Facebook
Google
GitHub
Linkedin
Hello there,
There are a few things about magnetic sensors that are most important. I'll try to outline some.
First, the type. The linear sensor is the one you want, not the switch type sensor. The linear type are more sensitive.
Second, the sensitivity. They make various sensors with various sensitivity spec's usually listed in Gausses or possibly in Teslas. Get the best sensitivity you can find. This plays in with the noise too. The better the sensitivity the better the signal to noise ratio, and there is no easy way to filter the noise because it is wideband unless perhaps you go with a microcontroller too.
Third, the distance. The distance is one of the most important things in magnetic circuits. The closer to the object, the larger the signal.
Fourth the type of magnet. The strongest magnet you can find, and biggest you can tolerate. Rare earth is a must. A bar type magnet will help ensure that there are only two possible orientations for magnet for the person holding the magnet since they want to keep it hidden.
Fifth, the physical construction. Using a lone sensor with nothing else offers the worst sensitivity with any sensor. Magnetically active materials that are made for magnetic applications like transformers offer some degree of magnetic amplification. They increase the apparent sensitivity of the sensor with liittle cost. Even a 10 penny nail might work better than nothing. A black binder clip works too, but the pressure on the plastic case of the sensor can crack the case. A block of material from an EI transformer would work pretty good, silicone grain oriented steel. That helps to couple the magnetic field to the sensor and and effectively amplifies the flux density and thus the output of the sensor becomes larger. That helps get past the noise floor too so that any amplification works mostly on the signal. Remember you've got two poles on both the magnet and the sensor, and the better you can couple those two poles to the sensor poles the stronger the signal. They must be coupled individually too, for example the tightest magnetic circuit would be that with two magnetic material blocks running one from the N pole of the magnet (touching it) to the N pole of the sensor (touching it), and the other one running from the S pole of the magnet to the S pole of the sensor. That would make a very strong signal (and reversing the magnet would simply reverse the output polarity relative to 1/2 of Vcc).
Sixth, electrical amplification is a given, although the wideband noise can be very problematic giving false readings. Using a microcontroller can help to minimize the noise using statistical methods. If not, a low pass filter with delay circuit at least for some help there.
Seventh, orientation. If the magnet in the hand of the person is oriented one way or another you'll get higher or lower readings. If this is a bar magnet, you will have the knowledge that it is probably being held in the hand like a roll of coins and thus you will know something about the orientation and that willl help detect it. You can then even detect WHICH way it is being held: norith pole to the right or to the left.
Lastly, i think it is a cute idea especially for kids who get a kick out of this stuff.
Last edited:
