After flight my quad tried to fly away even though motors were shut down

[Edgar Scott]

After a great flight all motors shut down.

I turned off myTx before my Quad batter disconnect, I know it went into FailSafe

I know why it may have done that but if the chopper is stopped and on the ground it should not take off again.  Lucky I grabbed it and brought it back but really I think 

A solution could be to always check that the throttle is > 10% before it ever tries to engage RTL to prevent it from taking off and smacking someone in the head like mine almost did if they forget to unplug their battery after a flight.  I don't want to see someone get hurt from a rouge flying quad.

Ed,

[Chris Anderson]

How can it check to see if the throttle is >10% if there's no RC input at all (because the Tx is turned off)? Instead, a better approach might be to not engage failsafe if the quad is on the ground (that's not always easy to detect 100%, but worth considering).  

[Edgar Scott]

:) That works

yea, I thought later that this wont work because my Failsafe is a real low trottle < 975

But whatever reason why is this happening if its not failsafe?

[Bill Bonney]

In ArduPlane the RTL fail safe does not engage unless the plane reports a movement of 10m/s. Though even with this you can get a false positive if the GPS is indoors. There is a big long thread about fail safe started by Monroe... Check that for more info on this.

[Edgar Scott]

I tested 3 times.  The craft was sitting on the ground about 2 minutes outside not moving.  It "appears" to me that the fail safe is trying to engage. 

[Chris Anderson]

Edger, Bill is referring to ArduPlane, not ArduCopter. Needless to say, the plane method of not engaging if you're not moving forward won't work for copters, since they can hover while planes can't. 

[Boston UAV]

Always DISARM when you land, then this won't happen :)

But yes, I've done the same thing.. only once though!

[Edgar Scott]

Yes I know that, but it should never take off if the craft is landed and not moving and stopped motors after a flight even if the Tx is turn off before the battery disconnect.  I want to say I tested this 3 times and if I turn off the Tx "After" a flight this always happens, turn off the Tx before the flight and it does not happen, it just sits there.

It was done all 3 times outside.

[Thomas the Nerd]

This is why there is an ARM/DISARM function. The copter doesn't understand your intentions (yet).

I've had a related problem: I tried using the "ARM" button in the Mission Planner, and the controller was off. When I turned the controller on, the motors were already armed and the vehicle went full-throttle. Luckily, it was tied to my test stand so it didn't crash into the ceiling.

[Edgar Scott]

Yes, I assume that it is ARMed but like the Arduplane it should not engage the RTL if not moving or I think Chris Anderson has a better approach for the multirotors that it should not RTL if the craft is already on the ground.

Its extremely dangerous for it to take off if the Tx is shut off, under no reason should the craft start its motors and take off when its on the ground and the motors are stopped.

Even the auto mode you have to throttle up a little bit, but in this case it just takes off and comes at whatever is in its way,

Its easily reproducible

[Edgar Scott]

I tested my f550 and the Naza and it does not take off after a flight when I change it to failsafe or shut off Tx

[Mac]

While this is an interesting bug that should be looked into, it sounds like you broke two cardinal safety rules:

1) For any radio controlled vehicle, NEVER EVER EVER shut off the TX before shutting down the vehicle.  You don't know what else is transmitting in your area.  Shutting off the TX is asking for trouble and it's super-bad to get in the habit of this.

2) For any vehicle with an arm/disarm function, ALWAYS disarm when you don't want the rotors to be spinning.  The arm/disarm function is there for a reason, use it!

As for how to fix the bug, I can't think of any quick and easy change change that doesn't introduce its own fun and creative in-flight failure modes.  Note that at the moment, there's no way for APM to "know" that it's on the ground.

[Edgar Scott]

Adam, 120% could not agree more, it was a mistake, even the noobs know this and I do too.

But it can happen and someone else will do this sooner or later,

Im not sure how the N product knows it is on the ground but "I think" I read that it was 10% throttle needed for failsafe, I can look into it again.