[ntp:questions] Recalibration for MSF move

PeterW peterw1_no+spam at f2s.com
Sun Apr 8 15:19:45 UTC 2007


"Richard B. gilbert" <rgilbert88 at comcast.net> wrote in
news:4618EC49.9060803 at comcast.net: 

> Danny Mayer wrote:
>> PeterW wrote:
>> 
>>>The UK MSF time-standard transmitter moved from near Rugby in
>>>Warwickshire to Anthorn in Cumbria at the beginning of April. For
>>>those with MSF reference clocks, this will make a difference of at
>>>worst case +/- 1msec depending on how distance from the transmitter
>>>changes. 
>>>
>>>I assume (hope) that most Stratum 1 clocks have been adjusted, but
>>>some may no have been, so check an MSF-based clock against GPS or
>>>DCF77a before relying on it.
>>>
>> 
>> 
>> Peter,
>> 
>> Do you have any reason to think that it is less reliable or accurate
>> than the old site?
>> 
>> Danny
> 
> 
> I think he means that the propagation delays have changed because of
> the 
>   move and that there may be problems with servers that have not made 
> the necessary adjustments.  I'm inclined to suspect that he's right
> and that many servers have failed to note the change and make the
> necessary adjustments.
> 
> OTOH, I can think of MAYBE ONE server that I can rely on for time 
> accurate to within a millisecond.  They all may have time accurate to 
> the microsecond but the problem is getting delivery over the network.
> 
> 
> 

It is the propagation delay I am talking of. It is just over 1 ms (in old 
money, 186 miles = 1 ms and the new site is about 200 miles away). The site 
is still locked to the NPL clocks in Teddington.

I had noticed one MSF site about 1ms adrift from GPS, but it seems OK now.

I now have a remote GPS S1 reference and MSF S1 reference in agreement to 
about 200us here.

Peter




More information about the questions mailing list