Disaster going to Autotune

After 7 months of inactivity I finally got back to investigating my CRASH of a RANGER EX and Pixhawk.
( I am now starting over with a new airplane; BIXLER 3 )

I set up my RANGER as I thought was best but apparently it was not.
It had Arduplane 3.8.0 beta6 loaded.
On my 3rd flight and first attempt to going to Autotune, according to my reading of the LOG, the Elevator and Ailerons went totally wild resulting in a very quick destruction of the entire thing.

I am not sure which way to look at RCIN and RCOU, which way is meant by each relative to the receiver or the Pixhawk?
Where do I begin analyzing the Log?
I do not want to repeat this.
Were my PIDS all wrong to start with?
The LOG Analysis Routine did not help me.
How do I attach the LOG file to this Topic?

Thanks for any help &/or suggestions

The steps before autotune IMO are

  • fly in Maual mode
  • fly in FBWA mode

If the plane is flying fine in FBWA mode there is no reasons with default parameters that you might have problems flying in Autotune.

Defaults Pids are quite prudents so you should not care about them in the first flights in FBWA mode.

Lucamax,
Thanks for you notes.
Can you help me with the LOG file upload question?

Thanks

PS:Anyone else out there want to chime in??

It’s likely that your control surfaces were double reversed - making the control normal in manual mode, but destabilizing in any stabilization modes. I’m assuming you didn’t do a FBWA check on the ground.

Nathan,
That is a very intriguing idea. It never occurred to me that would happen.
I thought I had done a FBAW check but I have no way of knowing for sure now.

On the new airplane I will do that test for sure; several times until its right.
Do you have any other good things to check?

Thanks

If the log size is small , less than 1 Mb, just drag it in the comment window , if you receive a message saying it is too large , you have to paste a link to a storage service as Dropbox where you uploaded your log file.

Here is the link to a Dropbox version of my LOG of my mysterious CRASH.
If someone could have an analytical look and comment, it would be very much appreciated.
Thanks

RCIN is related to the receiver , RCOUT is the pwm signal from the Pixhawk to Esc and servo.

I’m not an expert about log analysis…

  • Only Manual mode is recorded in your Log that might be corrupted, cannot see when you engaged Auto Tune mode.
  • You were flying quite fast around 100 km/h
  • At second 161 your RCIN.C2 and RCOUT.C2 diverge , guess it is your Roll

Here is a screen shot from my PC. It clearly shows AUTOTUNE was engaged at the 120 mark just after a couple of seconds of LOITER.
I used Mission Planner to produce this graph.

Does this help out.

Can anyone else take a look at this graph please.

That log file only has Manual modes in it, changed frequently.
You also seem to have RCin2 and RCin5 mixed in your transmitter, why?
Have a look at the RCin’s from the log, they are all over the place.

HI Mike,
Thanks for responding, I appreciate the help.

I am very confused. I can clearly see the word AUTOTUNE with word LOITER just above it in my black graph above. This at roughly the 120 marker.
I just reran the graph using the log on my Dropbox called Cal-RNG3.log which I uploaded to here a few days ago.
As a validity check, my log has 126,763 lines in it.
I was thinking that Dropbox may be clipping the file length so for a second check, I downloaded my log from the forum and ran the very same graph. It looks perfect.
Please help.

Mike,
The RCin2 and RCin5 are my left and right aileron channels and on my graph are in dead sync as I would expect.
Cheers

I can confirm that the .log file downloaded from DropBox is a text file with 126,763 lines.

I also can open it with Mission Planner, and see the AUTOTUNE mode switch on line 119661. (TimeUS = 161276420, ModeNum=8)

Hunt0r,
Thanks for that.
Can you see what happened to cause my crash?

No, sorry, I’m not practiced at diagnosing crashes