Fly away, log file and Yuneec's incompetence/bad faith

[record.play.rewind]

Thanks, what do you mean with entry and exit? These are not separate flights, it's just that radio contact was lost in between and apparently when that happens, a new file is created when contact is restored. Are you able to "decode" the stick input? My understanding is that in the last part of the flight RTH was activated but the craft was still not heading to the remote.

What is Yuneec CS?

Gotcha. Holy smokes, from 00114 it traveled 26,752ft and from 0015 to 0017 7,477ft (Google Earth estimated distance) and your voltage when disappeared was at 14.2v
You can also read your cvs telemetry coordinates via Google Earth.

CS = Customer Service

 

[ukaleq]

Again 114 is is a normal flight that took place before. I included it to show that everything was still fine at that time. The copter was powered off and transported by car to a new place where the last flight started. The last three files correspond to the flight when they flyaway occured.

 

[DoomMeister]

ukaleq,

I hope this helps you. I haven't delved into the Sensor files yet and am not sure if they are the ones to match the time stamps of the telemetry files. From what I am seeing in the telemetry there was a switch to Manual mode (Angle mode without GPS) shortly after takeoff on flight 00115 and does not appear to be pilot initiated. The TH+ seemed to quit responding to your inputs and a few attempts at using RTH were made to no effect. I am attaching a bunch of files to this post and they are summarized in the text file.

The following images are screen captures from Google Earth that I plotted your telemetry in from kml files generated in Q500log2kml.exe (a program developed by @h-elsner and available at Meine private Homepage - Programmierung, Kopter, Bilder in the downloads section).


good flight at top of the valley


start of fatal flight


reacquisition of signal











The following are plots from Q500log2kml showing ST-16s inputs and altitude telemetry:

 

[DoomMeister]

Here are the rest of the Q500log2kml plots.

 

[DoomMeister]

I have checked the Sensor files you posted and all of them were from an area near the airport at Qeqertarsuaq. The numbering system for the Sensor files is a pain with the TH+ as they don't match the numbering for the rest of the Telemetry files.

I thought I should give some clue as to the values for the plots above:

CH0 is J1/Throttle - 2048 is neutral, values greater call for ascent and lesser call for descent, a value of 0 is pressing of B3 Motor Start/Stop switch.

CH1 is J4/Roll - 2048 is neutral, left bank <2048< right bank

CH2 is J3/pitch - 2048 is neutral, back <2048< forward

CH3 is J2/yaw - 2048 is neutral, left turn <2048< right turn

CH4 is Flight mode switch, 3412 is Sport/Smart, 2048 is Angle, 683 is RTH (RTL on TH+)

If you have access to any of the other sensor files that match the date and time of the fatal flight post them, they may contain info on the switch to manual non-GPS control mode.

 

[h-elsner]

From the sensor files it claims sometimes compass errors:

D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00079.txt
19 INFO: Data link #0 lost.
20 CRITICAL: ALL DATA LINKS LOST.
21 INFO: (1,7)Warn geomagnetic field is disturbed.
22 INFO: (1,7)Warn geomagnetic field is disturbed.
23 INFO: Data link #0 regained.
8151 INFO: (3,51) BEGIN TO RTL
8152 INFO: (2,87) not angle mode obs not available
8681 INFO: (1,7)Warn geomagnetic field is disturbed.
8801 MAVlink Datensätze


D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00080.txt
1 MAVlink Datensätze


D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00081.txt
15 INFO: (1,7)Warn geomagnetic field is disturbed.
16 INFO: (3,26)ANGLE MODE
17 INFO: (3,81) obs open
18 INFO: Data link #0 lost.
20 CRITICAL: ALL DATA LINKS LOST.
21 INFO: Data link #0 regained.
1594 MAVlink Datensätze


D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00082.txt
15 INFO: Data link #0 lost.
16 CRITICAL: ALL DATA LINKS LOST.
17 INFO: Data link #0 regained.
6210 INFO: (2,86) altitude Less than 1.5m, obs not available
6401 CRITICAL: NOT DISARMING: Not in manual mode or landed yet.
6451 INFO: Landing detected.
6452 INFO: Disarmed by RC.
6833 INFO: (1,7)Warn geomagnetic field is disturbed.
7606 MAVlink Datensätze


D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00083.txt
34 INFO: (3,26)ANGLE MODE
35 CRITICAL: CAUTION: Avionics power low: 4.63 Volt
36 INFO: Armed by RC.
37 INFO: (3,14)home: 69.2387545, -53.5484411, 33.26
39 INFO: [logger] file: /fs/microsd/log/2018-08-17/16_18_3
779 INFO: (1,7)Warn geomagnetic field is disturbed.
1290 CRITICAL: NOT DISARMING: Not in manual mode or landed yet.
1334 INFO: Landing detected.
1335 INFO: Disarmed by RC.
1371 CRITICAL: PREFLIGHT FAIL: EKF INTERNAL CHECKS
2167 MAVlink Datensätze

Nothing about RTH, I do not know what the last critical message means, hope Yuneec will know.

Maybe this caused the fly-away.
It was in RTH and should come back but it didn't. Last speed was 5.8m/s.

This is the path over ground, 3 parts.




br HE

 

[ukaleq]

Thank you very much. I was flying my own payload, which I did many other times and in fact most of the time, but I suspect that the sensor files might come from the 5.8ghz link of the camera because I zipped the most recent ones and I recall the last time I used the c23 camera was in qeqertarsuaq so that would explain why no sensor files are available for the last two flights (telemetry 114 and 115-117).

Yes I remember I got a number of those geomagnetic errors but the flight behavior was normal so I didn't pay much attention to them. Oddly enough I don't think I got that kind of message on the day of the flyaway.

The dealer who sold me an h520 (and who is a 'yuneec commercial partner') got in touch with yuneec about my problem with the typhoon h plus and apparently they are now willing to investigate but they maintain that the telemetry files are different from those of the previous models so they need some help from their Chinese headquarters. It sounds strange given that they are csv files and that you are able to read them with a third-party software.

 

[ukaleq]

From the sensor files it claims sometimes compass errors:

D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00079.txt
19 INFO: Data link #0 lost.
20 CRITICAL: ALL DATA LINKS LOST.
21 INFO: (1,7)Warn geomagnetic field is disturbed.
22 INFO: (1,7)Warn geomagnetic field is disturbed.
23 INFO: Data link #0 regained.
8151 INFO: (3,51) BEGIN TO RTL
8152 INFO: (2,87) not angle mode obs not available
8681 INFO: (1,7)Warn geomagnetic field is disturbed.
8801 MAVlink Datensätze


D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00080.txt
1 MAVlink Datensätze


D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00081.txt
15 INFO: (1,7)Warn geomagnetic field is disturbed.
16 INFO: (3,26)ANGLE MODE
17 INFO: (3,81) obs open
18 INFO: Data link #0 lost.
20 CRITICAL: ALL DATA LINKS LOST.
21 INFO: Data link #0 regained.
1594 MAVlink Datensätze


D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00082.txt
15 INFO: Data link #0 lost.
16 CRITICAL: ALL DATA LINKS LOST.
17 INFO: Data link #0 regained.
6210 INFO: (2,86) altitude Less than 1.5m, obs not available
6401 CRITICAL: NOT DISARMING: Not in manual mode or landed yet.
6451 INFO: Landing detected.
6452 INFO: Disarmed by RC.
6833 INFO: (1,7)Warn geomagnetic field is disturbed.
7606 MAVlink Datensätze


D:\temp\Flight2Log_ukaleq\Sensor\Sensor_00083.txt
34 INFO: (3,26)ANGLE MODE
35 CRITICAL: CAUTION: Avionics power low: 4.63 Volt
36 INFO: Armed by RC.
37 INFO: (3,14)home: 69.2387545, -53.5484411, 33.26
39 INFO: [logger] file: /fs/microsd/log/2018-08-17/16_18_3
779 INFO: (1,7)Warn geomagnetic field is disturbed.
1290 CRITICAL: NOT DISARMING: Not in manual mode or landed yet.
1334 INFO: Landing detected.
1335 INFO: Disarmed by RC.
1371 CRITICAL: PREFLIGHT FAIL: EKF INTERNAL CHECKS
2167 MAVlink Datensätze

Nothing about RTH, I do not know what the last critical message means, hope Yuneec will know.

Maybe this caused the fly-away.
It was in RTH and should come back but it didn't. Last speed was 5.8m/s.

This is the path over ground, 3 parts.
View attachment 11521



br HE

Hi h-elsner,
You're the author of the software, right? Thank you so much! But doommeister is right, the sensor flights are from a flight the day before (I am just reading all this with my smartphone but I can see that the latitude is 69 degrees and I was at 67 when the copter flew away).

The odd thing is that if the copter switched to a flight mode without gps I should have been able to control it. There was very little wind on that day so the trajectory of the drone cannot be explained by the wind drift. I flew my h520 without gps a couple of times, and yes one has to ne careful not to let it drift too much but that was nothing compared to what happened with the th plus.

 

[h-elsner]

If compass is wrong, this will not getting better. It may lead to more problems in further flights. May lead! This is all only guessing.
At legacy Yuneec SW, compass erros were one of most reasons for autonomous fly (away) of copters. For PX4, there may be other reasons. I have not followed that in PX forums.

You are completly right, missing GPS is not the root cause here. 5-8m/s is more that simple drift, it's an "autonomous" flight (want to go back to China I guess).
And it take the RTH command and changed flight mode to RTH. I'm not sure, what is the difference between RTH12 and RTH13, what is the meaning. This is all Yuneec secret and we can only guess. At Typhoon H Plus many columns in the telemetry are not supported any more, like Error Flags, gpsAccH, gpsPosUsed is always 'false' due to SW bug and so on...

I'm still curious with the msg: CRITICAL: PREFLIGHT FAIL: EKF INTERNAL CHECKS.

You have an older SW running, thus you could not switch to manual mode which may a possible way out of the situation.

br HE

 

[h-elsner]

Now I think I found out what EKF means (see message CRITICAL: PREFLIGHT FAIL: EKF INTERNAL CHECKS. ): Extended Kalman Filter
Extended Kalman filter - Wikipedia

I guess those information should be in message ESTIMATOR_STATUS ( #230 - $E6) as ESTIMATOR_STATUS_FLAGS
common.xml · MAVLink Developer Guide

But I didn't found any of those messages. Maybe we do not see all in the Sensor files.

br HE

 

[PatR]

It was mentioned he was flying his own payload. Might the payload have caused issues?

 

[h-elsner]

Might the payload have caused issues?

No, with own payload (or with other words, without system camera) no sensor files will be written at Typhoon H / H+. But this message was found in sensor file.

br HE

 

[glider]

Now I think I found out what EKF means (see message CRITICAL: PREFLIGHT FAIL: EKF INTERNAL CHECKS. ): Extended Kalman Filter
Extended Kalman filter - Wikipedia

I guess those information should be in message ESTIMATOR_STATUS ( #230 - $E6) as ESTIMATOR_STATUS_FLAGS
common.xml · MAVLink Developer Guide

But I didn't found any of those messages. Maybe we do not see all in the Sensor files.

br HE

It appears the H+ has more 3DR lineage than I previously thought. EKF in large part is for preventing flyaways. If there were EKF warnings that is troubling. Did the controller give you the ok to take off with a pre-flight EKF warning?

 

[ukaleq]

Hi guys, sorry I didn't see you had posted more replies. Again, the sensor files are from previous flights that went well. Nevertheless, it is true that I frequently had warnings of the type "caution geomagnetic interference, fly with caution" but all flights appeared normal so I stopped paying attention to them.

What is 3DR lineage?

Regarding the payload it's just a camera weighing ca. 550grams, so that's hardly more than the original payload (C23 camera) plus the optional real sense so the typhoon should easily manage that.

I contacted the dealer who sold me my H520 and he asked yuneec, but they told him they have to wait to receive some special software from China in order to analyse the log files because they are different from the logs of previous copters. I think some of you mentioned that some people had their copter replaced after flyaways, at least in the US, so how did they do it without the special software? That sounds like utter nonsense to me. What's more, Yuneec refuses to sell the aircraft alone so I am left with a remote (that is still at their premises) and a C23 that I cannot use. Their advice? Sell it on ebay.

Horrible service.

 

[h-elsner]

3DR is a US drone company. They use flight controller on base of Pixhawk PX4 as Yuneec do at H Plus, H520 ans Mantis Q.

I have heard from some people that got a replacement H Plus here in Germany from Europe Service in Kaltenkirchen.
The problem may be that you don't have the copter anymore.

br HE

 

[ukaleq]

Oh yes ok!
Well, that's one more hint that they are talking rubbish. The question is why are they doing this with someone who bought 3 drones from them in 6 months? It's almost as if they are actively trying to alienate customers.