Issues With PPK and CRS conversion

Hi Guys,

I’ve just received my Reach RS kit which I plan to use for laying GCP’s from which I will require absolute accuracy for most of my projects. It’s my understanding that in order to get an accurate absolute accuracy without having an available pre-surveyed point I’ll either need to setup a NTRIP connection in the field or post process against my nearest station’s RINEX files. As some of my previous works have been done in places without reception to setup a VRS I’m looking at focusing on the post processing option.

With this in mind, my planned workflow has been (for sites without a pre-surveyed point available):

  • Place base station (Average single position) and leave RAW logging enabled for at least 20 minutes.
  • Perform survey with rover using LoRa RTK connection.
  • Post process base station files:
    • Convert RAW u-blox to obs,nav & sbs using RTKCONV
    • Covert .17d Rinex files to .17o using CRNX2RNX (if I skip this step, RTKPOST doesn’t seem to be able to read the files)
    • Process base vs CORS using RTKPOST as shown in the image below:
    • Calculate the average position values given in the output .pos file.
    • Calculate difference between post processed position vs the base positioned used during the survey.
  • Offset all surveyed positions by the value calculated in the previous step.
  • Convert values to GDA94 and AHD.

At this stage, I’ve tested setting my base station up on top of two pre-surveyed points and then tried post processing the raw logs to see how accurate my positioning is, with the distance between my base and the reference station being 16km and 12km. However, upon comparing my post processed average position with the surveyed position (translated ITRF2008->GDA94), I’m getting large errors. Today’s test resulted in errors of about, 0.96m E 1.93m N and 1.97m Z.

The files I’ve used are in this dropbox folder:

The base was placed on top of the following point(GDA94):

  • 27º 33’ 42.75271" S

  • 152º 56’ 25.52630" E

  • 494118.814m

  • 6951327.940m

  • 29.171m AHD

I’m going to keep investigating why I’m getting the error but I think it could be either:

  1. Incorrect post processing
  2. Poor base placement (base could only see 12-14 satellites, on a previous trial I had 20-22.
  3. Incorrect CRS transformation

Any help with troubleshooting where I might have gone wrong / any help with correcting my workflow to be more aligned with surveying practices would be greatly appreciated.


1 Like

It seems you are missing correction data for the last 15mins of the total 25min rover file. Also, a bit few satellites yes.

1 Like

Thanks @TB_RTK , I’m assuming you’re referring to the reference station file? The one I uploaded didn’t include the last 15 minutes of the observed data, sorry. I just downloaded the next reference station obs file and concatenated the two, so as to cover the complete data set. After re-processing my errors came to 0.956m E & 1.33m N. I’ll try and collect another data set tomorrow with better base placement to improve the satellite count, I’ll upload my results.

Yup :+1:
With longer log time, you migh get lucky and fix point, or at least statet that its not doable without better visibility.

Out of curiosity, how did you concatenate the .obs files? I’ve run into that issue, myself, where I have two separate raw logs that I’d like to treat as a single entity. Thank you!

If you use a wild card (*) at the end as your file path, it should read all data within that path.

1 Like

@Lauren_O, I just manually copied over the section below the header from the second file below the entries from the first. The header of the file I have doesn’t specify an end time, only a start so I didn’t have to make any changes there. TB_RTK’s method is easier though, so for example my file path became C:\Users\Drone\Desktop\Survey Logs\27-07-17*.17o

1 Like

@bwagjor @TB_RTK Those are both really helpful suggestions! Thank you!

This topic was automatically closed after 100 days. New replies are no longer allowed.