Unwanted max throttle limit in auto mode

Hi all. First post here (be nice :))
Anyway we have an issue with a pixhawk (3rd) that we cannot figure out the reason and how to solve.

The thing is that when switched to auto-mode and take off command, the AP only puts out 90% throttle. I also obvious from the failed launches that we have had.

The curious thing is that all parameters as set correctly (as far as we know) and that ins FBWA mode the throttle and ch3 output (controlled from RC), goes to 100%. Also the motors spin to max. We can even see difference in wind meter we put behind it.

In the parameters we don’t have anything limited. Here are the obvious 2 that come to mind.
thr_max and tkoff_thr_max . Both are set to 100%.
I am happy to share save parameters if it any help.

Frame is a fixed wing obviously. MyTwinDream, with twin motors.

Some more background information, incase it could help. I build the plane a tuned it for someone. It was perfect. Then I shipped it to him and in the journy the airspeed sensor tube got damaged. So we ended up disabling air speed completly. Again I do not know how this would mater. As I said in static ground testing the auto mode goes ful throttle to only 90%.

I would really appreciate any hints, ideas to fix this. Or at least where to look…
Thank you.

Did you try re-calibrating the controls?
Are you sure the esc’s are calibrated?

1 Like

Yes we did. And yes ESC is not and issue (it is calibrated). I would have thought that was basically obvious from the fact that FBWA mode the throttle is fine.
Also the telemetry (AP) itself reports that throttle out as only 90%.
That is when I select to show Ch3percent. If ESC was and issue then it would show what it believes as 100% to be 100%.
So in other words it knows that it is limiting it.

1 Like

Has this found any ideas yet???

You may want to try resetting your parameters. If you set a parameter such as THR_MAX it will stay until reset.
How do you know you’re only at 90%?

Actually it was always set to 100%. So we did not lower it then put it back.
However we did reset. We did full wipe and reinstall (we installed copter firmware then put back arduplane, this to make sure EPPRON is fully wiped).

We nor anyone else has a solution for this. I lost us a lot of time on a project and countless broken props.

How do we know that it is only 90%.

  1. We output (on mission planner HUD) the % out output of ch3 (which is the throttle). Also the value of Ch confirms this.
  2. In out ground test we even measured the wind power from props is auto mode.
  3. Plane crashes to the ground on launch due to being heavy and under powered.

Again to repeat, the issue is so weird, but yet intentional from pixhawk/arduplane. We did not change anything in the parameters. We just setup a camera trigger on one of the aux channels (ch10 I think). Then the next launch after that it started this behavior. We even rolled back to older parameters and still same. As if it has something to do with hardware, meanwhile that is impossible.

So what we get is. If we switch to stab/FBWx or manual mode, and push the trottle to max it will go full power 100% max trust. So both the HUD will show 100% and plane is full powered and will rise at launch. However is we switch to auto-mode (take off instructions) it will push throttle only to 90% and stay there. So if we launch like this it will be under powered and fly to the ground. There is no air speed sensor, it is also disabled. So nothing is limiting. This stay the same for 10 tries now.

Before you ask, you should know that even on ground test with RTL turned on (props off for safety) it will only push ch3 (throttle) to 90% of what it knows to be max.

Again we have re-calibrated RC, which did not help. It has nothing to do with RC it is a problem solely in auto mode and as I sated before with RC we get the max throttle.

It is such a scary problem and it is driving us crazy. No one has any idea.

Can you post a dataflash log? Without this its very hard to diagnose any issues.

I wish you or anyone else asked me this a few days earlier.
We only have tlogs and parameters available to post now. The plane is in another country. The friend that was using it had to travel overseas, after more than 1 week of continuous failure and no help or advance. Won’t be back to that particular plane for a few months.
Can the telemetry logs not be of any help?

So again to be clear, we had 100% throttle in auto up to a point. I was always watching every launch via remote access to the ground control PC running Mission planner.
Then after about the second day. This was the 3rd launch. This thing started to happen. At first we had no idea what was going on and why it always ended up on the ground right at launch. Then after going over the tlogs a few times I realized the issue. However I haven’t the slightest idea why it was doing it.
We tried everything.
I am sure you will agree that it is a software issue, not a hardware one. However it makes no sense.
I even compared the parameters line bybline to the save that I made before shipping. Apart from the gyro and rc recalibration, there was not a single difference. And even with that we rolled back to the version of the parameters saved that did work the first 3 launches (that is did not have the 90% limit problem).
It almost seems like a virus which I know is not a possibility.

Tlogs are definitely a starting place. The bins would have more information, but there might be enough available in a tlog to start investigating.

Th naming convention used is really bad, i’m trying to dig down the tlogs. Also I’m trying to replay them myself so I know I get the right one, but fast forwarding does weird things. I will continue to dig, and see if I get at least one of them.