Magnet Myths and Misconceptions

[TinselKoala]

@Itsu

I was using the same techniques taking a measure along the magnet side getting the same results. As well when rotating the probe on each end or pole I produced the same results as you have.


Now if you reduce the number of magnets so the width is near the same as the height then always point the same sensor face at the center of the stack you will see the field resembles two half spheres of opposite polarity. A cylinder magnet of equal diameter/height will produce better results.

The perpendicular scan refutes this interpretation.

This would seem to contradict the looped flux lines model in my opinion and there is a point above and below the magnet center where the sensor can be rotated 90 degrees with no change in magnitude nor polarity which raises many questions.


AC

Yes... like how can engineers who "believe" in the conventional picture possibly actually design things that work, if their beliefs are so wrong? 

That particular example gives no new information, if you actually know how the Hall sensor responds to orientation of flux lines and flux density.


Again, what is the part number of your Hall sensor, and can you provide a data sheet for it?

[minnie]

Koala,
        You've got an inventive sort of mind. Is there any way you could arrange
your magnets to give a "figure of eight" to satisfy Tinman's aspirations?
             John.

[allcanadian]

@Poynt99

Are you certain your sensor isn't just saturating and that's why there is no change?

There lies the question, It would seem like it is however once the area in between pole planes is crossed the polarity and measure remain constant when the sensor axis is aligned with the magnet NS axis as Itsu has shown.
Then when the the sensor axis is perpendicular to the magnet NS axis (sensor facing side of magnet) there is another plane extending from the center outward again the measure and polarity remain constant.


At the intersection of the two planes where the sensor can see a value of +1 parallel to the NS axis and a value of +1 perpendicular to the NS axis it can rotate 90 degrees with no change in the values. The values are not max values so I question if saturation is the cause. The whole thing is questionable because sensor stability along the axis of rotation is extremely sensitive and needs to be verified by someone else here.


AC

[allcanadian]

@TK

Again, what is the part number of your Hall sensor, and can you provide a data sheet for it?

I can barely see the damn sensor let alone the black writing on the black body printed on the atomic scale, lol. I tried and using my magnifier but it was still a no go, I'm fairly certain it was an allegro ratiometric hall sensor because I did not want a switch I wanted a ratiometric sensor when I ordered them a couple years ago. I order them for doing exactly this kind of work.


AC

[TinselKoala]

Koala,
        You've got an inventive sort of mind. Is there any way you could arrange
your magnets to give a "figure of eight" to satisfy Tinman's aspirations?
             John.

I don't know, perhaps by gluing two like pole faces together, or with an intervening bit of iron or steel, maybe.
But Tinman isn't exactly "wrong" in his picture either, as has been explained by several people in the thread up above.

My demonstration of the mapping with the plane of the Hall sensor parallel to the long axis of the magnet stack will be viewable in a few minutes; it's being rendered and will be uploaded to YT shortly.

@AC: The labelling on my A3503 sensors is visible in the closeup image in my latest video. Typically only the last three digits of the part number are shown on the small Allegro packages. So you see "503" in the top row, and a date and/or mfg. location code in the bottom row of digits.