I am a Physics student (4th year) and I'm trying to study the Einstein - de Haas effect in laboratory. That is what I got: a suspended Iron cylinder with about 5 cm height and a radius of 0.8 cm is put inside a solenoid which can create a magnetic field of 0.2 mT; we use a Titanium fiber of 0.05 mm radius to hold up the cylinder.
Now where is the problem? Well we expect that when we turn on the current and so create the magnetic field through the solenoid, the cylinder would rotate just a bit, for example 2 degrees because that is what is written in the Einstein - de Haas article... but not! We got rotation of 70° and more (to measure such a great angle we used a camera which records the piece while rotating)! Measuring the initial angular velocity with a computer, and substituting it in the formula:
$$ I \omega = -(2m/e) M $$
Where $I$ is the moment of inertia, $\omega$ is the angular velocity and $M$ is the magnetization of our material (which we can obtain), we get a value so far from $(2m/e)\simeq10^{-11}rad / s \cdot T$ in fact we got something like $10^{-5}$. I know that we should obtain something like the double of $(2m/e)$ but we are neither close to it!
In conclusion: should we force ourselves to have an angle of $1°\sim5°$ and then (and only then) measure the angular velocity because this effect is measurable only with little oscillation, or we got something strange, which means that we should change the system because something is interfering? Obviously we have isolated as we can the system from air currents and ground vibrations.
If you need more information about what we are using in the experiment, just say it and I will provide them. Thanks!
New information (about the comment of HolgerFielder): the wire diameter I've already given: Titanium 0.1 mm; if you mean the diameter of the solenoid's wire, that is 2 mm almost. Coil diameter = 4 cm. Number of loops = 200. Current $\simeq$ 1.5 Ampère. From these values and using this formula for magnetic field of a solenoid with radius r and with length l (in its "central axis"):
$$ B = \mu _{0} (N/l) i \cdot l/(\sqrt(l^{2} + 4 r^{2})) $$
one can obtain $B \simeq 2.2$ mT which is close to the value that we measure with the Gaussmeter (0.2 mT as I mentioned before, in the top of the question).
