Re: ✘PX1172RH_DS performance

[Greg Troxel]

"Gary E. Miller" <gem@rellim.com> writes:

> The part includes 2 RTK GNSS and you connect two GNSS antennas to it.
> You take RTCM3 from a "fixed base" and feed it to this parts "moving
> base".  Then the part computes distance and angels to its built in
> "rover".

How many angels fit on a GPS antennas?

But seriously, I'm not sure I am following.

I see the point of two antennas so you can have them at some fixed
separation on a vehicle and obtain orientation.  And I can see the point
of RTK from a base/network to get coordinates for the primary antenna
location.  (Then you can also get position for the "azimuth mark" (2nd
antenna) as it might be called for passive control.)  So questions:

  What are you using for fixed base?  An F9P locally with a third
  antenna?  ORGN?  What's the rough baseline length?

  I think your plot is for position from the primary antenna, and you
  are not showing data from the secondary antenna.  Is that right?

  Can it do azimuth without any incoming RTCM3?  Basically use one as a
  moving base, with position determined from navigation solution, and
  compute the vector between them, and have precise oritentation even if
  position is fuzzy at the 5m level.

> See attached plot.
>
> Actual 2D error:
>
>     CEP(50) = 0.007 m
>     CEP(95) = 0.012 m
>     CEP(99) = 0.014 m
>
> Actual Alt error:
>
>     EP(50) = 0.007 m
>     EP(95) = 0.022 m
>     EP(99) = 0.030 m

It is not really error, but a meaure of variance.   I am 99% guessing
that you don't have a mm-level position for the APC, and thus these are
the values relative to the average position.  One could, for a fixed
roof antenna, do

  raw observations for 24h and use PPP (both?)
  raw observations for 24h (same bits) and use OPUS
  Use F9P with ORGN baes
  Use PX1172RH_DS with ORGN baes

and compare them all.

> But this perforrmance is not easy to get.  Not just any antennas are
> good enough.  I had to sort for my 2 best L1/L2 antennas.  Even then a
> good skyview was not sufficient.  I only acheived RTKFIX by putting the
> antennas on top of my roof.

I find that getting into FIX mode depends a lot on sky view.  And, one's
base having more constellations helps make it work better.  With a 20
degree treelne at the horizon it is rapidly and solidly in FIX.  On
trails in the woods, iffy but FIX can be achieved at times.  There is a
new firmware for F9P that I need to update to and it is rumored to be a
lot bettter.

> Figuring out how to test the "compass" part will take some work.

I guess that answers my question above - you did what I would have done
first.

signature.asc
Description: PGP signature