[time-nuts] Optical link connects atomic clocks over 1400 km of fibre

[jimlux]

On 8/25/16 8:12 PM, Bill Metzenthen wrote:
> On 26/08/16 08:14, Hal Murray wrote:
>> billm713 at gmail.com said:
>>> If the conductor also has magnetic properties (e.g. if iron were
>>> used)  then
>>> magnetic saturation could be an issue.
>> Ah...  Sorry I wasn't clear.  How about
>>
>> Is skin depth an interesting concept if you are using materials
>> commonly used
>> for magnetic shielding, for example mu-metal?
>>
>> The electrical conduction is poor so the normal skin depth reasoning
>> probably
>> won't be useful in practical examples.
>>
>>
> For skin depth, the lower conductivity of mu-metal is more than
> compensated for by its higher permeability.  At 50 Hz its skin depth is
> calculated to be around 0.3 mm assuming that the conductivity and
> permeability are real (as distinct from complex) and everything is
> linear.  I haven't needed to consider what happens when these
> assumptions aren't valid so I could only guess how rapidly the skin
> depth increases when a magnetic material such as mu-metal is pushed past
> its linear region.
>
> The usual skin depth calculation is based upon assumptions about the
> form of the electromagnetic field and the conductor, which might not
> apply to the situation in which one is interested. Despite this, it is
> often used to get an estimate which is useful in a range of applications.
>

it's all about sqrt( rho/mu)... good conductivity and high mu are what 
you want.  soft iron works well... heck, steel works well for a lot of 
applications.

Mu metal (which as others have noted has all sorts of handling and 
fabrication issues) is for the more exotic cases..

But, could you not run a differential pair of fibers, and both would be 
affected, but with opposite signs....