[ntp:questions] So what? (Re: 500ppm - is it too small?)

unruh unruh at physics.ubc.ca
Wed Aug 19 18:24:30 UTC 2009


Nero Imhard <nim at pipe.nl> writes:

...
>wrt requiring rigorous argumentation for this precise value of 500 ppm:
>to me it seems that it is a tradeoff between brokenness of clocks and
>the desirability to set the limit as low as possible. So *any* limit is

Why is it desireable to set the limit as low as possible? There seems no 
justification for this claim. While a limit on how big a slew rate one wants may be desireable (I would think that getting the time error down to the 
lowest possible value would be more desireable), a limit less than about 1%
(100000PPM) on the compensation of the drift rate seems to me to have little
justification. IF the drift rate fluctuates by say 500PPM or even 10PPM that is
definitely bad. But ntp compensating for a constant drift rate, no matter
how large,  seems perfectly 
fine to me. 
>at least partly depending on what frequency error one is willing to
>accept, hence somewhat aribrary. Yeah. So what?

I want 0 frequency error. That is why I want ntp to be able to compensate
for large drift rates on the clock. 


>N




More information about the questions mailing list