Re: ✘ NEO-M9N

[Gary E. Miller]

Yo Greg!

On Thu, 20 Feb 2020 20:15:56 -0500
Greg Troxel <address@hidden> wrote:

> You say rtcm3, and I'm guessing that means the carrier phase
> differential information.

The older I get, the less I try to guess.

Best to think of RTCM2 and RTCM3 like NMEA.  They describe the message
structure, but say zero about what messages may or may not be in a
stream.

Samples of the local RTCM2 and RTCM3 streams that I use are in the
regressions:
    test/daemon/rtcm2.bend.log
    test/daemon/rtcm3.bndm.log

The matching .chk files show you what is in them.

The RTCM2 stream contains types 1, 3, 18, 19, 22, 23, 24

1  - differential GPS corrections
3  - reference station parameters
18 - RTK uncorrected carrier phases
19 - RTK uncorrected psuedoranges
22 - Extended reference station parameters
23 - Type of Antenna
24 - Reference station ARP

So the RTCM2 I get contains both carrier phase and psuedorange corrections.
RTCM2 can include GLONASS, but the feed I get does not.

Carrier Phase is the raw measured data, and the Psuedorange is the
same data lightly processed.  "Easy" to convert from one to other with
the ephmeris at hand.

The RTCM3 is similar, it just supports more message types, with some more
and more precise, data.

The RTCM3 stream I use contains types: 1004, 1006, 1008, 1012, 1013, 1029,
1033, and 1230.

1004 - GPS Extended RTK, L1 & L2
1006 - Stationary Antenna Reference Point
1008 - Antenna Description & Serial Number
1012 - GLONASS Extended RTK, L1 & L2
1013 - System Parameters
1029 - Text in UTF8 format
1033 - Receiver and Antenna Descriptor
1230 - GLONASS L1 and L2, C/A and P, Code-Phase Biases.

Note the addition of L2 and GLONASS, but otherwise similar, slightly
extended, data.

Since the NEO-M9N does not support L2 that is not helpful, but the
GLONASS potentially is.

> I am assuming the M9N has RTK builtin, like the F9P.

Yup.  By default.  You can see if a u-blox is configured to accept RTCMx
this way: ubxtool -p CFG-PRT -v 2

Uh, oh.  I see my M9N can do RTCM3 but is configured off.  Oooops, dunno
how that got changed from the default.  Guess I gotta rerun my RTCM3
test...

A quick reset: ubxtool -p RESET, and then CFG-PRT returns the defaults:

UBX-CFG-PRT:
 PortID 3 (USB) reserved1 0 txReady 0x0
  reserved2 [0 0]
  inProtoMask 0x23 outProtoMask 0x3
  reserved3 0 reserved4 0
    inProtoMask (UBX NMEA RTCM3)
    outProtoMask (UBX NMEA)

Notice inProtoMask shows RTCM3.  So the input will decode RTCM3.

But when I look at NAV-SAT and NAV-SIG, it only shows SBAS corrections
(basically RTCM2), but no RTCM3 corrections.  The input RTCM3 is not
doing anything...

So much for my "working" rtcm3.  I have no idea what is wrong...

>  Did you tell the M9N to do RTK, or does it just do that by
> itself, or ?

Well, the dafault is on, but as I found, that needs to be double checked...

> I wonder what happesn with rtcm2.

A u-blox understands RTCM2 or RTCM3, but not both.  UBX-NAV-SAT
describes each sat seen, and tells you if RTCM corrections were applied
to that sat.  UBX-NAV-SIG supplies similar and more.  Otherwise what
they do is opaque.

I guess gpsd should report those new bits: rtcmCorrUsed and corrSource.

> I'm not 100% sure, but my
> impression is that rtcm2 streams have pseudorange corrections for
> regular differential and rtcm3 have carrier phase information for RTK.

Both RTCM2 and RTCM3 can, but are not required to, carry both
carrier phase and psuedorange.  RTCM3 supports L2, more constellations,
and more precision.

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

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

pgp8KrJM6u8tk.pgp
Description: OpenPGP digital signature