[time-nuts] Those pesky Neutrinos again...
magnus at rubidium.dyndns.org
Mon Nov 21 15:30:14 EST 2011
On 11/21/2011 09:49 AM, Hal Murray wrote:
> michael.cook at sfr.fr said:
>> I heard on the BBC the other day that a repeat experiment is planned,
>> firing neurinos from the US into Canada. The labs were not cited, but I
>> expect it would be Fermilab to Sudbury Ontario. If this is the case, then
>> there will still be the problem of not being able to run a parallel fibre
>> as the Sudbury detectors are also deep underground. ...
> I don't think that running a parallel fiber will help much. I see two
> First, fibers don't run in straight lines. Even if you could accurately
> survey the pipes in the ground, there is slack in the fiber within the cable
> bundle so it doesn't break if somebody pulls on the cable.
> Second, the speed of light in fibers isn't known to the required level of
> accuracy. (It's probably temperature dependent.)
It is. About 10% of the change is due to length-changes in the fibre and
about 90% of the change is due to group-delay shifts in the fibre.
If you laser frequency (wavelength is the traditional value here) shifts
then the dispersion effect also causes shift in delay.
> In the context of fibers, having the detectors located deep underground in
> not a problem. They have to get power and data cables down to the
> instruments somehow. It would be easy to run a fiber in parallel with those
> There is also the delay in the amplifiers that you will need every 100 miles
> or so. Or the low rise time if you try to avoid the amplifiers because you
> don't need much bandwidth...
Depends on the amplifiers you are using. EDFA amps has fairly high
bandwidth and delay is like normal fibre of the same length.
More information about the time-nuts