Magnet Myths and Misconceptions

[TinselKoala]

@AC: Can you explain why a stack of disk magnets is not equivalent to a bar magnet?

Much of the rest of your response is adding complexity to the rather simple hypotheses I stated.  But it appears that you agree with what I stated, and you even have apparently confirmed that your data indicate parallel lines of field instead of the "peanut waist" picture:

That is exactly the issue I had and it was constant, the far field would appear constant however the near field not so much. I would think a change in sensor orientation would show an increasing/decreasing field strength dependent on the polarity at that point so long as the sensor rotates about it's own axis. As I said, even a small deviation from the sensor axis ie rotation changes the measure dramatically.

Am I right about that? Stripped of the extra stuff about rotation?

[MarkE]

I have some ratiometric Hall effect sensors, but unfortunately they only cost me about 50 cents each. Allegro A3503, data sheet attached below. Will those do, even though they are not $2.00 sensors?   

They of course work fine.

But I don't have a bar magnet! Can you believe it. All I have are disk magnets, ball magnets, small block magnets etc. 

1)Do we think that the community here will accept an experimental result obtained from a stack of disk magnets, instead of a solid bar magnet? If not, why not?

A stack of short cylindrical magnets magnetized through the thickness work fine.  Anything that ends up making a single composite dipole is fine.  I've done this with bar magnets, cylindrical magnets, plastic magnets etc.

The next thing to do of course is to try to imagine, that is _hypothesize_, what the sensor reading would look like in the various test conditions and under the two competing claims of the shape of the field. As a general principle of experimental design, this should be done _before_ any actual data gathering is done in an experiment.

I think it is obvious that if the conventional view is correct then the flux densitiy through the sensor and therefore perpendicular to the dipole axis approaches zero near the center of the dipole.  After all, all the flux is supposed to be parallel to the dipole in that region, leaving nothing to be perpendicular.  This should be true independent of which side of the dipole the sensor is located: east or west , and independent of whether the sensor is rotated for with its "south" sensing side towards the dipole center, or the "north" sensing side towards the dipole center.  If the figure eight idea is correct then there should be a big increase in flux density close to the dipole center.  It should manifest as two peaks of opposing polarity.

2)The Hall sensor of course will read its maximum value for a given field strength in a given location if the "field lines" are exactly perpendicular to the sensor surface. Right?

Yes

3)At the same location, an angled field, or equivalently an angled sensor, will give a lesser reading. Right?

Yes

4)In the "peanut waist" picture, the field lines are not parallel to the long axis of the magnet stack, but dip in at the waist or "Bloch wall"("arrow" pointing toward magnet body) and come back out again ("arrow" pointing away from magnet body) on the other side of the BW. In the conventional picture, the field lines are parallel to the long axis of the stack and there is no "dipping" in and out at a "Bloch Wall" waist or equator. Right?

Yes

5)So a Hall sensor held with its plane at right angles to the magnet's long axis, and scanned along the length of the magnet, would read very differently in the two cases. Right?

Yes

6)If the "peanut waist" picture is true, the sensor being held at right angles and scanned along the magnet will experience a changing angle of the field as the field dips toward the "Bloch wall" waist, and then an also changing angle as the field dips up out of the equator on the other side. This will cause changing readings as the sensor is scanned past the "Bloch wall equator".  Right?

Yes, moving from N-S the flux density through the sensor will first exhibit a peak that is opposing magnetic polarity from near the north pole on the same side of the magnet, and then moving a little closer to the south pole it should exhibit a second peak in the same magnetic polarity as near the north pole on the same side of the magnet.

7)But in the conventional case, with the field lines strictly parallel to the long axis except near the end poles, the sensor will experience the field lines straight through the plane of the sensor, and thus the sensor's reading will remain constant, and _at the maximum value_  as it is scanned along the magnet's long axis. Right?

No.  Near the ends of the magnet where there is lots of curl, there will be substantial flux, and therefore flux density that is perpendicular to the dipole axis and registered by the senosr.  The sensor should read just as the diagrams I posted indicate.  Near the middle of the magnet where all the flux lines become parallel the perpendicular flux, and therefore flux density is zero.  A plot of the sensor output will be a non-linear, but monotonic line moving along from one end of the magnet to the other.

I have asked seven specific questions, some in the form of testable hypotheses, related to an experiment. Before I proceed further, I would like to hear some opinions about these seven questions. Especially, if a stack of disk magnets is not an acceptable substitute for an actual one-piece bar magnet, please let me know right away, and be sure to tell me why.



p.s. Minnie, particle physics and the CERN research are not a load of crap. But then you knew I'd say that. The hypotheses of the researchers may turn out to be unsupported or even falsified by the data they gather, but that's not crap, it's science.

Of course _some people_ evidently can't tell the difference, even as they sit at their computers designed by people they think are idiots using models they think are crap. It's a good thing they aren't still using CRT monitors... that would be yet another layer of "crap" they have to look past and ignore in order to bolster their pet "theories".

[MarkE]

Well thats telling us what is happening,but not how it's happening. In fact,i would like anyone to show me a link that shows how magnetic fields attract and repell each other-what is the physical force that applies these two forces.

https://www.youtube.com/watch?v=evfUTmx0uh8

The internet features endless loads of misleading crap.  Caveat emptor.

[synchro1]

@NoBull:
Thanks for your response. So you agree with the "rights" on all of the 6 hypotheses. But we still don't know if my stack of disks is "good enough" for the experiment. I'd still like to hear from some of the other participants in the discussion about all 7 questions before I proceed further.

@ramset:  Personally, I think that the stack of disks is fine to use instead of a bar magnet. I can make a weak bar magnet by magnetizing a piece of mild steel barstock, of course, but stronger is probably better for this kind of test. Buying a special magnet from, say, K&J Magnetics would be kind of silly, since the shipping cost alone will be quite a bit more than the cost of the magnet, and there would be several days of delay in getting it here. By next Friday the issue will probably be forgotten already.
http://www.kjmagnetics.com/proddetail.asp?prod=D4Y8

"Other needs for this work"... heh, that's a good one. Some confidence that proper conclusions will be drawn by "the usual suspects" given the observed data, would be nice. 
It all depends on "how fancy" one wants to get. I can spend a day or more programming an Arduino and Processing to display some nice graphs on a computer screen, or I can just show a video of the scanning process and the raw voltage output of the Hall sensor as it is scanned along the magnet. I prefer to do the latter, since it is so simple and, to my thinking, unequivocal in its results.

Allcanadian has already shown data from a similar experiment, with a fancy display and everything. I am not sure from his description about the exact orientation of the Hall sensor and some other important variables. His graphs would make more sense if the x-axis was given in position along the magnet rather than time, though.

@TinselKoala,

"Buying a special magnet from, say, K&J Magnetics would be kind of silly, since the shipping cost alone will be quite a bit more than the cost of the magnet",

You're straight out of a tale from Charles Dickens you miserly thread ball Scrooge! Get a job!

[MarkE]

@TK

Quote

    1)Do we think that the community here will accept an experimental result obtained from a stack of disk magnets, instead of a solid bar magnet? If not, why not?

Already done it, not even comparable. I have the same problem you do and have every magnet under the sun except for a long bar magnet.

So have I and it works well with any dipole that is substantially longer than the Hall sensor area.

Quote

    2)The Hall sensor of course will read its maximum value for a given field strength in a given location if the "field lines" are exactly perpendicular to the sensor surface. Right?

I think of it as a sensor loop, it measures the magnitude and polarity of the magnetic field within the loop. Try rotating the sensor in the field as you will see what I mean.

It measures the average flux density of flux that penetrates the sensor area.

Quote

    3)At the same location, an angled field, or equivalently an angled sensor, will give a lesser reading. Right?

Exactly and even a slight off angle makes a huge difference. It is actually very hard to keep the sensor stable when amplifying the signal.

The Allegro parts seem to have little difficulty with this.

Quote

    4)In the "peanut waist" picture, the field lines are not parallel to the long axis of the magnet stack, but dip in at the waist or "Bloch wall"("arrow" pointing toward magnet body) and come back out again ("arrow" pointing away from magnet body) on the other side of the BW. In the conventional picture, the field lines are parallel to the long axis of the stack and there is no "dipping" in and out at a "Bloch Wall" waist or equator. Right?

I would agree the picture is incomplete and every time I fire up the sensor I generally learn something new.

Why the non-responsive reply?  TK asks what you believe the hypothesis predicts.

Quote

    5)So a Hall sensor held with its plane at right angles to the magnet's long axis, and scanned along the length of the magnet, would read very differently in the two cases. Right?

As it should I believe and a rotating sensor or 4 sensors at 90 deg to each other may be the answer but trying to plot this in 3D would be a nightmare.

The simple yes was adequate.

Quote

    6)If the "peanut waist" picture is true, the sensor being held at right angles and scanned along the magnet will experience a changing angle of the field as the field dips toward the "Bloch wall" waist, and then an also changing angle as the field dips up out of the equator on the other side. This will cause changing readings as the sensor is scanned past the "Bloch wall equator".  Right?

That's where a gyro/accelerometer would be awesome I think, I have a few on hand and have been throwing around code on how to plot a moving/rotating sensor .
Quote

Again, why the non-responsive reply?  TK asks you to make a prediction based on the hypothesis.  If a hypothesis cannot make testable predictions, then that hypothesis has little value.

    7)But in the conventional case, with the field lines strictly parallel to the long axis except near the end poles, the sensor will experience the field lines straight through the plane of the sensor, and thus the sensor's reading will remain constant, and _at the maximum value_  as it is scanned along the magnet's long axis. Right?

That is exactly the issue I had and it was constant, the far field would appear constant however the near field not so much. I would think a change in sensor orientation would show an increasing/decreasing field strength dependent on the polarity at that point so long as the sensor rotates about it's own axis. As I said, even a small deviation from the sensor axis ie rotation changes the measure dramatically.

Really?  You measured maximum flux density with the sensor face perpendicular to the magnet's axis when located at the center of the magnet?  I would like to see a picture.  You show me yours and I will happily post mine.

I'm kind of pumped your going to do this and I will be doing my thing on this end. Hopefully Sunday I can start on the gyro/accelerometer code but the processing is going to be the kicker. I'm not even sure how in the hell I'm going to plot this in 3D, lol.


AC