laceyboy Posted at 2017-5-20 08:36
Reading your DAT file in Viewer, I noticed a drop in battery voltage.

Just curiously thinking out loud regarding the voltage drop -

The drop in voltage and especially current (down to 0) simultaneously is interesting.  Just a measurement glitch ? Momentary battery disconnect or battery shutdown?  

If the battery (or battery connection) did fail momentarily, would it have been long enough to force a reboot of the AC or some of the systems in the AC?  That might explain why the motors eventually recovered.

Another March Mavic!

Griffith Posted at 2017-5-22 10:45
Just curiously thinking out loud regarding the voltage drop -

The drop in voltage and especially current (down to 0) simultaneously is interesting.  Just a measurement glitch ? Momentary battery disconnect or battery shutdown?  

This doesn't appear in CsvView/DatCon. But, it's probably a glitch because there is a glitch in other signals as well. E.g. relative_H, roll, compass_Z, etc. These signals are unrelated to the battery.

fansa74cc3db Posted at 2017-5-22 10:23
Thank you for the detailed answer. But I did not understand the most important thing:
I have a sense to send Mavic in warranty repair? Or is the incident entirely the driver's fault?

I wouldn't say it was entirely the pilot's fault. You could have prevented the incident by avoiding a launch site that is geomagnetically distorted. But, the Mavic should have switched to ATTI but didn't. The only way that can be fixed is with a firmware update.

I forgot to mention a bit of advice that I usually give for these types of incidents. Before launch look at the Go App and verify that the red triangle that represents the Mavic is pointing in the right direction. If it's not then the compass is incorrect, but, more importantly Yaw is incorrect. IMHO, this is the single most effective check that can be done to prevent a fly away like yours. Checking the compass numbers doesn't help much. Another check is to see if the GPS graph on the RC is showing 4 or 5 bars. Number of satellites isn't as important.

its nice to see we can use logs like this to help under stand what is going on in flight or crash and find out how is in the wrong

BudWalker Posted at 2017-5-22 16:26
This doesn't appear in CsvView/DatCon. But, it's probably a glitch because there is a glitch in other signals as well. E.g. relative_H, roll, compass_Z, etc. These signals are unrelated to the battery.

Except that they require battery voltage to operate.  But then they also require battery voltage to record.  I suspect you're correct that it was just a recording glitch.

Consider this:  could there have been a momentary (a few milliseconds) failure of one of the systems, not long enough to shutdown the main processor but long enough to shutdown the biggest current hog, the motor system?   Don't know, just throwing it out.

Griffith Posted at 2017-5-23 05:11
Except that they require battery voltage to operate.  But then they also require battery voltage to record.  I suspect you're correct that it was just a recording glitch.

Consider this:  could there have been a momentary (a few milliseconds) failure of one of the systems, not long enough to shutdown the main processor but long enough to shutdown the biggest current hog, the motor system?   Don't know, just throwing it out.

That's a good point. It's hard to accept this as a glitch that happens to coincide with a flight where there was an abrupt loss of power. OTOH, I've been trying to duplicate this glitch using DatCon and haven't been successful. I've used Viewer to look at 2 other flights with an abrupt loss of power and there do not appear to be any glitches like this.

I had some very strange battery glitches the first time I flew with 2 of 3 of my batteries. After further recharges everything wad ok. This glitch wad during heavy acceleration in sport mode only. Maybe it is the same issue and I was just lucky to have enough juse for the circuits. I can try to find this logs and post here.

BudWalker Posted at 2017-5-23 07:02
That's a good point. It's hard to accept this as a glitch that happens to coincide with a flight where there was an abrupt loss of power. OTOH, I've been trying to duplicate this glitch using DatCon and haven't been successful. I've used Viewer to look at 2 other flights with an abrupt loss of power and there do not appear to be any glitches like this.

If I have a fail in sport, be sure to send you the dat. It'll probably be from a shoe though from someone who rolls around with a bottle of whiskey. Don't remember complaints of them falling out the sky last year, could be qc or a recent bad batch of parts.

BudWalker Posted at 2017-5-22 16:38
I wouldn't say it was entirely the pilot's fault. You could have prevented the incident by avoiding a launch site that is geomagnetically distorted. But, the Mavic should have switched to ATTI but didn't. The only way that can be fixed is with a firmware update.

I forgot to mention a bit of advice that I usually give for these types of incidents. Before launch look at the Go App and verify that the red triangle that represents the Mavic is pointing in the right direction. If it's not then the compass is incorrect, but, more importantly Yaw is incorrect. IMHO, this is the single most effective check that can be done to prevent a fly away like yours. Checking the compass numbers doesn't help much. Another check is to see if the GPS graph on the RC is showing 4 or 5 bars. Number of satellites isn't as important.

Here is a screenshot from the GPS-Test program.
As you can see, in that place is caught 26 satellites, in use 15

Hi there!

First of all: sorry for my bad english, hope you will understand me.

I have exactly the same problem as the author of this thread.

Here are my DAT files:

I am not allowed to post links?

And here the data from my Go App:

I am not allowed to post links?

Thank you in advance - hope you can help me.

Greetings andreas

Hello,

i found a solution for my link problem:

Dropbox-Link

Hope you can open everything.

Thanks in advance and sorry for the double post - i didn't find the edit button.

Greetings Andreas