Understanding Autotune

I’ve just fitted a new pixhawk to my X5 wing and tried the autotune feature, when in autotune or FBWA mode the aircraft will barely roll or pitch, so I need to switch back to manual to make a turn and fly back all the time, I tried 3 flights running autotune but nothing seemed to change. Having read a little more I think that as I had to keep changing back to manual to turn before the 10secs had been achieved that autotune must have been resetting to the previous values all the time, do you think this is correct? Does anyone have any starting point values for the X5, I have tried the Makedrones values but the X5 does not hardly roll or pitch, at the moment the roll and pitch P value is set at .8, how high should these be?
I’m assuming that I need to get it to the point where it can be flown around in autotune or FBWA mode without needing to switch to manual to turn, then try the autotune from there, would you agree?
Thanks for your help

The basic autopilot PID parameters should allow stabilised or FBW flight. If you can’t maneouvre the plane in these modes, you have too much damping.

I would however do some trouble-shooting first, if not done yet. When you hold the plane in your hand in the mode you intend to fly, you should see control surfaces move in the opposite way to injected roll or pitch movements. If some PID parameters are set excessively high, you should see extreme movement of the control surfaces responding to small roll or pitch movements of the fuselage. From memory the P and D parameters will affect the immediate response magnitude. The I parameter should determine how the difference between actual attitude and reference attitude is reduced.

Just some thoughts to help understand. Any other thoughts out there?

I don’t have parameters for your plane.

Hi, when the plane is being hand held in FBWA mode it reacts perfectly, correcting and movements and also moving the control surfaces in line with my stick inputs, however in flight it is perfectly stable and level just barely reacts to my stick inputs, not enough to turn and fly back, is that simply a P value that is set too low?

It is probably a too low P value, but we can’t really tell for sure without a log file. It could also be a roll limit or another parameter.

Hey Tridge, Ill try and upload some of the log files later, bank limit is set at 45 deg I think. Havent had chance to test fly it this weekend yet

Hey guys,
Had another couple of test flights the other day and managed to leave it in autotune for longer duration’s, it was a very wind day and I was really pleased with how well it coped. The roll pitch had started to be altered so just going to continue with some more flights to refine it, I did notice that on my logs the is a bigger nav roll demand on left turn compared to right turns? I’ve re-calibrated my radio to see if that fixes it unless you have other suggestions. Also on the pitch inputs the values are very low even though these were full stick movements! I noticed when calibrating that it doesnt show full movement on the pitch channel until you add roll in as well, is that something to do with elevon mixing?
Thanks for your help

I’ve had a similar issue with my wing. Sice I went from elevon mixing to elevon_output everything is perfect. In elevon mixing mode it was barely turning around, and the bank angle wasn’t the same I set in mission planner. I was also able to do the autotune successfully.

OK thanks Funnel I’ll give that a try, will report back after a test flight.


do you have a telemetry receiver installed and RC telemetry enabled? I experienced the same “dead airplane” symptom by using Futaba’s FASSTest receiver. The problem was fixed by using a “dumb” Futaba 7ch receiver.