Trouble with RTKLIB and Phantom 4 RTK Data

[michaelL (Michael Lambert)]

Good morning all! I am trying to assist a user on the DroneDeploy forum that is trying to PPK their P4 RTK data and is getting some really strange results in the Emlid version of RTKLIB that I have been able to reproduce. He also has used another version (b33) and the GPS track looks correct, but is not able to get a fixed condition. Please help! Here’s the history. We can continue here and I can cross-post back if we can come up with something. Thanks!

How To - PPK Processing with RTKLib (Phantom 4 RTK)

Here is a previous Emlid Community thread dealing with some of the same information, but it has been closed.

DJI Phantom 4 RTK Post-Processing

This is not for Reach specifically, but comes from a user who is trying to post-process data from the actual P4P RTK and I told him he would have a better chance of getting help from you all than the folks on DroneDeploy’s forum. Thanks!

[michaelL (Michael Lambert)]

Here is the link to download the raw files. The NAV file from the “base” was used.

https://drive.google.com/drive/folders/1_-4NaFBSYohRuRPjzdcmVC3OGac58BqM?usp=sharing

[Magda_Chmielewska (Magda)]

I post-process data in RTKPOST ver.demo5 b33 and i’ve got a solution

BTW. Any idea why same input data in different version of RTKLIB gives totally different solution?

[tiedtke (Thiago - T2R Tech)]

Hi, I did the processing with RTKpost, with no problems.

And everything works like expected. Works fine in several configurations, like: only GPS, GPS+GLONASS, and AR in Fix-and-Hold, Continuous, etc

See:

100_0069_Rinex.pos (187.7 KB)

[wizprod (Christian Grüner)]

Firstly: because the initial/default parameters are different from branch to branch it seems.
Secondly: later versions can have improvements to the algorithms used for processing, I.e. the current version for the RS2 can use the E5b signal, the RS(+) version can’t.

[Magda_Chmielewska (Magda)]

OK, I got it, but if i would set up by myself exactly the same parameters in different versions of RTKLIB i would expect to get exactly same solution.
On the other hand bugs in program’s source code are quite common…
So far problem solved

[michaelL (Michael Lambert)]

I think it has more to do with when they were coded and what hardware was available at the time. Kind of like programs that won’t run on Windows 7… I’m still very interested in the crazy pattern we got in Emlid’s version.