ZED-F9R RTK: Large Position Shifts When Rover Is Moving (Stable When Stationary)

GPS/GNSS Development Boards
1

Hello everyone. I’m using a SparkFun RTK base + rover system, with a ZED-F9R (rover) connected to a Raspberry Pi 5. I’m currently testing by recording GPS positions on the rover every second. While I am able to log data properly, I’ve noticed an issue: the data sometimes drifts randomly, as if the RTK fix is lost for a short period.

Our main problem is shown in the lower-left illustration. It shows a 7-meter shift in the reported position within just one second. This shift moves most of the dataset to the west and north. Otherwise, the dataset looks reasonable. It seems clear that RTK fix was lost at that moment.

1755566498060blob
1755566498060blob961×573 58.8 KB

By contrast, testing at a known stationary rover position very close to the base does not show any movement when plotted on the same north/east scale. The bottom illustration only shows small, reasonable variations in the reported location.

1755566588679blob
1755566588679blob446×486 11.6 KB

Has anyone experienced similar issues? I’ve noticed that when I leave the rover stationary next to the base, the RTK LED on the F9R stays solid. But as soon as I start moving, it begins flashing. I understand that the F9R includes IMUs for dead reckoning, but I’m not sure if that relates to the shifts I’m observing in the corrected data.

More info about my setup:

Hardware / setup specifics

  • ZED-F9R rover, ZED-F9P base and they are connected via telemetry radios

  • Rover Antenna placed about 1.5 meters above the rover. Base antenna placed at about 8 feet. No building are around, and I’m using metal plates for antennae.

Software / configuration

  • On software side I’m running a simple python script similar to sparkfun example code which just records data to csv file every 1 second

  • I used the same configuration for rover and base as in this tutorial Setting up a Rover Base RTK System - SparkFun Learn

  • Test conditions

    • I tested driving the rover at normal speed, not very fast. Also tested walking but similar results, actually worse when walking so that might be a clue.

    • Distance between base and rover is about 10-15 meters at the time of testing.

    • There is a shed nearby where I test and a big tree but I don’t think those are major blockers for the signal since I’m on a farm and there isn’t much around me.

  • Symptoms

    • The big 7 m jump I saw happens almost as soon as the rover starts moving. I see good data while the RTK fix is there but then upon movement I see issues.
2

Do you happen to know why you’re losing RTK fix? Are there obstructions, hills, trees, cloud cover?

3

Hi TS-Russell, that’s what is puzzling to me. I’m losing RTK fix but i can’t figure out why. I’m testing on a farm and I don’t have very big obstacles near it. I’m testing in front of a barn, and I placed the base about six feet away from it, and about 8 feet high. And I’m driving the rover in the open field without any obstacles around it, antenna about 5 feet high and clear view of sky. I noticed that when i start recording it and i just stand there stationary, RTK led on the rover becomes solid green, but as soon as i move the rover antena even a little bit, the LED starts blinking indicating that I lost the RTK fix. Could this be some hardware malfunction or misconfiguration on my end?

4

Hmmm.

How much can the antenna move? If it sways more than a few mm it’ll throw things off. Maybe share a photo of the rover setup

Is it on a 4-wheeler or something that has its own grounding circuit? How is it being powered?

5

I’m not personally familiar with the F9R, but since it’s the version for dead reckoning and sensor fusion - maybe that’s an issue ?

IE: Moving the Antenna without agreement from the “additional” sensors causes problems for the FIX ?