[FOSS-GPS] help with RTKPOST of GPS file please

[David Kelley]

ebennb:
I finally got some time to play with this data today...   I found a 
fixing rate of 23.9% with a continuous AR setting and 2.9 threshold, 
elevation angle of 15 deg, reset set to defaults.  More or less the same 
as what Michele got, perhaps the additional AR fixed % is just due to 
minor setting changes.  I see a very odd think to report and ask others 
about; looking at the debug traces there is a constant stream of cycle 
slips on anywhere from five to all of the SVs, it never settled in this 
respect.  Even with the somewhat low Cno values, there should not be 
this constant loss of lock.   Also, when the platform is on the ground 
and getting a smattering fixed points, it is still shifting by a couple 
of meters, so unless you were hovering over it (blocking the signals), 
this is quite suspect. Finally, in the air, on a couple of the loops, as 
the platform moves its heading from directly north to slightly west, 
there is a noticeable (i.e. repeatable) jump. It is probably traceable 
to part of the airframe blocking a key SV each time.  I remain 
suspicious of the antenna used, and kind of suspect your can double your 
rates with a better on.  Was this a small patch antenna and was there 
any sort of ground plane?

When you do the next "warm up" try to keep anyone a good 10 meters away 
from it and then we can process that segment using other method to 
stationary observation and perhaps learn more about this odd constant 
cycle slipping.


On 4/17/2016 11:30 PM, ebennb wrote:
> Very informative! I've included a number of photos now of the GPS unit (it is
> indeed the LEA-6T) as PrecisionHawk as it integrated into their UAV. The
> photos are under multimedia in the dropbox folder, or a direct link below:
>
> Full folder:
> https://www.dropbox.com/sh/o7y4p7bplnocvqu/AACVBsxV87n531fstkEEUTvOa?dl=0
> Photos only:
> https://www.dropbox.com/sh/4qkpjrikqi44fcy/AADhSBN4sW47xf6LD3zB7qEBa?dl=0
>
> I modeled parameters on your responses and found I was able to get much
> better solutions than before asking for help.
>
> I will try next another trial flight with a 15 minute GPS warm up period
> prior to the flight, perhaps that will boost the number of Q=1 solutions.
> Thanks!
>

-- 
Regards,
David Kelley