Hello everyone!I've read a lot of stories of Inspire 1 flyaways and finally it happened to me. Couple of days ago I took off near my country-side house, where I had around 20 flights previously without any problems. Weather was very good. I calibrated the compass, checked IMU... everything as usual, but right after take off the camera tilted to one side a bit. I didn't notice it right away otherwise I would land immediately, because I've seen a lot of videos with that camera tilt which doesn't bring anything good So I took few photos, flew around the neighborhood and suddenly 'No Signal' appeared on my iPad screen when Inspire was right above me at 300 meters or so. I waited for few seconds hoping signal would get back to normal, but nothing happened, still no picture. At this moment I realized something bad happened and initiated RTH. As the drone was pretty close to me I expect it to come back in a minute, but it didn't! After another minute RC stopped beeping indicating RTH fail. I pressed RTH again and after some time video signal finally started to appear. The drone was coming towards me from completely opposite direction, not where I was looking for it. This was crazy, I managed to land it with 10% of battery left, so my guess that is automatically initiated RTH whereas my commands from RC didn't ever reach the drone. After looking into the video and recorded flight data I've found there were wrong compass readings. Here is a video of an accident:



What do you think, what can cause such poor communication between RC and Inspire and that camera tilt effect?
Inspire 1 Firmware v1.2.1.06

Just seeing the first part of the video, I can tell you that the problem you had, many be that you calibrated next to a huge metal tank. It's not a good idea to calibrate the compass near any metal.

SAFAERIALS Posted at 2015-6-17 00:06
Just seeing the first part of the video, I can tell you that the problem you had, many be that you c ...

You mean that greenish thing which is in front of a take off point? It's a plastic tent, I'm sure it has no affect on the drone

One thing I noticed is that you are within what looks like 5m of power lines.  This can seriously affect compass calibration, which I have seen myself when at this distance.

Did you undertake an automatic take off.  I ask this as I note the RC assisted take off message on the flight record.  I have seen this on my flight records even when I don't actually use this function and never have.  Perhaps DJI could look into this as a potential signal or problem/ conflict being noted in the system somewhere when electro interference is close by.

Your first mistake: noticing there was a problem and still taking it up to 300 metres.

The gimbal pitch / roll is heavily influenced by the compass data that comes from the FC. You can test this pretty easily: start up your aircraft and level it on a bench and then take a piece of ferrous metal and put it near the compass. The gimbal will start moving to compensate for what it is perceiving to be a change in aircraft attitude, but is actually just incorrect data from the compass module.
The fact that you started seeing bad roll data on the gimbal, based on all your other recorded data, shows there was definitely a compass issue that you needed to correct before you went up to almos 1000 feet.

The best way to takeoff is:

- Motors on.
- Take off to 2-3m (without grabbing height).
- Verify the aircraft can hold position in GPS mode. If it can't (is toilet bowling, etc) land.
- Verify the aircraft can hold position (allowing for wind) in ATTI mode. If it can't, land.
- Pitch slightly forward / back / left / right. Verify it's moving in straight lines. If it can't, land.
- Yaw left. Yaw right. If it starts drifting like crazy, land.

This all takes 30 seconds or so on takeoff; not bad for safety in a 15+ minute flight. After all that is done, raise your gear and fly.
At least you got this one back.

Do you guys really think that the total loss of signal is because of a bad compass calibration ?
I don´t  think so ...

"Tilted horizon" is a common bug and is nothing that would cause me to worry about flying my Inspire 1.
I get that in ALL my flights!

Power lines could probably affect the video feed but experienced RC pilots says that it wont have any affect on the controls.

When the RC and the air craft lose connection it doesn´t matter how many times you press the RTH button ... That just wont work!
With loss of signal the Inspire will engage the RTH automatically, after a while, if the connection isn´t restored.
If the Inspire also looses GPS it will start drifting away with the wind until GPS is restored. "Fly away"

When the LED on the remote is red instead of green, indicating no signal.
First thing thing I would do in this situation is to restart the RC and see if it binds with the bird again.

Make sure that the loss of signal isn´t only referring to video feed ... as this can be confusing in the pilot app!
If your Inspire still responds to stick input ... don´t panic ... just land it

GB44 Posted at 2015-6-17 00:33
One thing I noticed is that you are within what looks like 5m of power lines.  This can seriously af ...

Flying over power lines -- that's the only logical explanation I have in my mind.
As for auto take off -- I never actually tried it, but somehow Pilot App always displaying that message in my flight records.

SimplePanda Posted at 2015-6-17 00:34
Your first mistake: noticing there was a problem and still taking it up to 300 metres.

The gimbal p ...

Yes, seems like there was an electromagnetic interference with the power lines followed by compass error as you mentioned. That's definitely make sense for me.
Also thank you for the tips, I will definitely follow them before taking Inspire up!

w1der Posted at 2015-6-17 01:16
Do you guys really think that the total loss of signal is because of a bad compass calibration ?
I  ...

Maybe those are two separate problems: compass error was caused by power lines and something else triggered RC signal loss.
I just can't get how it could lost the signal when flying over my head. In the previous flights I reached 1.7 km distance flying in the same area with out any signal loss issues. Maybe it has something to do with solar wind?

SimplePanda Posted at 2015-6-17 00:34
Your first mistake: noticing there was a problem and still taking it up to 300 metres.

The gimbal p ...

I'm going to agree with this guys post.   I see that  he just took off.    You should hold position for 5-10 seconds after autotake off and make sure it hold position.  During that time, go into the sensor settings are good under the MOD value.  should be close to 1500.  I've had some unsuccessful compass calibrations where this value was 1200 and it was all over the place.    Check those values.    I won't fly unless its around 1500 and holds.    Just because you did a compass calibration, doesn't mean you did it correctly.   

The rest of it.   I feel you had manual control over GPS until it lost contact and GPS took over (with improper compass calibration).    It then didn't know where to go.   It went out.     Battery got low and it initiated a RTH.   Not enough battery to get completely back and then it started auto land sequence.  If you had control when that happened, did you force it back up until it cleared the trees or did you just watch it?   I don't know if the video loss was because of SD card full or if it shuts off that function to conserve energy.   Was the card full?  If not, then it shut down video to conserve juice.      My phantom did this on a long return flight.   SD card had plenty of space but video shut down.   

I think it's probably compass calibration error.  Look at 4:30 in the video with camera roll.  When you notice that, you need to shift to atti mode and land it.  You don't have good calibration.   There were tell tale signs and you continued to fly.   I'm not chasticing you, but with experience comes knowledge and you should of known to land it.

InspirelessAggi Posted at 2015-6-17 01:55
I'm going to agree with this guys post.   I see that  he just took off.    You should hold position ...

I agree that it was my mistake to takeoff and immediately fly away. I will be more attentive next time.
As for RTH -- actually it didn't start to land over that trees it was me: when I first saw the video signal came back it took me couple of seconds to realize where the drone is located, so I stopped it and while observing the area I started to descent. Finally Inspire landed in manual mode on the starting spot. SD was full for real.

SimplePanda Posted at 2015-6-17 00:34
Your first mistake: noticing there was a problem and still taking it up to 300 metres.

The gimbal p ...

As Kung Fu Panda says, simple little things at commencement of flight give reassurances that everything is well before increasing altitude and flight distance, which should be part of every flight.  

I also re-confirm home point manually and reset IOC at this point when it is in a hover above take-off point.  All very quick but simple to do, I then know the home point is recorded properly and if I use the IOC I know if I pull back on the stick I know exactly the orientation.

GB44 Posted at 2015-6-17 02:54
As Kung Fu Panda says, simple little things at commencement of flight give reassurances that every ...

I need to learn IOC better.  I need to read the manual and figure out how to set it's function for the phantom.   How does IOC work if GPS has a problem.  Does it just know overall direction?

m037op@gmail.co Posted at 2015-6-17 01:53
Maybe those are two separate problems: compass error was caused by power lines and something else  ...

Someone can correct me if I'm wrong but I'm pretty sure I've seen radiation pattern diagrams for the Phantom 3 / Inspire 1 style controllers from DJI and they have the pattern going outwards conically, perpendicular to the antenna. Page 29 of the Inspire 1 manual (optimal transmission range) also seems to suggest this.

If | is the antenna and < is the radiation / reception pattern, it looks like this:

|<

So if your antenna's are angled like this (forward and down as i'd guess they often are):

/

You're probably sending (and listening for) a lot of signal to/from the ground and very little straight up over your head where your aircraft is.

As a digression, 5.8ghz analog video transmitters / receivers (FatShark goggles, for example) are notorious for this "overhead deadzone" problem as the radiation pattern from the circular polarized antenna makes a donut around the operator with very little reception straight up (or transmission power straight down). Always amusing when you do a 1km flight without issue only to return home to a burst of analog interference at 0m distance (but 40m altitude).

InspirelessAggi Posted at 2015-6-17 03:10
I need to learn IOC better.  I need to read the manual and figure out how to set it's function for ...

To set the IOC you simply make sure you have selected the Multiple Flight Mode on the App, then once in the air, you just flick into the IOC mode then a screen will pop up and you then press to reset.

Once you have done this the App will give you confirmation that it has been set then you can return to GPS mode.  Should you then flick into IOC mode the aircraft will operate on a forward, left, right, backwards from the direction of the aircraft nose when you set the IOC manually.  

This is useful if you are flying at distance or height and you become disorientated and can't figure out manually which way your aircraft orientation is.  it just helps to quickly orientate your aircraft or assist in bringing it home.

Try it out on the simulator.

GB44 Posted at 2015-6-17 04:58
To set the IOC you simply make sure you have selected the Multiple Flight Mode on the App, then onc ...

Will do.   I'd given anything to have this on my helicopters.   Orientation is one of the hardest things for me, especially with aging eyes.

It looks like it stuck with the last signal and just kept using the last control sticks position instead of triggering RTH. Here you go, a 3000$ drone that can't be controlled