For the tuning, I may have misunderstand somthing, but I canāt get any values to change. This window is not going away, I canāt hit ok.
Looks a bit strange
For the tuning, I may have misunderstand somthing, but I canāt get any values to change. This window is not going away, I canāt hit ok.
Looks a bit strange
I donāt think autotuning is fully supported yet
Flight test 250 quad: Pixracer, Q estimator
ACRO and RATTITUDE modes are good. ALTCTL and POSCTL also worked well
Above has nothing to do with Auto-Tuning.
@Mark_Whitehorn No, Iām working on my macbook pro 15". Itās an old one 8years but still up to date.
Itās not autotune that Iām taking about!
Found a bug with accelerometer calibration, it shows only 3 picture (as well as mag calibration) but expects 6 rotations:
Thanks guys - in trying to simplify mag cal bad images for accel cal slipped in. We have fixed this now. More testing with the most recent QGC daily and the beta branch would be appreciated, in particular if you had a well tuned setup before.
@Mark_Whitehorn Would you be interested getting some mission tests in with your tablet?
Had a flight with the Hexa today using latest Beta from today. Flight was ok, but still some issues with Battery Measurement IMHO. It seems more sensitive than in Version 1.2. Finally I got new Power Supply Cable (AWG 12) which I exchanged this evening. Iāll try to do some test flights maybe Friday/Saturday to recheck if this was due to too small Cabling.
Hello,
I am also experiencing odd Battery level readings.
I use a pixhawk with its power-board.
The Battery is a 3S with 4500mAh.
It is at 0% at 10V (thatās where the Voltage breaks down)
It is at 100% 12.46 or 12.6V (charged twice, this is the variance)
PX4 gives me an alarm at 11.5V and 1842mAh
I have used these Settings:
C_Scaling 0.013
V_Scaling 0.0082
These are calibrated, i.e. I did measure the Voltage and Current under load (ājust before hoveringā)
Greetings,
Hendrik
Ok, I finally found the stupid mistake: I had switched two variables in the battery estimation filter which led to the unfiltered charge level to be used for the failsafe handling. Which means that noise on the voltage lines could cause the alarm to trigger.
This is now fixed for good. Test flights (wait 10 mins, then flash beta) would be highly appreciated. Iāve increased the default velocity control gain as well, so position hold should be slightly better. Please let us know if its too aggressive now (weāve flight tested this).
With the release of QGC 2.9.5 are we supposed to use the daily build still for the beta testing or the stable build?
Always use beta, the Firmware versions do not differ between QGC versions. As long as 1.3.0 is not released you need to use beta.
Flashed beta to PixRacer using QGC 2.9.5. QGC complained it was unable to fetch some parameters after flashing. Disconnected QGC, unplugged/replugged USB, then QGC connected with no complaints.
Attempted to set up a mission to be flown at the field using desktop machine at home: map was very slow to pan and zoom. Left the desk to give the map time to fill in and when I returned it had moved back to vehicle location. āFollow Vehicleā checkbox is gone, and choosing āhomeā or āmissionā takes me halfway around the world. Hence Iām unable to perform mission planning at the bench. Maybe it will work at the field on Android.
When I try to send a mission to the vehicle, QGC reports āā¦too far: 8318876 m, refusingā¦ā
According to the display, āhomeā is a few meters from the first WP, not >8 million. user error?
Let me quickly repro here.
Installed QGC 2.9.5 with latest Beta Build. I did not fly yet noticed some different behaviour:
Log from an fmu-v4 test flight: http://logs.uaventure.com/view/qBX3ywh7na2eJiFXbymFoA
Altitude control was very erratic, Mission mode didnāt do anything resembling the programmed waypoints.
QGC did not display any map imagery, even though I had tried to set up offline files.
Question: is the mission definition logged anywhere to verify that it is being followed correctly?