We have mounted a M2+ on a drone, which gets correction data from our Reach RS2+.
The measured position comes in at about 3hz. It seems quite consistent at some points, but has large jumps at a seemingly regular (~9s) interval while the time between datapoints remains consistent.
Our Base also got correction data from a nearby reference station, shown in the emlid app to be coming in at intervals of 0.6 to 1.5 seconds.
Below the measured datapoints are shown in red markers, with a blue line connecting them.
The overall shape of the path matches the real path that our drone made, besides the periodic “teleportation”. Our speed was reasonably consistent (0.2-0.4m/s)
How are you moving the data and to what?
Here is a quick schematic of how we move our data through a Nuc computer through a ROS network (real-time), to be recorded and eventually plotted (afterwards, not real time) with Matlab.
My guess is the NMEA parser you are using is unable to handle additional periodic satellite in view messages, either because it is not coded to handle them, or your communication rate is too slow and you are ending up overrunning the buffer and its garbling the data.
At 9600 baud 1 byte takes 1ms to transfer, so one NMEA message at 82 bytes takes 82ms to move trough the buffer, when the GSV’s come up every second it could be 20 or 30 messages at once flooding the buffer.
Always choose the fastest baud rate that you can successfully connect at without problems.
