Magnet Myths and Misconceptions

[MileHigh]

Tinman:

Do you agree with the magnetic field around a wire as shown in the attached picture?

If you do then look at the second picture with a coil.  The coil is just a straight wire bent into a shape.  Just do the right-hand-rule visualization of the magnetic field generated for each loop and combine it with the vector addition for each loop and you should see how it all comes together.  The field around the coil comes directly from the field around a straight wire.

MileHigh

[MileHigh]

After his test-with a hall sensor as suggested many times by the big guns here-turns out the field was shaped like a peanut-whats the chances of that.

No way, those tests were inconclusive.

What you would have to do with a Hall sensor would be to find the actual direction of the magnetic field around the magnet.  Let's say to keep it simple you place the magnet on your bench table and just do tests at a series of points on the surface of the table located around the magnet.  At every position for the Hall sensor, you would have to rotate the sensor around two axes (pitch and yaw, see graphic) and find the position where the magnetic field is the strongest.  If your setup is nice and symmetrical you will see that the magnetic field only varies on the yaw axis as you go around the magnet.  The angle that is normal to the surface of the Hall sensor is the direction of the magnetic field.

Do that and you will get the standard pattern for a magnetic field outside a coil or a bar magnet.

Then just put the Hall sensor in the air above the magnet and repeat the process and you will be able to see where the magnetic field is pointed anywhere above the magnet.

I saw how AC's Hall sensor software gives him a gain function for so I am assuming that this test is quite feasible.

MileHigh

[MarkE]

Up until now, I have seen no evidence that supports a looped field. I have seen far more that supports my figure 8/peanut shaped field. So far I have seen the results of one test here, and done by some one that is pretty handy with electronic equipment. This some one also said before his test, that I was not seeing what others were trying to show me-the looped field. After his test-with a hall sensor as suggested many times by the big guns here-turns out the field was shaped like a peanut-whats the chances of that.
But we know the onslaught of why the test failed to show correct results is just around the corner. This is when the big guns get to explain why the test they recomended didnt show a looped field from one end to the other.

This thread is a hoot
TK-where you at ?

Tinman you never answered me whether you think this representation is fair:

[MarkE]

Here is an experiment that anyone interested in this topic can perform at home for about $5.00.

Parts required:  Allegro Microsystems linear Hall effect sensor A1324LUA-T  http://www.digikey.com/product-detail/en/A1324LUA-T/620-1432-ND/2728144
5V bench power supply or
DIY 5V supply:
  9V battery
  Battery clip
  LM78L05
  10uF input and output capacitors
.1uF capacitor for the A1324, solder directly across pin 1 and 2.
fine lead wire 24AWG or smaller
heat shrink

DMM

Bar magnet, or stack of cylindrical magnets. 

Solder a 12" power lead to the A1324 pin 1.  Cover the exposed metal with heat shrink tubing.
Solder a 12" power return lead to pin 2.  Cover the exposed metal with heat shrink tubing.

Solder a 12" signal lead to the A1324 pin 3.    Cover the exposed metal with heat shrink tubing.
Solder a 12" signal return lead to pin 2.  Cover the exposed metal with heat shrink tubing.

Twist the power and power return leads together with at least two twists per inch.

Twist the signal and signal return leads together with at least two twists per inch.

Take the 16 measurements shown in the attached drawing.  Compare your results with those in the drawing.  Take additional measurements along the magnet as desired.

For added fun make the magnet longer by stacking additional magnets together.  What do these results say about flux supposedly returning to the middle of the magnet?  What do they say about the idea that near the middle of the magnet domains change direction?

[NoBull]

After his test-with a hall sensor as suggested many times by the big guns here-turns out the field was shaped like a peanut-whats the chances of that.

The chance is 1:3.

This is because there are 3 common ways of scanning the magnetic field of a cylindrical bar magnet with a single Hall sensor.

1) Scanning parallell to the magnet's major symmetry axis (B_X)
2) Scanning perpendicular to the magnet's major symmetry axis (B_Y)
3) Scanning perpendicular to the magnetic field (B_B)

Only the method #3 yields the standard shape of the B_B field around the magnet.
Using method #1 yields the peanut shape.

Scanning with the sensor oriented in a constant orientation to the magnet is easy in methods #1 and #2.
But keeping the sensor aligned with the direction of the magnetic field requires more effort in method #3.

MileHigh vividly describes how to do this:

At every position for the Hall sensor, you would have to rotate the sensor around two axes (pitch and yaw, see graphic) and find the position where the magnetic field is the strongest.  If your setup is nice and symmetrical you will see that the magnetic field only varies on the yaw axis as you go around the magnet.  The angle that is normal to the surface of the Hall sensor is the direction of the magnetic field.

Do that and you will get the standard pattern for a magnetic field outside a coil or a bar magnet.

...and using an attraction of a small soft ferromagnetic object to scan the space surrounding a bar magnet, maps out the gradient of its field (not its density!). 
Such mapping can also be done in these 3 directions, as with the Hall sensor, yielding 3 additional methods of measuring the gradient of field.

So in total, we can have 6 common ways of scanning the field of the bar magnet, each showing a different shape.
Most people don't even distinguish between these different methods of mapping out the field thus it is no wonder this thread is such a "hoot".

This thread is a hoot