RTK Receivers and PPP Correction Procedures and Errors

[os2man4]

I am attempting to get my RTK Facet to utilize correction data from an NTRIP server. Thanks to help on this forum I’ve gotten the Facet to accept NTRIP data and enter the correct mode.

Since there is no station local to me I would like to setup a Base Station myself and I’ve been unsuccessfully trying to follow the instructions located on the tutorial

https://learn.sparkfun.com/tutorials/ho … tation/all

I run into two issues, my correction data is either inaccurate or the CSRS immediately kicks back an error.

The data I have been able to submit shows

The estimated coordinates ITRF14 2022-10-26 for the COM7_221026_221346.obs RINEX file are as follows:

Latitude N40° 39’ 30.5461" ± 0.478 m (95%)

Longitude W74° 28’ 14.7017" ± 0.564 m (95%)

Ellipsoidal Height 53.428 m ± 1.107 m (95%)

Orthometric Height CGVD2013 CGG2013a 87.3130 m

[40.65848502,-74.47075048,53.428]

UTM Zone 18 (North)

Northing 4500981.419 m

Easting 544739.744 m

Scale factor (point) 0.99962464

Scale factor (combined) 0.99961626

[4500981.419,544739.744,53.428]

Cartesian coordinates

X 1297267.618 ± 0.647 m (95%)

Y -4668539.782 ± 0.910 m (95%)

Z 4133761.330 ± 0.725 m (95%)

[1297267.618,-4668539.782,4133761.330]

Orbits and Clocks Used: NRCan Rapid

GNSS Data: GPS

GRS80 ellipsoid used for (x,y,z) to (lat,lon,h) transformation

The example shows getting a location less then 0.006m and I am easily 100x that. I tried two separate times with two sets of data. The data was collected with a GPS RTK2 from Sparkfun, the two datasets were both little over 28 hrs long and both had equally meh results.

I’m not sure where to start, but some things seemed odd and I am not sure if they are relevant.

When I have UBlox open and observing the data it shows 28 more satellites, however when I plot the data after capture using RTKLIB and RTKCONV is shows less then 10. I’ve attached two images showing what I see in UCenter and RTKPLOT.

So to try and rule out my setup I attempted to utilize the Facet itself to collect the data. I can’t use it as my base station and rover simultaneously, but wanted to see if I could collect good data from the Facet and get a precise measurement from that device instead (this would have hopefully ruled out bad cables, poor antenna, or other hardware problems). With the Facet, I can’t seem to get good data at all. I’ve tried varying some settings and resetting to factory, but in every case when I convert the UBX file to OBS and submit it to the CSRS PPP portal I get and error like the following.

COM9___9600_221029_011041B.obs Error : The RINEX file contains formatting errors in the observations at epoch 2022 10 29 1 11 29.99. Please correct and resubmit.

The data itself looks fine to me, I can plot position, number of satellites and it plays back just fine in UCenter. I’ve tried taking multiple data sets and they all receive the same error.

I’ve tried to put relevant information into this post, but I am sure that more information would be required so let me know what else if anything I should post up and I will do so.

Thank you very much,

Jim

[sparky]

Thanks for a great support post!

I agree, ± 0.478 m is really bad for a PPP survey. You should be sub 0.010m at 28 hours.

Many thoughts in no particular order:

• rtkconv will default to only GPS and GLONASS satellites so I’m not too surprised you’re seeing a discrepancy. Don’t worry about it. We use the default settings with CSRS all the time.

• How good is your reception? Any buildings around? For PPP logging, a clear view of the sky is very much needed. I see you have reception to a good handful of satellites (~25) but, according to your screenshots (thanks!), it’s less than I would expect. I regularly see 35 or more satellites (see attached image - Note: this was taken using Express+TOP106 antenna). The ZED-F9x receiver will use a subset of these to obtain a fix (SIV on the front of the unit is the # of sats used for a fix, not just in view).

• The Facet is not designed for permanent external mounting (it is not waterproof enough for all environments). I recommend an Express + TOP106 antenna as it’s my go to combo.

• CSRS is likely to reject a RINEX file with oddities. Log corruption is a known issue with RTK product firmware v2.5 and below (https://github.com/sparkfun/SparkFun_RT … issues/345) with a fix in the works. If you’d like, migrate to this release candidate to improve log validity.

download/file.php?id=7659

[os2man4]

Ok, so finally got a chance to take a look at this some more. In an effort to improve things I remounted my antenna to the top of the roof. It is now a full 30 feet taller and I got results that were twice as good! Since I was 200 times more then I was looking for when I started, that’s still unfortunately not great.

Cartesian coordinates

X 1297277.096 ± 0.302 m (95%)

Y -4668540.907 ± 0.383 m (95%)

Z 4133763.177 ± 0.299 m (95%)

I did see approximately 35 Satellites instead of 25, so I was able to increase the number of Satellites in view.

I noticed two other items that could potentially be an issue?

I noticed in Sparky’s message above that he is using an L1/L2 band antenna and I noticed that in his Satellite view it appears as if his charts have two lines and text at the top L1C / L2C (or similar its a little hard to read). The screenshot that I took is not the same and does not have those lines, so potentially my antenna is only single band or I have the receiver configured improperly?

Also when I submit the file to CSRS I notice two differences between the example on the Sparkfun webpage and what I receive. (See attached picture)

The example shows that it utilized GPS & GLONASS NRCan Ultra-rapid where my result showed that it only used GPS NRCan Ultra-rapid. It also gave a warning that said it could not use ambiguity resolution PPP because it had no GLONASS products.

I had GLONASS, BEIDOU, and Galileo satellites in my solution, what would cause them to not be utilized or saved?

I feel like my problem is related to the dataset and that I either need to get an L1/L2 antenna or somehow get the GLONASS data saved. Not quite sure if any of that makes sense, I’m just trying to figure out differences and grasp at anything that might make sense.

Thank you very much,

Jim

[sparky]

The screenshot that I took is not the same and does not have those lines, so potentially my antenna is only single band or I have the receiver configured improperly?

Eagle eyes! Nice job. I agree, it looks like your Facet is only receiving L1 band. I have never seen an antenna fail a partial element like that. Very interesting. But, that is your problem: you will never get good results with L1 only.

You are welcome to return the unit for repair (https://www.sparkfun.com/returns). But before we do that, let’s make sure the configuration is not the issue.

It is possible to turn off L2 reception of the GNSS but it’s very advanced and the RTK firmware doesn’t even really support it. Generally, you have to do it through u-center. To be really sure, I recommend doing a factory reset from the system status menu (https://docs.sparkfun.com/SparkFun_RTK_ … em_status/) - ‘r’ and then ‘y’ to confirm. You can also do this over WiFi.

[os2man4]

Ok, so here is where I am at, BLUF - It’s great.

The biggest issue I did not realize was that multi-band antennas mean, L1 / L2 Bands not GPS, GLONASS, BEIDOU. I was under the mistaken impression that multi-service meant multi-band. So I was using an incorrect antenna for the level of accuracy that I was looking for.

Once I got a TOP106 antenna I saw the L1/L2 bands being received by the receiver. I was able to capture data and successfully submit it to CSRS and have my position localized within 0.003!

Thank you very much for your help,

Jim

[sparky]

Excellent news! Congrats and happy RTK’ing.