# Hall sensor confusion with unipolar and bipolar as well as magnetic induction sensitivity

I want to create a simple bicycle computer that shows my speed and RPM. I'm thinking of using a TLE4905L or TLE4906L (which I think is more sensitive) - both unipolar.

Now I'm a bit confused about unipolar vs bipolar. When my wheel spins, the magnet will pass in front of the sensor (sensor is static, magnet is rotating with the wheel on a spoke) with the same orientation, as it passes, the magnetic field increases, then decreases. With unipolar sensors this should count as a rotation (a rising and decreasing edge at the output pin - of course I'll use just one of the edges at the controller level). On the other hand with bipolar sensors I think I'd need two magnets, one near the other oriented with opposing poles towards the sensor, one would make it go high and the opposing pole from the second magnet passing by would make the signal go low, am I understanding this correctly?

Now for the sensitivity of the sensors. Let's assume, for the sake of the argument, a typical fridge magnet at 5mT as per Wikipedia. With the unipolar versions of the sensors with the TLE4905 (page 8 of the data sheet linked above) the strength of the magnet would not be enough to trigger the sensor which has a Bop=7.5mT. The TLE4906 is a bit more sensitive at Bop=6.5mT, but still not enough.

Is this right? Now I either need a stronger magnet or a more sensitive unipolar sensor, right? If I got the unipolar vs bipolar part right, then the answer should be yes, otherwise I need some explanation on why not.

with bipolar sensors I think I'd need two magnets, one near the other oriented with opposing poles towards the sensor, one would make it go high and the opposing pole from the second magnet passing by would make the signal go low,

No. A single magnet has both north and south poles, so with correct orientation it can do both. Two magnets would only be required if you wanted the output to go high and low at widely separated positions on the wheel.

Let's assume, for the sake of the argument, a typical fridge magnet at 5mT as per Wikipedia. With the unipolar versions of the sensors with the TLE4905 (page 8 of the data sheet linked above) the strength of the magnet would not be enough to trigger the sensor which has a Bop=7.5mT. The TLE4906 is a bit more sensitive at Bop=6.5mT, but still not enough.

Is this right? Now I either need a stronger magnet or a more sensitive unipolar sensor, right?

Yes. Best solution may be to use a Neodymium magnet, which is much stronger than a fridge magnet.

• A single magnet has both north and south poles, so with correct orientation it can do both - Yes, but on a wheel, the magnet is fixed on a spoke, only ONE pole of the magnet is oriented towards the sensor during an entire ride. The magnet's poles do not reverse positions from one rotation to the other, so only the south pole may be oriented towards the bipolar sensor, hence, not triggering it since the magnetic field with only be either positive or negative. Or have I got this completely wrong?
– Paul
Commented May 28, 2020 at 4:48
• (Since I think what you said applies only if the magnet flips from one rotation to the other, ie. a magnet flip counts as a rotation, in my case I cannot flip the magnet from one rotation to the other)
– Paul
Commented May 28, 2020 at 4:50
• If the magnet poles are aligned with the direction of movement then both north and south poles will be presented. If they are perpendicular (ie. along the spoke) then only one pole will present and you will need a unipolar sensor. A block magnet could be attached in either orientation, but a bar or cylinder magnet might need to run along the spoke and so (depending on the direction of magnetization) need a unipolar sensor. But you are looking at unipolar sensors anyway, so... Commented May 28, 2020 at 5:53
• Ok, I see what you mean now. The unipolar vs bipolar stuff was for my understanding.
– Paul
Commented May 28, 2020 at 7:31