Re: ✘NTRIP and RTCM

[Gary E. Miller]

Yo Greg!

On Fri, 03 Dec 2021 15:39:19 -0500
Greg Troxel <gdt@lexort.com> wrote:

> "Gary E. Miller" <gem@rellim.com> writes:
> 
> >> > All from the Oregon ORGN network.  Many thanks to them for
> >> > allowing me cess.    
> >> 
> >> Indeed, it's very nice of the states to allow public access.
> >> MassDOT runs MaCORS which anyone can use.  
> >
> > Still not as easy as many other countries.  
> 
> Yes, I guess they could run a caster that does not require login.

And a lot of the USA has no nearby casters.  Login or not, nothing
available.

> >> I am fuzzy but I gather that RTCM2.3 typically is pseudorange
> >> corrections, and GPS only.  It may be L1 only.  
> >
> > Yes, in this case, GPS and L1 only.  
> 
> So you lose 3 of 4 constellations, and you lose the abiltiy to
> estimate iono delay from L1/L2.   Interesting that it makes things
> worse, but I am not super surprised.

ONe would hope the firmware would recognize the degradation and
stop using the RTCM2.3.  Just like it stops using any satellite
that provides bad data.

> >> Do you have SBAS disabled, so it is really "no corrections".  
> >
> > I do not have SBAS disabled, but that F9P rarely uses it.  SNAS is
> > almost exactly RTCM 2.3  
> 
> So no L2C, no GLONASS even?  I suppose someone(tm) should get the SBAS
> stream and decode it.

Thera are so few L2C GPS satellites, not much use.  RTCM 2.3 does support
GLONASS, but not available to me nearby.

> >> Indeed, and you can see the resolution of the coordinates which
> >> looks like most of 1 cm.  I wonder if this quantization is in the
> >> raw logs or if it could be in the analysis chain.  
> >
> > I'm running in normal mode, with none of the "High Precision"
> > messages. Moany of the UBX-NAV-PVT data items are in mm.  Worth a
> > test.  
> 
> I've been using high-precision NMEA, because my real goal is to log
> data in Vespucci for mapping/surveying, and I think I'm seeing well
> sub cm quantization.

What is "high-precision NMEA"?  NMEA does not specify any resolutions.

u-blox HP mode gets to 0.1 mm.  UBX-NAV-PVT (default) is 1 mm.

I'll recheck gpsd JSON path to confirm it can hadle 0.1mm.

> >> What I would find really useful is:
> >> 
> >>   record fix mode with each point as metadata  
> >
> > I have all the raw data.  Nothing more can, or could, be recorded.
> > And it is all 3D.  
> 
> Great, so you should be able to associate with each position what kind
> of fix it is.

They are all 3D differential.  Nothing more to say.  u-blox does not
report RTK float of RTK fix.  AFAIK they never really meant anything.

> >>   carry that through analysis and color the dots based on
> >>   autonomous/differential/RTK-fixed/RTK-float  
> >
> > u-blox is bad at reporting that.  
> 
> Do you mean there are no messages that say if the solution is from RTK
> FLOAT, RTK FIX, just nav satellites, differential etc

Correct.

> like you get in
> the mode field of the main NMEA message?

Uh, no.  The NMEA mode field can also report RTK fix, RTK Float, and
other variants.  Badly defined and inconsistently implemented.

> >>   be able to use an input position in the analysis instead of the
> >>   average.  I bet your tight value in fix is pretty much right.  
> >
> > Lost me.  Explain?  
> 
> I think the analysis is computing an average over the hour, and then
> the CEP is the distance such that half the points are closer to the
> average.  But that's not really "error" because you don't know the
> right answer.

I could go to a geodetic benchmrk, and repeat.  But even they are not
accurate to 0.1mm.

> I am suggesting, more concretely:
> 
>   have the analysis program output the computed average in geojson or
>   something into a file, maybe optionally

gpsprof does that.

>   have an optional argument to read a "correct answer" from a file in
>   that same format

If I could get a "correct answer", but I have none.  So not useful.  And
the math to add a constant error is trivial, if I knew what that was.

>   when given a position this way, compute in terms of that position
>   being the truth, rather than assuming that the average of *this
>   session* is correct.

See above.

> What you would do is use the output from the MAX which appears to be
> RTK FIX, to process the rest.

It is the only output I have, so I dont care what it appears to be.

>  That would show you biases among them,
> even if you aren't still really sure whihc is right.

Lost me again.  What "them".


> >> I am curious what antenna  you are using.  
> >
> > The u-blox one, in a window sill.  
> 
> Do you mean the slightly larger mouse-style magmount one that is about
> $60?  I have one and need to try it.

They onlye sell one:

https://www.u-blox.com/en/product/ann-mb1-antenna

I find it works well.

> >> What I don't understand is why you aren't able to get into FIX
> >> with a single reference a mile away.  
> >
> > I don't know if I can or can't.  But that is my long term
> > experience, and other see simiilar.  
> 
> I would love to see data with something like the survey or calibrated
> from this line.  I am guessing you have the MB.
>   https://www.ardusimple.com/simpleant2b/

Unclear what models they are selling.  I have one of the mushrooms.  It
has a 40db LNA, so good for feeding a power splitter.  I have not tried
it with the F9P.

> So it would be really interesting to see data from your receiver/ant
> with the antenna fully outside in a repeatable postion.

Givne the data cluster at 1cm, I would prolly not be able to see any
improvement.

> I have a
> particular deck railing nail I place the stand over.

Ditto.  Now if any 2 GPS ever gave me the same position of that nail...

I am running into one big problem.  When I use the iMAX data, the
F9P CPU gets overloaded.  It stopps sending all but the most basic
messages, the txbuf is at 99 to 100% full:

UBX-MON-TXBUF:
 pending   0 12031 0 0 0 0
 usage     0 99 0 0 0 0
 peakUsage 0 99 0 11 0 0
 tUsage 99 tPeakUsage 100 errors xc0 reserved1 0
  errors (mem alloc) limit (0)

and constantly sending error messages:

$GNTXT,01,01,00,txbuf alloc*61

So the extra precision comes at great cost to other available data.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        gem@rellim.com  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin

pgpik4E38s5h5.pgp
Description: OpenPGP digital signature