Basic Tests Master 3DR X8 Pixhawk

[David Pawlak]

In the first log you can note that it is stabilizing.

Here I did two short flights, in stab. AHRS_EKF_USE = 0

I have the X8 auto-tuned so it is real sensitive, and I noted small twitches as mentioned by others.

For the most part, although the HUD indicates rolled right and pitched up, it's  level in flight, I had a fair bit of left roll and back pitch to keep it flat.

In the last part of the flight, it landed on uneven ground. You can see the pitch and roll. When I tried to rearm  the HUD doesn't move and the controls were very twitchy.

The throttle would jump dramatically coming out of zero, and it wasn't possible to take off, it would roll to one side. I moved it to flatter ground, you can see that the HUD moves, but even on flat ground, it had heavy roll and pitch. Couldn't take off there either.

In the second log, I had power cycled to see if that would help stablize the drift.

First two "flights" AHRS_EKF = 0, second two =1.

It seemed a lot more docile, although there were still small twitches.

The third and fourth "flights" were with AHRS_EKF_USE = 1. It seemed more stable and confortable to fly. Might be that it was "warming up" so I went back to one flight in EKF_USE = 0 , going to loiter and one with EKF_USE = 1, going to loiter. Before these last two flights, I did an auto calibration to tune out the roll and pitch inputs from the RC.

Loiter was not as stable as 3.2. Seemed to wonder a little over a meter. Both with and without EKF, about the same.

[Paul Riseborough]

Loiter stability is highly GPS sensitive - varies from day to day.

[Jonathan Challinger]

I've noticed that sometimes the copter won't take off in stabilize. It falls over. Requires a power cycle to fix it. Maybe a division by zero bug somewhere or something? I'll try to get a detailed log of it.

GPS  is pretty highly variable.

Auto trim and save trim don't work with ekf if I am not mistaken.

--
You received this message because you are subscribed to the Google Groups "drones-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to drones-discus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[David Pawlak]

Isn't that a part of the initial temperature drift factors?

Is it possible to anticipate those and compensate for them as they settle in?

Perhaps if a unit remembered  what it learned from previous flights about it's compensations, as it "warms up" after start-up, it could see that it's off, and know where it's going, thereby tracking the drift and compensating real time.

[David Pawlak]

As far as the auto trim, it sure seemed to make a difference. Gave me enough stability to take my hand off the roll/pitch to switch to loiter.

My Iris had similar comportment.(Admittedly without EKF), but the auto trim would only be temporary, because as the AHRS  started to line itself up, the trim was now working the other direction.

Didn't seem to happen here, although perhaps I really didn't fly too long after the auto-trim.

On Sunday, March 30, 2014 4:59:02 PM UTC-3, Jonathan Challinger wrote: