I have an F9P and F9H setup finally going on my lawnmower rover and a pixhawk4. The good news is if I sit still with full view of the sky, everything is great. GPS heading is spot on, and all is well. The second a tree comes nearby, it drops out, pixhawk goes to dcm mode, and sits there until I’m just about to get to another obstruction, switches back to ekf3 then right back to dcm. I never lose rtk fix on the main F9P, so something on the H must be saying don’t output unless everything is super optimal! I’ve ordered another F9P (through mouser) since the pixhawk seems to want to know the locations of the antennas to do its own heading calculation anyhow, and it seems to be how the other rover guys are doing it as well…
so I have no doubt the F9H works quite well in the sky, but unless there is a parameter to tune to make it more forgiving of the loss of a couple of satellites, it’s useless on a rover that’s not painting a football field :). (Or track and field lines, or similar)
Should be the same performance as with the F9P. It is more difficult for a rover to calculate position out of a moving base than from a static base.
An idea to test if not already done, mount the antennas on ground planes (20cm square side or 20cm diameter round), or move to survey antennas.
I’m using survey antennas, and it does give heading with a full clear view of the sky. The F9P though will remain in rtk fixed with small trees as minor obstructions, or the house blocking part of the sky, the F9H will stop sending heading though. It seems really picky about when it will send a heading. I was hoping there was a tweak for that!
Using F9P/F9H combos here with FW 1.13 HPG/HDG
The parts should be really consistent.
Are you saying UBX-NAV-RELPOSNED is not reporting or that some of the sub-fields and flags need better decoding?