Hello Fellow Yuneec Pilot!
Join our free Yuneec community and remove this annoying banner!
Sign up

H920 lost connection (fly-away)

Joined
Mar 30, 2020
Messages
23
Reaction score
22
Age
38
Hi, may I ask you a question: I have purchased Tornado H920 (not plus) with ST24, ST12, and CGO4. During my second flight test, drone lost the connection with ST24 and flew away... It happened in an open field, when the drone was only 80 m above ground, and approx. 200 m away from the controller... Did you experience anything similar with your H920? Is there anything I could do to improve the connection (e.g. install better antenna)? Thanks!

*********************************************************************************
Summary (after discussion in the following posts):

1. For best connection between ST24 and H920, keep both antennas 90 degrees to surface of ST24 (in my case, I did not do that).
2. Let both ST24 and H920 get all satellites - turn everything ON, and let drone sit on the ground for 5 min or so (I believe I did this).
3. After first disconnect, I panicked and accidentally turned off GPS, instead of turning ON Return To Home (thus, H920 drifted with wind).
4. The response of H920 was slow, because rate knob was towards turtle. Therefore while in panic (me bursting my lungs while running through the field) it seamed that H920 did not react to ST24.
5. Drone was able to land in between trees without hitting anything (I hold down control on the ST24 while running all the time).
6. No damage neither to drone nor to other's property was done.
7. In test flights after the accident, nothing similar happened. Drone responded as it should. From FlightLog, it was concluded that H920 behaves like any other H920.

"Fly-away" happened due to my fault and inexperience.

Thanks to all, who responded at this post/issue.
 
Last edited:
Itelite makes an antenna for the ST24 or you can use the same one as the ST16. There was a similar question recently but I can't find it now.
Can you copy the flight log from the ST24? That would provide better information about that flight.
 
Hi Steve, thanks for your replay. I located FlightLog folder in ST24. Could you please direct me either to information how to read log, or what to look for?

Also, I thought too that there should be some information online regarding fly away of H920. However, I could not find it yet..

After the first shock is gone, I try to remember how it happened:

1. Drone was at 80m altitude and some 200 m away from me. I could still see it quite well.
2. Promptly I got notification on ST24 screen that "Connection lost" or smth like that.
3. After this, drone started to fly to the right.
4. I got scared and started to run towards drone, trying to minimize the distance, hoping connection will be re-established.
5. I was able to shorten the distance somewhat, however, drone did not reconnected with ST24, because I was time to time (while running through the field) trying to press down on ST24 to lower drone altitude.
6. After some time (approx 2 min.) drone started to descent into trees.
7. When I finally came to it, I found it landed on the ground. All intact. Somehow, it missed trees and high power lines...
8. The landing feet was on the ground, however, I do not recall that I put them down while was running. Meaning that drone itself landed as it should. If so, question remains - why it did not fly to HOME? Could it be that GPS is not properly calibrated? However, in the manual it says that you do not need to calibrate GPS unless the appropriate light is blinking..
 
@skorius ,

200m is an awfully short distance to lose RC connection with the aircraft. I am curious how you have the antennas oriented on the controller. Both antennas should be perpendicular to the face of the controller, especially the 2.4GHz whip on the left. If it is pointing at the aircraft, you are aiming the dead zone at it resulting in very poor signal reception.

The actions exhibited at connection loss are very much reminiscent of poor GPS acquisition for the aircraft, controller, or both. The flightlogs should be able to tell us. I will get a procedure for you and post it soon.
 
DoomMeister, thanks! This is a super good point, regarding the antenna orientation. Honestly, I did not pay too much attention to that, and I believe the long-slim antenna although was not pointing to the drone, but it was not 90 degrees to controller either (more like leaning to the side). Thanks to you, now I know.

With that being said, should not drone return to home (from where it took off), when the connection is lost? In my case, it flew to almost opposite direction.

Thank you for your help.
 
DoomMeister, thanks! This is a super good point, regarding the antenna orientation. Honestly, I did not pay too much attention to that, and I believe the long-slim antenna although was not pointing to the drone, but it was not 90 degrees to controller either (more like leaning to the side). Thanks to you, now I know.

With that being said, should not drone return to home (from where it took off), when the connection is lost? In my case, it flew to almost opposite direction.

Thank you for your help.
If you do not have a good GPS fix on both the aircraft and the controller the RTH function will not work properly. The action taken by the H920 also leads me to believe the battery was marginal and with no connection to the controller that was not reported. It appears to have executed a low battery failsafe emergency landing. Thankfully avoiding obstacles in the process!!! These are speculation on my part.

Download the Way to Fix Drones Project pdf file in this post Way To Fix Drones project and go to the attachment for retrieving the FlightLog for a Typhoon H. I am quite certain the procedure should be very close if not exactly the same on the ST24. You can then upload the resulting zip file by using the Attach files button in a reply.
 
I have attached the FlightLog. I noticed that the date when the files were created is 2013..This is wrong. I have assumed that date will be corrected automatically, when GPS gets connected. At this point, I think the following:

1. If ST24 does not automatically catches the time when GPS is connected, then the last LOG should be the one, when drone lost the connection and flew away. I did not fly drone after the accident.
2. If ST24 does correct time automatically when the GPS is connected, then it means latest log files were not created.
3. Accident happened during the first and only flight that day, after approx. 4 min of flying (I know this from the continuous CGO4 recorded video, which I started when I started the drone). Because batteries are completely new and freshly charged, I do not think this happened due to batteries running dry (I used 3 batteries).

At this moment, I recognize my stupidity:
1. I did not fold antennas as suggested for optimal connectivity (could be the reason for fly away).
2. I did not set proper time on ST24 (I do not think this could be the reason).
 

Attachments

  • FlightLog.zip
    419.9 KB · Views: 16
Maybe I'm missing the obvious but I had 2 fly aways back to back. It is only my opinion that they were due to GPS issues. Did you let the 920 sit for 15 minutes or more before flying? If the GPS database isn't substantiated it will have wrong date, location and GPS coordinates.

If my drones sit for a long time or I travel 200 miles to fly I use a low battery and let the drone sit for 15. The 2 fly aways back to back burnt a hole in my brain to let them sit for 15 minutes.

Both flyaways were the same. I put the bird into a level hover at 8 to 10 feet and then start to maneuver them around. After hover and the first stick command the bird took off with no control. The logs showed that I was off the coast of Africa I think and that RTH was initiated but I think it was trying to return back to the congo.
 
  • Wow
Reactions: DoomMeister
Interesting, thoneter. Before flying, I have tried to connect ST12 to CGO4 and ST24 to drone. This attempt took me indeed 10 to 15 min. However, it was not successful. So yes, drone was on for 10-15 min. But after failed attempt to connect ST12, I turned H920 off, and did the re-binding with ST24. Everything worked as it should, and after less than 5 min I was already in the air with ST24.

So, you suggest to turn drone and ST24 on, and let it sit for 15 min before flying?
 
The logs showed that I was off the coast of Africa I think and that RTH was initiated but I think it was trying to return back to the congo.
For what it's worth, this is GPS "Ground Zero". (Where the Greenwich Meridian crosses the Equator)
Ground Zero.JPG
I suspect this comes into play for certain fly away events. If the drone has an accurate fix on where the controller is, but the drone itself "thinks" it is off the coast of Africa, it sets up some issues. One would be if you go to "RTH". I suspect it will start moving in a line that it "thinks" is moving towards the controller. In actuality, it would be moving in the opposite direction from "Ground Zero" (and irrelevant to where you are standing}. It would probably continue in that direction until it looses connection with the controller, or the battery dies, and eventually auto-land.

At least that's my theory. Never heard anyone that actually knows discuss it.
 
  • Like
Reactions: thoneter
Hi all,

I have tried to plot Telemetry and RemoteGPS coordinates. This what I got:

1. The location where drone started and landed is correct. I did not find any spots to be in Africa, or somewhere else. Everything from Long and Lat coordinates, which I found in ST24, is correct.
2. The red dots are from Telemetry *.csv files and blue is from RemoteGPS.
3. Red dots show where drone was flying.
4. Blue line shows either were I was running to pick the drone (with ST24 in my hands), OR where I was walking back with drone and ST24 in my hands. This I cannot say at the moment (need closer look at data).
 

Attachments

  • Telemetry and RemoteGPS data.png
    Telemetry and RemoteGPS data.png
    1.9 MB · Views: 34
  • Like
Reactions: thoneter
Interesting, thoneter. Before flying, I have tried to connect ST12 to CGO4 and ST24 to drone. This attempt took me indeed 10 to 15 min. However, it was not successful. So yes, drone was on for 10-15 min. But after failed attempt to connect ST12, I turned H920 off, and did the re-binding with ST24. Everything worked as it should, and after less than 5 min I was already in the air with ST24.

So, you suggest to turn drone and ST24 on, and let it sit for 15 min before flying?
If it's been a while sitting in mothballs I do let them get reacquainted with the satellites. Sometimes I wonder if it's voodoo though but it doesn't hurt to do it as I use a spent battery.
 
@skorius,
My apologies. The post above was related to @thoneter's comment, not related to your event.

The outstanding item I see in your data is very poor signal strength between the controller and the drone. I suspect this will show to be a large contributor to the event, but without input from more experienced data readers, I would not put too much faith in that yet. And it does not, by itself, explain why the H920 was moving. A simple loss of signal should have resulted in the drone holding position. And it did not.
 
Hi skorius,
It would be also nice to know how long you have had the H 920 for and your experience and whether you have read all the manuals. It seems that you have not as you did not realise that when you hit the RTH switch the 920 goes up to a predetermined ht and puts its legs down to land automatically when it gets back to the landing area near your controller. I am very puzzled why you would risk such an expensive machine without doing a compass calibration or a GPS warm-up for your area. I would suggest that you do a lot of reading the manual and read it again and fly without the camera as your camera is very expensive to fix if you have a close association with an immovable object. when you have done this for a while and you are confident keep practising until you are competent. There are heaps of information threads on her for you to scrub up on for your flights and learning experiences. Wishing the best of safe flying and have lots of fun times with the 920. Please don't take this the wrong way, you can never learn too much about flying these MACHINES.
Johnno Hennessy.
Keep flying on the GREEN side of the GRASS.
 
  • Like
Reactions: Steve Carr
@skorius ,
Here is what I see. Although control signal was often weak, there we only two periods of actual control loss. During the "drift" across the field you had control, but response seems very slow. Here are some charts and some notes of what I see. This is primarily a low signal condition. There is either a problem with your controller transmitting the signal, or with the drone receiving it.
Flight review.jpg
 
  • Like
Reactions: Steve Carr
Hi stories,
With all the documentation coming out, I think that you would probably need to clean the sticks on the st24 , have you checked it in the hardware monitor, how long has it been since the 920 was flown before you purchased it. You need electrical contact cleaner and search the threads about how to do it. DoomMeister will probably give you a link as I would not have a clue how to do it. It sounds awfully like dirty sticks to me, my two bobs worth. Dont waste your money on a antenna range extender as you do not need it and put the time into maintainance on your machine .Johnno Hennessy.
 
Last edited by a moderator:
Hi skorius,
It would be also nice to know how long you have had the H 920 for and your experience and whether you have read all the manuals. It seems that you have not as you did not realise that when you hit the RTH switch the 920 goes up to a predetermined ht and puts its legs down to land automatically when it gets back to the landing area near your controller. I am very puzzled why you would risk such an expensive machine without doing a compass calibration or a GPS warm-up for your area. I would suggest that you do a lot of reading the manual and read it again and fly without the camera as your camera is very expensive to fix if you have a close association with an immovable object. when you have done this for a while and you are confident keep practising until you are competent. There are heaps of information threads on her for you to scrub up on for your flights and learning experiences. Wishing the best of safe flying and have lots of fun times with the 920. Please don't take this the wrong way, you can never learn too much about flying these MACHINES.
Johnno Hennessy.
Keep flying on the GREEN side of the GRASS.

Hi John,
thanks for your comments. I will try to answer point by point:

1. I bought H920 in 2019 October. I was not able to test it right away, which I did only in 2020 March. I did read manuals and watched some youtube videos on how to set up drone for flying.
a) During first drone test (without CGO4), I tried RTH - which worked as it should. I was flying until battery (2 of them) got to 60%, approx. 15 min or so of starting, flying, landing, starting, flying, etc.
b) During second drone test (with CGO4), I also tried RTH - which worked fine as well. Although wind was strong, all landing and taking off went smooth. No problem at all.
c) During third test (with CGO4), everything started as normal. But then fly away happened.

From what I mentioned above, I believe I red manual as much as it was needed for me to be able to fly. My question is two-fold: i) why connection was lost (it could be the orientation of antenna -my fault); and ii) why after loosing connection drone flew away, instead of keeping the position (some suggest battery, however, I used 3 new, fully charged batteries (ST24 showed it to be full during fly away).

2. You said you were puzzled why I did not do GPS calibration. Well, in the same manual it says that it is not needed if drone itself does not ask for it. Therefore, I followed manual, and did not do it, because drone was not asking for it (did not blink specific code). As per GPS-warm up for area. I do agree that it is a good thing to do. However, I always kept drone on the ground for approx.5 min before flight. Also, ST24 showed above 10 satellites to be connected.

In the future, I will be cautious about these aspects. Thanks!
 
Hi all,

I have tried to plot Telemetry and RemoteGPS coordinates. This what I got:

1. The location where drone started and landed is correct. I did not find any spots to be in Africa, or somewhere else. Everything from Long and Lat coordinates, which I found in ST24, is correct.
2. The red dots are from Telemetry *.csv files and blue is from RemoteGPS.
3. Red dots show where drone was flying.
4. Blue line shows either were I was running to pick the drone (with ST24 in my hands), OR where I was walking back with drone and ST24 in my hands. This I cannot say at the moment (need closer look at data).
@skorius ,
Here is what I see. Although control signal was often weak, there we only two period of actual control loss. During the "drift" across the field you had control, but response seems very slow. Here are some charts and some notes of what I see. This is primarily a low signal condition. There is either a problem with your controller transmitting the signal, or with the drone receiving it.
View attachment 20983

Dear WTFDproject,

thank you a lot for such detailed analysis! It really seems that a lot of stuff happened during this incident. Thanks to recorded data, we can analyze the situation. Now, with all the information that we have, I could try to conclude:

1. I must have done GPS calibration, although drone was not asking for it. I purchased drone from Germany. GPS calibration was not done after purchasing.
2. After powering everything up, I should have left drone to sit 10 min on the ground, to "to get used to the environment".
3. Most importantly, ST24 antennas had to be perpendicular to the face of the controller. Which in my case most likely were not..
4. Should reduce drone speed somehow, that next time I could catch it up.
 
  • Like
Reactions: thoneter

New Posts

Members online

Forum statistics

Threads
20,955
Messages
241,587
Members
27,284
Latest member
csandoval