[time-nuts] Precise manual survey (was Common sky pps...)

SAIDJACK at aol.com SAIDJACK at aol.com
Wed Jan 7 04:05:20 EST 2009


Hello Peter,
 
that's a good question. Some users of our Fury use rented/borrowed survey  
equipment (Leica, Javad, etc) to get a very precise position of their antenna,  
or they call in a surveyor.
 
It is also possible to let the receiver do the auto survey multiple  times, 
preferably during different times of the day, then manually average the  
position fixes, and enter them into our Fury unit as the new position-hold  
position. I described this method some time ago. Position error should be  reduced by 
this type of averaging.
 
One could also use the Fury NMEA GPGGA output command to send NMEA position  
data to a PC over a long time, say a week or more, then crunch all that data  
down to one averaged position fix and manually enter it into the Fury.
 
And your idea of trial and error is not bad either, but I would think  height 
errors of <10 feet, and position errors of less than 5 feet would be  
sufficient to try. This requires a good reference to compare against though, and  
LOT'S of patience I would think.
 
>From what I can see, all commercial GPS receivers have a really tough time  
generating an accurate antenna height measurement due to the way GPS works. 
They  don't seem to have a problem generating good long/lat coordinates. But in  
timing, all three axis need to be precise unfortunately.
 
bye,
Said
 
 
In a message dated 1/6/2009 03:23:53 Pacific Standard Time,  
pvince at theiet.org writes:

Said  wrote:

>When we say units typically have 25ns unit-to-unit variation  on the 1PPS on 
 
>un-calibrated units, then I believe most of this  is caused by the 
auto-survey  
>position errors of the GPS  receiver. One could get much better performance 
by 
> manually entering  the exact position-hold position of the antenna, and 
then   
>calibrating for antenna cable delay (in 1ns steps).

A naive  question, if I may: how do I go about doing a precise manual 
position survey?  Is it simply a case of letting the unit self 
survey, then manually  entering different co-ordinates (say +/- 10 metres) in 
each of the three  directions in turn, watch the 
signal output, and try to tune for minimum  wobble?

Thanks,   Peter





More information about the time-nuts mailing list