Midflight (Almost) Crash and recovery

Log Analysis -- deprecated Copter Logs -- deprecated
1

Hello All,

As I am a newbee at this forum a quick introduction.

I am Erik, and since I bought a Parrot AR Drone last year, I decided to upgrade to the real stuff, after a flyaway of the Parrot… (Never recovered the drone)

This weekend I had selected for a maiden flight of my RcTimer Spider with a Pixhawk controller, running Arducopter v3.1.2
After a solid takeoff, and a very sensitive throttle response, the quad was almost lost just after a left turn (Yaw), which are acting very aggressive (So definitely needs some tuning here)

Due to the altitude (around 10 meters) I could recover the drone before it did hit the ground, but I couldn’t explain the sudden tumbling at all (around 1:10 in the video), as there was almost no wind…

After this first flight I had a couple more, and I didn’t see this behavior anymore.
So tested altitude hold (much easier to keep the brick in the air), and RTL due to a fence breach worked perfectly, and brought the Quad within a half meter from launch location (10 Satelite GPS lock)

If one of the experts can give me a hint in figuring out why the Quad tumbled I feel a little bid more secure the next flight… :slight_smile:

Setup:
RCTimer: Spider Frame, HP2816-810KV,
ESC 45A (Opto) SimonK, Boscam HD19
Kypom Lipo 5100 mAh, Props Carbon 10x4,7
Spectrum Dx6i, AR6210 Receiver + Satellite,
PPM Encoder, Pixhawk, Arducopter v3.12

So I have attached the log, and a link to the video
youtube.com/watch?v=puEwwHd8 … e=youtu.be

2

Flips that are quickly recovered from like that are often caused by the “ESC calibration” setup step being skipped. Otherwise, it is a hardware issue such as ESCs losing sync.

3

Thanks John,

I did the 4 in one calibration bij using the Pixhawk, boot ones, with full throttle, reboot, wait for the beeps at full throttle, and bring down the throttle to low, and wait for the beeps again.
That seemed to work as the all spinup at the same time, and full throttle gives similar power over all 4 motors.

The transmitter trim was set to neutral for this calibration

Before performing this setting, I had set the throttle calibration in the setup section set up with the trim down, to be able to capture a lower throttle than usual to activate the transmitter failsave (losing connection)

Should I bring down the frequency from 400Hz down to 300 or so, to make sure they wouldn’t lose sync in the Air ?

4

Are you saying you had done the ESC calibration prior to your crash, or that you’ve done the ESC calibration now?

Do you have sync problems? If so, I don’t think changing the update frequency is the way to fix it.

5

I have done the calibration before this first flight, otherwise it wouldn’t get of the ground I think,

The drone is flying, It seems to lost sync like you suggested mid way during the flight and resync during the fall down, so I could rescue the quad.

It actually fell about 10 meters before I regained control about 3 meters above the ground, so it didn’t
Crash at that moment…

After looking at the video it seems it lost control of the right front engine as that lost power just before the tumbling…

6

Update,…

After quite a few test under load, full throttle with the quad mounted on the ground with a few heavy rocks to prevent the takeoff, it became apparent that the ESC of the right front Motor was broken…

At a certain moment after recalibration it didn’t do any more than twitches, without spinning up.

Replacement of the motor didn’t resolve the problem, so I ordered a few new ESC,

Tonight I had a new try, in a windy area, so constant updates to keep the Quad around the same spot in the air was necessary.

By reducing the Yaw parameters to rotate in a less aggressive way it was behaving like expected…

I read on a different forum that the Sync problems you may expirience with simonK flashed ESC could be very well introduced by aggressive Yaw while you maintain the same flight level.

At a later moment I switched to Loiter, to let the GPS position the drone, which worked very well.

Due to the high wind the drone was constantly correcting the position in the air to prevent a drift away.

Next time with less wind I activate tuning again, which should be the ultimate test to determine of my ESC Sync problem is gone.

If I experience this again, I reflash the ESC’s to remove the SimonK software…

In other words this was indeed a hardware problem…