Hello, I am trying to obtain the correction data by connecting to satellites and average their data. I have the GPS RTk 2 ZED-F9P board. I connected it to RTKLIB, but I don’t think we actually obtain the corrected data. We then need to send this data to PX4. If anyone can help me by providing steps on how to obtain this correction data.
Thank you!
Check out our guide GPS-RTK2 Hookup Guide - SparkFun Learn
I’m trying to understand what you’re attempting.
What is the GNSS on the PX4 side, what sort of “corrections” are you trying to feed it?
The ZED-F9P can output RTCM3 messages directly, although the MSM messages aren’t corrections, but rather observations of the satellite signals from a fixed/known/established location.
I’m trying to set up a GNSS receiver that will provide navigation data to the navigation system on the rocket. And then, I need to create a base station that can send correction data to the receiver.
Ok, but WHICH GNSS Receiver do you have on the Rocket? That’s the question…
Many don’t support/use the 1980’s method of RTCM2, ie Pseudo Range Correction / Range/Rate, and it was never very good, gets you to perhaps 3m when there was 100m of Selective Availability Dither. But not sure that’s been a thing for decades
You need something supporting RTCM3 MSM Carrier Based methods
So a ZED-F9P as a Base, pushing RTCM3 1005, 1077, 1087, 1230, etc. to a Rover also using a ZED-F9P or NEO-M8P. The Rover takes these, you’ll need to push or tunnel them over your radio link. Watch you have a radio with TWO different frequencies so you can push this data over one band, and get your telemetry back on another, so you don’t jam a single channel / half-duplex link. The RTCM3 data is a ONE WAY broadcast of data from the Base Station to the Rover(s).
You’d really want to be using a L1/L2 dual band as L1 recovery from loss-of-lock is MINUTES
There are perhaps Jerk/Acceleration limits, see AIRBORNE 4G, or better if you have a High Dynamics receiver. COCOM/ITAR limit Consumer/Commercial GNSS Receivers to around 515m/s aka Mach 1.5
Doing some artisanal/home-cooked method than has to un-bake an LLA solution, and “fix” it is not workable.
Conceptually this might help with the mechanics of RTK
I’m guessing you’re using the SparkFun ZED-F9P on the Rover, but you need a second one on the ground at a fixed/known location to broadcast the RTCM3 data. Think of RTK as a triangle geometry problem, rather than using differences aka differential corrections, describing what’s wrong and how it’s changing.