That isn’t useful information. We already know it isn’t getting correct heading.
I’m nearly certain the value you’re looking at is heading to the NTRIP station and has nothing to do with the problem at hand.
Load the configuration files from ArduSimple again. You keep mucking about in uCenter, but I’m afraid you’re only making the situation worse by using that software after loading the pre-configured files.
Reboot the autopilot after making significant changes to ensure that everything is initialized correctly.
Is there anything I can do to resolve this gps2yaw constantly being 65535? I have set gps position offsets multiple times, nothing seems to have worked out correctly.
So to make this clear: I have the Ardupilot, 1 antenna in front of it in 0.5 meters and back from it in 0.43. I tried to set GPS_POS1_X to 0.93 (the diff between antennas) and POS1_X to 0.5 and POS2_X to 0.43. Neither seemed to work.
Any other configuration I could use for the simpleRTK2B and Lite modules?
I have reloaded the configuration files, the issue is still there.
This is why I don’t recommend using this hardware combination. And I don’t have any more advice to give. It does work, but it needs to be configured exactly per the instructions that you seem to have followed. But any little step missed results in disaster.
My honest recommendation:
Return this hardware in favor of two SimpleRTK2B (full sized) boards that can be auto configured by ArduPilot.
If you don’t want to do that, maybe someone at ArduSimple would be willing to get on a support call with you.
Ahh, now I see why you said this is a PITA…
I contacted them, they told me that okay, follow the instructions, but I’m already through that point multiple times…
My teacher said that if we can read positions in RTK Fixed precision, we could calculate the yaw, pitch, roll orientations on our own. What do you say about that? If we have a way to log positions measured from the two antennas and append the orientation, it is fine at this point, he said.
What if I disconnect the 2BLite from the 2B, and handle them separately. RTK Injection should be sent to both modules, right because of GPS_INJECT_TO parameter, right? Then I will still be able to fetch RTK Fixed signals, maybe the GPS2 won’t get decreased to 3d dgps either, and then can read positions from GPS1 and GPS2 with RTK Fixed precision. What do you think about this?
For now it would be nice to log the positions and orientation of the rover with RTK Fixed. Right now this is the use case.
As far as I know, if we know RTK Fixed position, the calculated orientation will be that precise as well. My question is, is it okay if I disconnect the 2B and 2BLite modules, and let Ardupilot configure them starting from the default configuration of the GNSS modules (I have to reset again then) with GPS_AUTO_CONFIG set to 1?
I have no knowledge of your GPS-devices and also no information about ArduSimple for now.
But some more general information about GPS and terrestical navigation.
If you have a fixed GPS base station which can send DGPS correction data and you have a GPS rover station which can receive the DGPS data than the position of only the GPS rover can be correctetd relative to the base station- So if the base station has a known position also the GPS rover has a corrected absolute position.
If you now use both positions you can easily calculate the direction from GPS rover to GPS base. This is the baseline. If you now place both GPS on a moving platform than you still can calculate this baseline between the two GPS but as now the GPS Base station has not a precisely position also the position of the GPS rover has only normal accuracy.
So if your GPS system can do all this jobs as I assume from your discussion than first
Don’t use the NTRIB signals. This is in the first setup not helpfull.
Than setup the GPS base to a fixed position and send the correction data to the GPS rover. So only the GPS rover will receive any correction data and must show that it is receiving and using this data.
Afterwards you can calculate the baseline out of both GPS positions.
Actually, I should’ve mentioned this much earlier. I tested this some time ago, and it works fine. Perhaps it should be my recommendation for anyone with this frustrating hardware.
Physically separate the two boards. Reset both to defaults in u-Center. Then use ArduPilot’s own documentation to configure:
Alright, thank you very much. First thing in the morning will be to check this out. Sorry that I kept repeating stuff, I just wanted to make sure I understand your guidance correctly, but I’m less-less confident about handling this issue.
Now the heading works fine! Thank you for this Ardupilot-based configuration!
We are injecting NTRIP corrections with mavproxy, and GPS2 gets it so it is RTK Fixed, GPS1 however stays at 3D dgps. Is it acceptable or do I have to change something?
The position starts to shift away from the real position. Could it be because we are between high buildings?
I can happily tell that today’s test was a success!
We received RTK Fixed positions + yaw from both GNSS-modules, finally.
This last solution works just as it should.
Thank you very much, Yuri and Juergen for guiding me around this topic.
There was a question about how we check the precision, how precise are we located. Is there a way to measure accuracy or fetch it from somewhere?
I think this is called the DOP, the dilution of the precision.
