[time-nuts] Positional accuracy of the M12+T

Magnus Danielson cfmd at bredband.net
Thu Jan 4 12:36:37 EST 2007


From: "Tom Van Baak" <tvb at leapsecond.com>
Subject: Re: [time-nuts] Positional accuracy of the M12+T
Date: Thu, 4 Jan 2007 09:12:30 -0800
Message-ID: <005a01c73024$5c99fce0$d28a8843 at computer>

Tom,

> [ sorry, let me try that again ]
> 
> > Hmmm... with 25m position accuracy (and 100ns is
> > about 30m), how do they really get time down to a
> > few ns. Clever engineering!  :-)
> 
> You're comparing apples and oranges: accuracy and
> stability.
> 
> The 25 m value you often read is probably a valid
> figure for the positional accuracy of an M12.
> 
> But do not confuse this with the "few ns" timing value
> you read about here. No one I know has an M12 that
> is accurate to a few ns. What many of us have are
> M12 that exhibit residual timing jitter of a few ns, when
> properly sawtooth corrected and averaged over time
> compared to a local cesium standard.
> 
> In other words, a properly filtered M12 is *stable* to a
> few ns over some time frame. But it's highly unlikely it
> is *accurate* to a few ns. A couple of ten ns is much
> more likely. And this is in the same ballpark as the
> "25 m" positional value.

Yes¸ let's not confuse our statistical deviation of samples with the accuracy
in their mean. Good point there.

Cheers,
Magnus



More information about the time-nuts mailing list