Replication of the Jan P. Cack device?

[CLaNZeR]

Quick update here. The magnets I wanted from Germany never turned up as they did not even bother to reply to my email, so used some rod magnets that I do not think are long enough.

But anyway, picture of the finished rig, excuse the Glue Gun attachment job, but it does the job of holding them in place.

Anyway it did not self spin, but fun making.



Cheers

Sean.

[Omnibus]

Sean, your work seems impeccable mechanically. However, no need to mention, what?s at play here are the magnetic fields and that?s been a problem with all these devices. Thus, I don?t believe it?s only the length of the rod magnets that?s problematic here. Do you have or could you possibly borrow a gauss meter to have these fields measured? You may place the device on the spindle of a milling machine, measure the induction at small increments and then plot it so that a quantitative picture is obtained of these fields. You may do this separately for the planets on the one hand and the sun on the other. Also, it may be worth trying magnets other than neo?s as well as non-magnetized steel pieces of the proper shape as the parts making the sun. Probably you may try using steel pieces of the right size for the stator (respectively, the rotor) in your Torbay replication.

[Low-Q]

Also just noticed this link on the same site as well. Provides for a better visual.

http://www.fdp.nu/cack_movie/JanPCackFemm.htm


Mikey

The problems with FEMM simulation is that if you "house" the magnetic device inside a small circle, as the animation shows, you don't have the correct magnetic event to calculate the correct torque - no matter how high the resolution you have or how many samples you measures in one revolution. The magnetic fields which is facing the outer circle (the "housing") are not all taken into account, and they are simply "absorbed" so the magnets closest to this circle is not symetrical. As we all know, all magnets are in real life allways symmetrical in the way they display their magnetic fields. In case of a FEMM simulation this is not allways true, hence you sometimes get positive or negative (Other than zero) torque.

In practice, when the device is made, it will therefor not work.

If you get this FEMM simulation from the inventor, and modify the housing outside it to a much bigger one, you'll probably see that the total torque in the device is approaching a zero value.

I have myself a small magnetmotor simulation that provides 7Nm in average. Very exiting at the first, but when playing with the allowed space to work in (The housing), this torque alters to more or less as I decrease or increase the housing around it.

Br.

Vidar

[mikey]

Quick update here. The magnets I wanted from Germany never turned up as they did not even bother to reply to my email, so used some rod magnets that I do not think are long enough.

But anyway, picture of the finished rig, excuse the Glue Gun attachment job, but it does the job of holding them in place.

Anyway it did not self spin, but fun making.



Cheers

Sean.

Hi Sean,

It's a shame it didn't work, especially after your superb craftsmanship!! You deserved better!! But at least this idea has definitively been ruled out now, which is a shame.

Best Regards,

Mike

[Omnibus]

But at least this idea has definitively been ruled out now, which is a shame.

The above statement is premature. This idea hasn't been ruled out yet. Now that the groundwork has been laid, it needs further exploration.