.POS or .LLH?

Hi Everyone!

I am very new to GPS and processing data so my apologies if I am not clear.

I am doing some preliminary testing with the Emlid Reach rs+ and went through system update just fine and the getting started tutorial. I have had what seems like some success for the last couple weeks. However, this situation pops up frequently:

I go through RTKpost in order to convert RINEX files, and but when I ‘drag the .pos file’ to the RTKplot or just click ‘plot’ i get a giant mess and nothing that looks anything like what my walking path was. However, when I ‘drag the .LLH file’ to RTKplot it gives me what looks like a perfect plot of the path I walked. My problem is I need to convert my data into a .GPX file. This is easy in RTKpost but the .POS file is a mess and you can’t do this with a .LLH file.
I’ve never seen the discrepancy this extreme but does anyone know what’s going on just by looking at this image and a usual reason why a .POS file does this after conversion ??

-The green line a plot of the .LLH file and resembles my actual route .

Something similar just came through another thread. What version of RTKLIB are you using?

RTKLIB ver.2.4.3 Emlid b28

That should be right for the RS+. You might need to post some screenshots of your settings windows in RTKCONV and RTKPOST.

You might check this to see if there is anything out of the ordinary.


Sure. A picture of the rtkplot is the one in the OP and here is a picture of rtkpost.
I’ve followed the directions to the T, and double checked it just now.

It only does this sometimes. Not always.

Go into Options and get the Settings 1, Settings 2 and the Positions tabs.

Positions Settings1 Settings2

The only settings that look different from what I run is it appears GLONASS AR is off and I have Min Fix/Hold Sats at 3. If AR was off in the RTK settings then you can’t turn it on in RTKPOST, but neither should cause what you are seeing. I only use PPK for my drone so I always run fix-and-hold, but if you are surveying points on the ground I have seen people run Static. Beyond that we will wait for a Ground Survey PPK genius to chime in.

1 Like

your rover and base station file names are rather similar, you got the rover file from the rover and the base from the base station correct?

rtklib_reach-rs_GNSS_1hz.zip (27.4 KB)

hey i have just recently been developing my skills with RTkLib
here is a set of config files i have been using,
(attached to my post hopefully)

and this site is really helpful


Oooh. You’re definitely correct about Glonass AR being off in the Settings2. Will look into that, thanks!

1 Like

if your doing PPK i think you need to select combined instead of forward.
from the RTKlibExplorer site

This is the direction in time that the kalman filter is run. For real-time processing, “forward” is your only choice. For post-processing, “combined” first runs the filter forward, then backwards and combines the results. …

1 Like

A little over my head, but will try that as well!

you just unzip the file and inside there are *.conf files
click on “load…” and you can load them into rtklib
first save your config file so you can compare them in something like winmerge

Hi Jonathan,

May I ask you to share your data so I can check them and try to post-process?


1 Like

Sure that’d be great. Thanks!

So when I download my sent my data from Emlid Reach RS+, I get Base .OBS, Raw (Rover) .OBS and .NAV files and I also get a folder that is called ‘Solution’ which contains a .LLH file. That’s the file that when I plot it, it works. This forum will not let me upload it here though. It says .LLH files are not allowed. So included here are Base .OBS, and Raw (Rover) .OBS and .NAV

base_201910071446.obs (226.9 KB)

raw_201910071446.nav (4.2 KB)

raw_201910071446.obs (899.8 KB)

1 Like

Hi @fangjian , you only have 2 minutes of observation in your files, and the Base has very few satellites and very poor signal.
My result Q=2 (47.0%) Q=5 (53.0%)

1 Like

[quote=“gabrielgal76, post:19, topic:16488”]
Hi @fangjian , you only have 2 minutes of observation in your files, and the Base has very few satellites and very poor signal.
My result Q=2 (47.0%) Q=5 (53.0%)[/quote]

Thanks for checking it out.

Before experiments are officially conducted, I’m just going out there and walking clear paths that I can verify afterward with ArcGIS and check its accuracy. I knew that for this one that the satellite signals would be poor because I did not set up the base in a clear area with 30 degree visibility. I was a little shocked at just how off it really was though. Also, the .LLH file was correct (green line in image uploaded earlier) and also the rover said it was in float mode at the time (sub-meter accuracy) so I assume it was getting base corrections. Although just not centimeter accuracy. And does anyone know what the .LLH file is ? Can I use this file and somehow convert it to .GPX ? …

You can’t check accuracy if you don’t have proper SNR in place, especially with L1 GNSS.

1 Like