Hi all,
I’m using RS3 as rover with an RS2 on a tripod as base. Both receivers get a fix in a couple of minutes and location seems to be normal (about 1cm error).
The problem is heading. It is quite frequent that heading starts with an offset and/or drifts more than 10 degrees/second. These errors are shown in the app and are also confirmed in the NMEA stream.
After restarting the RS3 a random number of times, it tends to fix, but obviously is not an ideal situation. The situation is not solved by disabling corrections, disabling the IMU and reenabling it…
Do you have any pointers on how to fix it? No error is shown apart from the obvious figures in the heading.
Do you have a dataset to share and show the problem? The heading value can change a lot when the rover is pseudo-stationary because the positioning errors are predominant over the displacement between 2 consecutive measurement epochs.
You can either share them directly in this thread, or if confidentiality is an issue, I would suggest to open a ticket to the Emlid support support@emlid.com and share the files to them.
There has been a lot of communitation with the support team and multiple data exchanges, so it took some time to find the issues. Kudos to everyone involved.
The summary:
Heading drifting is a result of insufficient movement during the callibration step (the “move the receiver” stage). Even if the recommended 8-pattern was performed multiple times, full system reports showed the receiver callibration was successful even it was not.
Also, regarding frequent “tilt compensation” deactivations requiring new callibrations. The firmware asked for recallibration of the IMU after every transition from FIX to FLOAT even if it was not strictly required. Why the receiver is losing FIX so frequently (multiple times an hour) is still something that needs to be debugged.
The dev team comments the “heading drifting over time” and “IMU recallibration after losing FIX” are going to be fixed in a firmware update.
