GPS Yaw Troubles/Internal Compass

I am running dual simpleRTK2B boards on my pixhawk 2.4.8. I used the steps in GPS for Yaw(AKA moving baseline). I also have another simpleRTK2B board as a base. I do achieve a heading in the data screen but I have a couple issues.
Here are my issues:

  1. With three boards, I have a base, moving base, and rover. The current mission planner rover firmware has trouble with the heading setup. It works fine, but does not allow GPS1 to get rtk fix from the base. Injecting rtk through MP instead of letting it send rtk through xbee radio, throws the pos and heading way off. Fortunately the MP beta rover firmware handles my current setup just fine. Both GPS’s have RTK. I do feel like I’m missing something though. I tried all sorts of board combos(1.13 and 1.32 firmware, two rovers or rover and moving base, let MP Auto_Config) but nothing seemed to let me achieve rtk on both moving boards. Weirdly but thankfully, the beta allows me to do this, but I’d rather be using a stable rover version.
  2. I’ve disabled every variable possible(I think) that mentions or affects the internal pixhawk compass. No matter what though, it still inputs heading. We’re converting a zero turn mower, and have switched to dual gps for heading, because the compass has too much interference. In some moments, the data screen spins wildly because of vibration and interference from the motor. With the new dual gps setup, we really need to disable the compass completely. @Yuri_Rage We’re pretty much trying to replicate your setup with the zero turn. Is it possible to completely disable, without it falling back on the compass?

This is starting to make more sense with the loads of research through discussions. While the forum is literally a goldmine of info on ardurover, it’s somewhat overwhelming to navigate through. It’s difficult to determine which components do what sections of thinking. Some people describe plugging everything into the pixhawk and letting it do the majority of the work and others seem to jumper wire here and there and communicate between boards, independently of the pixhawk. There are just too many streams of data flowing. Trying to continually wrap my head around different functions of Mission Planner.
I apologize if I’ve missed an existing, applicable discussion, but any help would be appreciated.
RoseyMow_Aug24b.param (15.3 KB)