthat in for this project.  There are different typesavailable, and their cost goes up with features andfunctions.  A basic one that will work well is noted in thetable above.

There!  Going with theperf board, it is only $6.08!!!  With the A3515EU from RadioShack and the breadboard, it will be about $16!  These willhave great relative accuracy!  For better absolute accuracy,it will cost about $71.   (Again, batteries and voltmeter notincluded.)

Now, how do you makeit? 
    Connect the + (red) of the battery clip to theinput of the 7805 (pin 1).
    Connect the - (black) of the battery clip to thecommon of the 7805 (pin 2).
    Connect the +5V input of the Hall device (pin 1)to the output of the 7805 (pin 3).
    Connect the common of the Hall device (pin 2) tothe common of the 7805 (pin 2).
    Set the voltmeter to read 20Vdc max.
    Attach the + of the voltmeter to the output ofthe Hall device (pin 3).
    Attach the - of the voltmeter to the common ofthe 7805 (pin 2) or the common of the Hall device (pin 2).
    You are now ready to snap a battery onto thebattery clip.

Here's a schematic of thecircuit (using the 3503 Hall-Effect Device):

   
    With no magnet near the Hall device, measure andnote the output voltage reading.  Call this V0.  Itshould be about 2.50Vdc.
    Now, with a magnet near the Hall device, youwill see the output voltage change.  If it is a South pole,the voltage will increase.  If it is a North pole, the voltagewill decrease.  Call this voltage reading V1.
    We will say that the sensitivity of the Halldevice is 2.50mV/G as found on their data-sheet.  Call thisk.
    Therefore, the Magnetic Flux Density you aremeasuring from that magnet can be calculated as:
        B = 1000*(V0-V1)/k, inGauss.
        Please note that with acalibrated Hall device, you would be given actual data measurementsfor the V0 value and for the k value.
        For example, suppose youmeasured 2.48Vdc for V0 and 1.32Vdc for V1.  Then B =1000*(2.48-1.32)/2.50 = 464Gauss, North pole (because it ispositive).
        For another example,suppose you now measured 4.56Vdc for V1 with the same Halldevice.  Then B = 1000*(2.48-4.56)/2.50 = -832Gauss, Southpole (because it is negative).

    See howeasy that is?  You can make your own plot using Excel so youdon't have to calculate all the time.  If you're takingmeasurements, just write down the output voltage and do thecalculations later.  You can simply use it to tell you if youhave a North if the output voltage decreased from V0, or a Southpole if the voltage increases from V0.

Here are some photos ofthis simple, inexpensive Gaussmeter.

Photo 1 is anoverall photo of the breadboard circuit.  Let's look at theclose-up in photo 2.  The 9V battery is at the bottom, the7805 voltage regulator is on the top left, the Hall device is onthe top right.  The red lead from the 9V battery goes to pin 1of the 7805.  The black lead from the battery goes to pin 2 ofthe 7805.   The output of the 7805 (pin 3) is connected by agreen wire to pin 1 of the Hall device.  Pin 2 of the 7805 isconnected by a black wire to pin 2 of the Hall device.  Please note that the marking on the Hall device (giving its partnumber) is facing the camera.  The voltmeter common (black) isconnected to pin 2 of the Hall device.   The voltmeter input(red) is connected to pin 3 of the Hall device.  ( I got thevoltmeter from a Home Depotstore near here for about $20.)  That's all there is!  Great, or what?!

Photo 3 shows the voltageat pin 3 of the voltage regulator.  Ideally it is 5.00 volts,but we measured 5.02, which is close enough.

Photo 4 shows the output ofthe Hall device when no magnet is nearby.  Ideally it is 2.50volts, but we measured 2.59.  This would be our V0 as notedabove.  The Hall device I have here is an Allegro UGN3503U,with a sensitivity of about 1.3 mV/G.

With a disk magnet sittingon top of the Hall device, the voltmeter is measuring 1.94volts.  This means that the Gauss measurement is1000*(2.59-1.94)/1.3 = 500 Gauss, North pole.

With the disk magnetflipped over, the voltmeter is measuring 3.22 volts.  Thismeans that the Gauss measurement is 1000*(2.59-3.22)/1.3 = -485Gauss, South pole.   You will notice that the placement of themagnet with respect to the Hall device is very critical, since themeasurement varies across the surface of the magnet (as it issupposed to, being strongest at the edge, not themiddle!).

With a NIB magnet sittingon top of the Hall device, the voltmeter is measuring 0.99volts.  This means that the Gauss measurement is1000*(2.59-0.99)/1.3 = 1231 Gauss, North pole.

With the NIB magnet flippedover, the voltmeter is measuring 4.30 volts.  This means thatthe Gauss measurement is 1000*(2.59-4.30)/1.3 = -1315 Gauss, Southpole. 

Now, the absolutevalue is not going to be correct since I don't have a calibrationchart with this device, but the relative measurement will be asaccurate as the Hall device, typically within 10 Gauss for theA3515 and the A3516 devices from Allegro.   From themeasurements, I know that the NIB is 1231/500 = 2.46 times strongerthat the disk magnet!  So, thisgaussmeter will work well for measuring the variation ofa magnet's flux density with respect to temperature verywell!