Differences between RTK and PPK solution

[joe.w.long (Joe Long)]

I typically perform surveys with RTK link to a nearby CORS station. But I may have some situations in the future (multiple units but limited logins) where I need to rely on PPK. I’m also fairly new to the Reach units/Emlid Studio software but I’ve been through all the materials/docs/videos.

I was doing a few tests to compare the two methods and make sure I had a suitable workflow but I am having some trouble interpreting the results.

Here is what I did:

Connect to the CORS station, wait for a fixed signal, and survey a nearby manhole cover averaging over a 3- minute time period. When collecting that point, I also recorded the full position log (LLH) and the raw RINEX.

I stayed in the same position but turned off the NTRIP corrections. Recorded the position log and raw RINEX with the ‘single’ signal

Uploaded my non-RTK observation file to Emlid Studio as the rover, my nearby CORS station observation file as the base, and a navigation file (trying both the one from the CORS and from my rover) to get a corrected solution.

I tried doing the PPK in static and kinematic mode. I realize this was sitting at a fixed point but when I am out surveying, it is often with a unit in a backpack and I wanted to see the result from both methods.

The PPK corrections do not result in a fixed solution and they just shift the data rather than increasing the precision of the individual observations. This may be too short of a time series (3-minutes) but I am interested in understanding why the RTK points have good precision for the 3 minutes but the PPK cannot achieve that. My understanding is they are essentially being corrected by the same base time series just not in real-time. I’m hoping someone can help me understand what might cause these differences.

Thanks in advance!

[julia.shestakova]

Hi Joe,

Welcome to the community forum!

Can you please share the logs for PPK and the LLH log? I’ll check them, process them myself, and will get back to you with my thoughts. You can send the files here or via support@emlid.com.

[Zaz5400 (Zaz Villanueva)]

what is the base logging rate set at for the particular CORS that you are using? 1, 2, 5, 15, or 30 seconds. In the unlikely event that it is 30 seconds, you would want to log for at least 10 minutes in my opinion.

[joe.w.long (Joe Long)]

Thanks for the reply! I downloaded the CORS with a sampling rate of 1 second.

I’ve emailed a zip file with three folders (unable to upload here as a new user):

• base: downloaded from the CORS site
• single: logs collected with out corrections (no NTRIP enabled)
• fixedrtk: logs and positions collected with NTRIP and fixed signal

[julia.shestakova]

Hi Joe,

I’ve processed your files and achieved Float only as well, which is sub-meter accurate. So I checked the logs and noticed that both files from your Reach RS2 rover don’t contain info about GPS signals. Have you turned off recording GPS in the Logging settings?

RTK algorithm on Reach RS2 always uses GPS observations in calculations. And we recommend recording GPS data for PPK as well. Your logs quality looks ok to me. So turning on GPS constellation for future surveys should allow you to achieve Fix in PPK.

[joe.w.long (Joe Long)]

Thanks, I will give that a try!