[LinuxPPS] Experience with current linuxPPS kernel patch.

[Hal V. Engel]

On Thursday 04 March 2010 03:51:38 am Miroslav Lichvar wrote:
> On Wed, Mar 03, 2010 at 08:03:55PM -0700, clemens at dwf.com wrote:
> > I still have the feeling that something is not quite right (but mabe
> > elsewhere in
> > the kernel code) since the accuracy is just about 1us.  That is, the
> > 'fuzz' of the timestamps around the mean is just about 1us.  I would
> > expect a factor of x10 better than that, and would not be surprised if
> > there is not still some bad ns <-> us conversions in the kernel.
> 
> What is your PPS source?

Since this was in reply to Reg's note I will venture a guess since I have a 
pretty good idea what Reg is using.  Reg is the maintainer of the Oncore 
driver for NTP and I think he is running Oncore VP and M12 timing GPS units.  
The Motorola Oncore timing receivers are considered the gold standard and are 
very accurate.

The older Oncore timing receivers like the VP, UT and UT+ had a PPS that was 
spec'ed at +-50 nanoseconds with SA.  Since SA has been turned off they should 
do significantly better.  But no specifications are available for these older 
units without SA.   Current Oncore timing receivers like the M12T are even 
more accurate.

In addition the Oncore units provide PPS saw tooth data to the driver and I 
think the current NTP Oncore driver uses this data to correct for saw tooth.  
The latest M12 Timing units are supported to be accurate to +-1 nanosecond if 
the sawtooth correction is applied.   Of course this assumes a near perfect 
installation with a very accurate position used for the antenna and other 
things like cable delays and such being correctly configured. 

Saw tooth error for the UT+ Oncores is about +-40 nanosecons and I believe 
that for the M12T units it is on the order of +-25 nanoseconds.

> 
> I'm using GPS 18x LVC and I'm seeing about 1us dispersion too. But the
> specs say it's accurate only to 1 us, so I'm not sure if I can expect
> anything better from it.
> 

Probably not.  Even those of us using specialized timing receivers like the 
Oncore units will never see dispersions any where near the limits of our 
receivers specs. because of factors downstream from the receiver that 
introduce much larger levels of dispersion.  These include variations in the 
response time of the interrupt handler and other factors that probably limit 
the minimum dispersion to perhaps 100 nanoseconds even with a +-1 nanosecond 
receiver.  In your case you are probably limited in this regard by the 
receiver since its PPS timing spec. is big enough to be a major factor.  In 
Reg's case most of the dispersion he is seeing is probably down stream from 
the GPS.

Hal