KakutesH7/H7v2/H7mini: use EKF2

[julianoes]

This switches Kakute F7 and H7v1 to EKF2 to match Kakute H7v2 and mini.

Untested, could you test @vincentpoont2 ?

[bresch]

you might also want to set EKF2_MAG_TYPE 5 (none)

[julianoes]

I didn't do that for the other Kakutes either. So you're saying I should?

[bresch]

Nope, in fact it's fine, the EKF will still initialize the heading using the yaw estimator if there is no mag data:

PX4-Autopilot/src/modules/ekf2/EKF/gps_control.cpp

Lines 126 to 141 in d4c7ea7

	if (((_params.gnss_ctrl & GnssCtrl::HPOS) || (_params.gnss_ctrl & GnssCtrl::VEL))
	&& _control_status.flags.tilt_align
	&& _NED_origin_initialised
	) {
	// if GPS is otherwise ready to go other than yaw align
	if (!_control_status.flags.yaw_align && gps_checks_passing && !gps_checks_failing) {
	if (resetYawToEKFGSF()) {
	ECL_INFO("GPS yaw aligned using IMU");
	}
	}
	if (_control_status.flags.yaw_align) {
	mandatory_conditions_passing = true;
	}
	}

[julianoes]

Oh, ok, so I'll drop the commit again.

[bkueng]

Can you disable these ekf parts: https://github.com/PX4/PX4-Autopilot/blob/main/boards/px4/fmu-v2/default.px4board#L26-L31?

[julianoes]

Ok, even if a GPS might be connected potentially? And would you do it for all Kakutes or just the F7?

[bkueng]

GPS will still work. We can do it for all of them as they're generally low on flash.

[julianoes]

Alright, will do. Thanks.

[julianoes]

@bkueng it looks like there is no way to fit EKF2 onto 1MB flash on the KakuteF7. Any ideas what to do? Retire it, or going back to Q?

[bkueng]

Let's keep Q then.

[dagar]

There's more we can carve up in EKF2.

[vincentpoont2]

@julianoes @bresch Please merge.
@julianoes Can we backport this to 1.14.1?

[bkueng]

@vincentpoont2 did you test this?

[vincentpoont2]

@vincentpoont2 did you test this?

Yes, we did a test flight. The only thing we notice is the following warning.

[bkueng]

Thanks @vincentpoont2, then let's get this merged.
For reference, can you share the link to the log?

[vincentpoont2]

Thanks @vincentpoont2, then let's get this merged.
For reference, can you share the link to the log?

https://logs.px4.io/plot_app?log=82c6495e-4196-4985-aa9d-8b8a9ebfe7c6