Rockin' and Rolling - not in the good way

Hi All,

Some background info on me quickly; I have been building my drone with a friend over the past few weeks and boy there is a lot to learn! We haven’t done that badly with only a couple of crashes under our belts. I introduced my drone to a tree pretty early on! :slight_smile:

I want to pop a mobius gimbal on it shortly but need it running right before I do, I did complete a successful autotune yesterday and did confirm different PID results in Mission Planner.

I’ve locked down my design on the S500 frame and I did my second FPV flight in drift (Airplane) mode today. Upon reviewing some the logs through the graphing, I noticed the pitch commanded vs pitch looked good but it got the wobbles occasionally on roll. I could really induce them passing around 50kph (13.8 metres per second) and it was very noticeable while flying in FPV, hearing the props really working to stablize the drone during turns. A point of interest is in Drift mode when I released the stick while flying forward, it quicked raised the nose and washed of the speed nice and quickly. Upon completing some orbits around the field it would rock left and right quite a bit through the turns and on the straights.

I was wondering as a quick check, would someone be able to review some of the key information in my logs and point me to any shortfalls I might have, not necessarily roll (but would be nice) but also generally? I did notice a GPS error in the auto analysis of the logs but checking the info, it looks to be at the end of the flight?

Logs and files: HERE

Build

  • Holybro PX4 2.4.6 running 3.5
  • S500 Frame
  • 10x4.5 Props
  • Multistar Elite 2810 750
  • 30A Mini ESC w/ OPTO Afro race spec
  • ZIPPY Flightmax 5000mAh 4S1P 20C

Auto Analysis of logs

Duration 0:21:59
Vehicletype ArduCopter
Firmware Version V3.5.0
Firmware Hash 633501f9
Hardware Type
Free Mem 0
Skipped Lines 0
Test: Autotune = NA -
Test: Brownout = GOOD -
Test: Compass = FAIL - Large change in mag_field (135.16%)
Max mag field length (784.22) > recommended (550.00)

Test: Dupe Log Data = GOOD -
Test: Empty = GOOD -
Test: Event/Failsafe = GOOD -
Test: GPS = GOOD -
Test: IMU Mismatch = GOOD - (Mismatch: 0.45, WARN: 0.75, FAIL: 1.50)
Test: Motor Balance = WARN - Motor channel averages = [1408, 1540, 1421, 1535]
Average motor output = 1476
Difference between min and max motor averages = 132
Test: Parameters = GOOD -
Test: PM = NA -
Test: Pitch/Roll = NA -
Test: Thrust = NA -
Test: VCC = GOOD -

Default

Autotune

The current design, popped 2 props on to show what they cover on the frame

The Flight

Thanks again guys, apologies if I missed uploading something! Really enjoying this new hobby.

Those PID’s seem low to me, for that size copter. I’ve seen other copters that size have roll/pitch P and I around .07 and .9.
How did it fly w/ the default PID’s compared to the auto-tuned ones?

Did you have any problems during the auto-tune? Was it windy or drifting all around or anything?
I’d either try manually tuning a bit, or I’d do another auto-tune.

Be sure everything is fastened tightly to the frame and there isn’t much wind and auto-tune again.

1 Like

Hi wicked1

Thank you very much for having a look!

It was a little windy and it was drifting which required me to correct it, but I thought that was ok (if not excessively windy) where it would continue auto tune after I release the sticks?

I will run the auto tune again today and see how I go.

In reference to your question, I think it does fly better now than it did before… I think the pitch is better but the roll really struggles.

Could prop wash over the pixhawk cause an issue as well?

Thanks
Sean