One more thing comes to mind as what could have happened: I loaded the flight plan from file and for some reason flight plan was loaded with Absolute Alt instead of Relative Alt. Is there any way of checking that from the log?
I saved the plan I downloaded from the Pixhawk separately, but I am not sure it is the same.
Andre K. > Tamás L. Páll (Tom)March 29, 2016 at 9:39am
Yes
Andre K. > Tamás L. Páll (Tom)March 29, 2016 at 8:01am
the log shows executed commands, so I only see your takeoff command, altitude types are fine as altitude error grows while the plane were loosing altitude.
Yes, it looks scary, after the desired takeoff pitch is achieved, it starts to drop while the AS is technically below minimum, makes me think of the TECS correction related to undersped in 3.5.2, but somebody else should confirm that.
takeoff pitch was set to 15 deg. I do not recall seeing that pitch but it happened fast... Anyway, I have 3.4 on this plane.
Was the TECS tweaked since 3.4?
Andre K. > Tamás L. Páll (Tom)March 28, 2016 at 7:48am
yes, I know all three things you mention, yes, (log contain version too) TECS have been tweaked since then , and pitch of 15 was not achieved, only >13, very shortly after the throw(?/ or catapult). - so it's a bit different from what I remembered there, but still strange
I launched with right hand and I have Mode2 RC controller so I was pulling the "elevator" and the plane up from the moment I got my hand back on the transmitter, because I saw the plane was not climbing as it should. I thought I can avoid crash by giving it a little manual input... did not happen. :-(
One thing I can think of is an obstructed Pitot but that I cleaned the pitot after the previous crash right before this one.
The other might sound too paranoid but do you think there is a possibility of purpose crash when a new FW version comes out?
Andre K. > Tamás L. Páll (Tom)March 28, 2016 at 11:09am
no, your airspeed data is fine.
and the second;
no, because there is no evil corporation behind that slows down windows XP, or bricks and iSheep product because it detected that you replaced a display for 1/30 of the price.
ArduPilot is open source, running on open hardware, anyone can read the source, and see that there are no evil code to promote sales.
nor is there any reason for such, as it can run on a huge amount of hardware, including Raspberry Pi's and other hardware, there is not a single business that profits from it.
you can be sure of that with FOSS projects in general.
Replies
One more thing comes to mind as what could have happened: I loaded the flight plan from file and for some reason flight plan was loaded with Absolute Alt instead of Relative Alt. Is there any way of checking that from the log?
I saved the plan I downloaded from the Pixhawk separately, but I am not sure it is the same.
the log shows executed commands, so I only see your takeoff command, altitude types are fine as altitude error grows while the plane were loosing altitude.
Where do you see altitude error?
I could only find NTUN / AltErr as altitude error. is that you are looking at?
Yes, it looks scary, after the desired takeoff pitch is achieved, it starts to drop while the AS is technically below minimum, makes me think of the TECS correction related to undersped in 3.5.2, but somebody else should confirm that.
takeoff pitch was set to 15 deg. I do not recall seeing that pitch but it happened fast... Anyway, I have 3.4 on this plane.
Was the TECS tweaked since 3.4?
yes, I know all three things you mention, yes, (log contain version too) TECS have been tweaked since then , and pitch of 15 was not achieved, only >13, very shortly after the throw(?/ or catapult). - so it's a bit different from what I remembered there, but still strange
I launched with right hand and I have Mode2 RC controller so I was pulling the "elevator" and the plane up from the moment I got my hand back on the transmitter, because I saw the plane was not climbing as it should. I thought I can avoid crash by giving it a little manual input... did not happen. :-(
One thing I can think of is an obstructed Pitot but that I cleaned the pitot after the previous crash right before this one.
The other might sound too paranoid but do you think there is a possibility of purpose crash when a new FW version comes out?
no, your airspeed data is fine.
and the second;
no, because there is no evil corporation behind that slows down windows XP, or bricks and iSheep product because it detected that you replaced a display for 1/30 of the price.
ArduPilot is open source, running on open hardware, anyone can read the source, and see that there are no evil code to promote sales.
nor is there any reason for such, as it can run on a huge amount of hardware, including Raspberry Pi's and other hardware, there is not a single business that profits from it.
you can be sure of that with FOSS projects in general.
Second was a joke. ;-)
Sorry, it was a bad one...