What is the possiblity I stumbled upon (at least a partial cuase) of the DJI Fly-Away

[coreyperez]

What is the possiblity I stumbled upon (at least a partial cause) of the DJI Fly-Away

I was recently the recipient of a 14SG and used the two links below to setup my radio to my Naza Mv2.

http://www.rcgroups.com/forums/showthread.php?t=1973186
https://www.youtube.com/watch?v=NJ8lkBwu_AY

The Video was a 90% solution, where as the RCGroups link gave me the failsafe information. I thought every thing was good and went on my merry way. When I went to fly the MR the first time I couldn't get the motors to arm, so back to the assistant. Long story there was the Elevator was backwards, quick fix there had the motors spinning up. BUT...

On a whim I was showing a friend with the FrySky Taranis the remote failsafe-Futaba (power off/lost connection while connected to the Naza Assistant). In doing so, I had my AIL pushed left (merely on accident, when I hit the power button I noticed that the controls stayed in the "unusual attitude", (Last input). It is my belief that we (me anyway) would want my craft to come to stick/control neutral AND throttle at half/(maintain hover). My MR would have maintained last known power setting, BUT.. would have appeared to continue to "Fly-Away", at least before the failsafe kicked in.

I don't know, this MAY be the way other's would want to have their MR respond, but to me it seems exactly what we do NOT want them to do. I had to go back in to my TX and set each channel (affected) to F/S (Fail Safe) and get it adjusted properly.

I'd prompt anybody (At least those with a Futaba) to connect to the Naza Assistant, put their craft in a turn/climb, etc and kill the power on the TX to see what the assistant is going to have the MR actually do. Mine would have just kept flying in some weird direction for however long it took to kick the fail safe in. If you are expecting an immediate halt and don't get it, you start flipping fail-safe, GPS, ATT switches, power button, etc. all the while the MR is going to keep doing what it is told! It would have been your fault for the fly-away. You TOLD it to keep going, no stopping with your incorrect fail safe settings. This would be continued because you keep cycling power (failsafe), etc!!!

Keep in mind, I JUST graduated from the Turnigy 9XR (FrySky module/RX), the RX in that case was much easier to set, when the F/S was "programmed" it took exactly whatever setting the TX was sending at that time. My sticks were center (Throttle inc.), GPS activated with F/S mode selected, and acted exactly as such when the power off F/S was initiated. The Futaba was not the case. I had to go in and individually set each channel. Other radios I cannot confirm either. My radio/RX was the 14SG and R7008SB.

Open for discussion!

Corey

 

[Bartman]

................I'd prompt anybody (At least those with a Futaba) to connect to the Naza Assistant, put their craft in a turn/climb, etc and kill the power on the TX to see what the assistant is going to have the MR actually do. Mine would have just kept flying in some weird direction for however long it took to kick the fail safe in. If you are expecting an immediate halt and don't get it, you start flipping fail-safe, GPS, ATT switches, power button, etc. all the while the MR is going to keep doing what it is told! It would have been your fault for the fly-away. You TOLD it to keep going, no stopping with your incorrect fail safe settings. This would be continued because you keep cycling power (failsafe), etc!!!
........................

that's a good point Corey, what happens when failsafe kicks in and the helicopter isn't nice and level like it was on our bench when we tested and set it up the first time?
only caveat/caution....remove the props before powering it up, holding it in your hand, and testing stuff. there's so much potential for injury with these things it can't be said enough, remove the props....remove the props.....remove the props

 

[RTRyder]

Most modern radios have two failsafe modes, maintain last stick position and full channel failsafe. The difference is a multirotor needs to have full channel failsafe set for proper RTH to work when the radio is switched off as the default on most RX is hold last stick position. Binding for full failsafe on all applicable channels is usually a different method or a second step to normal binding depending on the radio set. On the Spektrum DX7 the full channel failsafe binding is done by removing the bind plug after powering up the RX in normal bind mode, each mfgr has their own method. In order for this to work correctly on either loss of signal or TX power off, the sticks should ALL be in the middle of their travel, the autolevel switch should be ON, and if the flight controller has the capability to assign a failsafe switch it should also be ON. With DJI you can easily test this on the bench, do a full channel bind with the TX setup as I described, while connected to the DJI Assistant change to the TX tab and note where the channel sliders are. Now power off the TX and you should see the throttle go to midpoint and the failsafe is active, if you do you have successfully set full channel failsafe and should NEVER encounter a flyaway situation due to improper setup which IMO is the cause of the majority of flyaways that happen.

Ken

 

[coreyperez]

that's a good point Corey, what happens when failsafe kicks in and the helicopter isn't nice and level like it was on our bench when we tested and set it up the first time?
only caveat/caution....remove the props before powering it up, holding it in your hand, and testing stuff. there's so much potential for injury with these things it can't be said enough, remove the props....remove the props.....remove the props

I should re-state, this was intended to be attempted with the Naza Assistant in use, this, using the slider read-out as a indication as to what would happen when power was lost.

Corey

 

[coreyperez]

Most modern radios have two failsafe modes, maintain last stick position and full channel failsafe. The difference is a multirotor needs to have full channel failsafe set for proper RTH to work when the radio is switched off as the default on most RX is hold last stick position. Binding for full failsafe on all applicable channels is usually a different method or a second step to normal binding depending on the radio set. On the Spektrum DX7 the full channel failsafe binding is done by removing the bind plug after powering up the RX in normal bind mode, each mfgr has their own method. In order for this to work correctly on either loss of signal or TX power off, the sticks should ALL be in the middle of their travel, the autolevel switch should be ON, and if the flight controller has the capability to assign a failsafe switch it should also be ON. With DJI you can easily test this on the bench, do a full channel bind with the TX setup as I described, while connected to the DJI Assistant change to the TX tab and note where the channel sliders are. Now power off the TX and you should see the throttle go to midpoint and the failsafe is active, if you do you have successfully set full channel failsafe and should NEVER encounter a flyaway situation due to improper setup which IMO is the cause of the majority of flyaways that happen.

Ken

Ken,

I had used a FrySky with my 9xr and when the failsafe was "programmed" all the sticks were in the desired position, so everthing worked as I expected. BUT... When following ALL the tutorials for Futaba I only seen the people programming the VPP (I think it was called) which was what mixed the failsafe into working. BUT, there was no process described to arrest the maneuver.

As I said, I would see somebody losing orientation, getting to a point of panic, flipping the "fail-safe", which the RX is now telling the Naza "Keep going!", which it would, now the person just keeps flipping the fail safe on and off, and the MR would be getting different instructions from the user. The user thinks they lost control, this in turn prompts them to "Hail Mary" it and shut off the transmitter, in WHAT EVER configuration they left it in and once again, the craft keeps going. Probably into the dirt at that point, thus becoming another "Naza Fly-Away", all due to User Error.

Just thought I'd bring it up, it was not a "hard learned" lesson, but it could have, with one single flip of a switch.

Corey

 

[Bartman]

Most modern radios .................................should NEVER encounter a flyaway situation due to improper setup which IMO is the cause of the majority of flyaways that happen.

Ken

I agree Ken which is why I continue to advocate NOT setting up advanced features (like programmable failsafe in the flight control) until basic flying skills are learned and an understanding of the interface software is very well established.

 

[Bartman]

I should re-state, this was intended to be attempted with the Naza Assistant in use, this, using the slider read-out as a indication as to what would happen when power was lost.

Corey

I got that Corey, thanks for reinforcing it if I wasn't clear about it.

 

[Quinton]

That is a very interesting point I use this combination albeit with a Wookong, I have never attempted failsafe when flying in a certain direction, only when stationary, however even if you were flying in a direction failsafe kicks in very quickly and it would simply come back home would it not, and not continue to fly off in the direction?.
Also I only ever had my throttle failsafe set at 20% the minimum required, never thought about the rest..time for another look I think.

 

[econfly]

It's a good point. The flight controller only knows what the receiver is telling it. The Futaba 14SG (at least with the 7008 receiver) has per-channel programmable failsafe and it is very useful. For example, you can program the receiver to lower gear, center sticks, and put the controller into failsafe mode if the link is lost. However one decides to set it up, recommendation to test (safely, of course) is spot on. I wouldn't want my first dead link test to be in the air!

 

[blade strike]

This is why I always test my failsafe with some kind of stick input when turning off the TX during setup. This is the only true way to know that you have FS setup correctly.

Futaba is easier than spektrum since it's done on the radio side..

People on spektrum need to pull the bind plug out before they bind. Even then people will redo a standard bind and kill their fs bind.

You should always test failsafe after any major change. Like Bart said please remove props before you do any type of bench test.

 

[coreyperez]

Just looking at these messages really has me thinking I got "lucky" with my first 9XR configuration. I just figured "That is the way it is". I didn't realize how many different way to configure (different radios), nor did I even consider a re-bind issue dumping the previous fail-safe.

While I tend to over analyze things, plus I tend to check, re-check an check things again (I was an Analyst in the Air Force and currently a UH60 Pilot) one can never ensure things are perfect. Especially in Aviation.

Needless to say, I was sure I had my settings correct but just had that nagging feeling something wasn't correct. So I waited until I had a friend here to help me with the first new radio flight, and I just couldn't get the motors to arm. Once again, I was 100% sure I had it all correct. It wouldn't have been and would have ended VERY badly for my new landing gear, gimbal, GoPro AND 550.

Crap.. There are so many variables!!! Just when you think you got it all figured out... I'm glad that Naza gave us the Assistant software! (Although I feel it needs a LOT of work...)

Bart,

Is there a way we can incorporate this learning point in the Group build? I'm not really sure if you plan to get to an "Advanced programming" part or not. But I'm sure there are more (than not) possibly making these same mistakes.

Time to go pre-flight. I've got 5 day off and today is BEAUTIFUL (multi-rotor wx)!

Corey

 

[coreyperez]

That is a very interesting point I use this combination albeit with a Wookong, I have never attempted failsafe when flying in a certain direction, only when stationary, however even if you were flying in a direction failsafe kicks in very quickly and it would simply come back home would it not, and not continue to fly off in the direction?.
Also I only ever had my throttle failsafe set at 20% the minimum required, never thought about the rest..time for another look I think.

My thought was, in that instance of panic, that brief moment that Naza is waiting, BUT still receiving info from the RX to "Keep Going" will seem like forever. I could see the panic set in, an the user flipping the failsafe or power button on and off trying to get it to "RTH". BUT, each time the fail safe is cycled I'm assuming that the internal Naza timer resets, thus appearing that it never worked.

Corey

 

[djw43]

i have to admit,that after reading this thread I was curious about the RTH and failsafe in my system. WKM with 14 SG mode "D" SBUS.
I had only throttle set with a failsafe, slightly above middle, the other sticks were set at hold. First I tested in the assistant, If I moved a stick to the side and powered off the radio transmitter, the stick stayed at that last position. So I took out my aircraft to test it. And in both RTH and F/S if the stick was displaced, there was a momentary movement of the craft in that direction, but only very briefly after the power was shut off, then it when into the RTH mode.
I just needed to confirm for myself.
thanks

 

[coreyperez]

That is good to know. I may have to give it a try with the -M V2 and see if it performs the same.

(Some day in the future).

I just got back from doing a max alt/distance (well, new record for me), I flew it 1/2 mile out and over 1000' up! I had a 500FPM rate of descent coming down too. I've probably used up any "good luck" I had in reserves at this point.

http://www.flytrex.com/coreyperez13/

Corey

 

[djw43]

nice, that's a great bunch of badges!