[time-nuts] Rb Mag sensitivity

[Bruce Griffiths]

Bob Camp wrote:
> Hi
>
> I finally found a number listed for the external magnetic field sensitivity of a fairly small Efratom rubidium. It's specified as 3 x10^-11 per 0.1 mili tesla.  I did *not* make up those units, that's the way they stated it. If that's correct *and* my very rusty physics is also right:
>
> A 1 A change in a wire one inch away will produce  0.08 mili tesla of field change. That should produce  2.36x10^-11 frequency change. Obviously it's not quite that simple since the wire is a lot closer to the device than the device is large. I suppose a more proper measure would be 1" away from the physics package. Even then the field is not uniform over the entire physics package.
>
> The reasonable conclusion would seem be that amp or multi amp conductors at an inch or more from the package might indeed be an issue at the 1x10^-11 level. At the very least twisted paris for DC in the area sound like a good idea as opposed to random wire routing. It also sounds like a good idea to simply not have that sort of current running right by the physics package.
>
> Ok, what did I miss that time. Hand calculating DC fields is something I haven't done in ... ummm,... errr ... let's leave it at "quite a while".
>
> Bob
> _______________________________________________
>    
Yes, its about 80uT at 1" radial distance from an infinite length of 
straight conductor carrying 1A.
B~ 2E-7(I/r)

Smaller for finite wire lengths (but in this case one has to consider 
the geometry of the other conductors forming the loop).

Since the earths field lies in the 30uT to 60uT range, one may be able 
to use the Rubidium standard as an insensitive compass.

Bruce