Welcome to PhantomPilots.com

Sign up for a weekly email of the latest drone news & information

why does firmware not appear to address 1 missing motor?

Discussion in 'Firmware' started by yorlik, Feb 26, 2015.

  1. yorlik

    Joined:
    Feb 9, 2015
    Messages:
    547
    Likes Received:
    91
    So 1 motor or ESC can fail in a flight. Why does the firmware (fw) not address this real possibility?

    I would assume if monitored properly, the fw could see a dead motor, and based on the IMU data, modulate the other 3 to make up for it. Should not be so hard to do. It could even use motor current drop to show a prop broke or spun off....

    Why on a $ 1300.00 sophisticated device like this is that not addressed?
     
  2. LuvMyTJ

    LuvMyTJ ADMINISTRATOR
    Staff Member

    Joined:
    Feb 9, 2014
    Messages:
    6,811
    Likes Received:
    4,376
    Location:
    Live! From New York!
    If you lose one corner of a quad it falls from the sky. Three props is not enough to support flight.
    If you had a hexacopter you could possibly compensate for a failed motor.
     
  3. skyhighdiver

    Joined:
    Dec 19, 2013
    Messages:
    1,007
    Likes Received:
    211
    Location:
    South Dakota
    http://youtu.be/bsHryqnvyYA

    all they need to do is install it in the firmware :evil: :evil: :evil:

    [youtube]https://www.youtube.com/watch?v=bsHryqnvyYA[/youtube]
     
  4. Meta4

    Meta4 Moderator
    Staff Member

    Joined:
    Jan 8, 2014
    Messages:
    7,767
    Likes Received:
    3,478
    Why didn't Henry Ford put disc brakes on his Motel T?
    Multicopters are quite recent inventions and are still evolving.

    And have enough power in reserve to carry the weight of the quad.

    Just because it has been demonstrated to be possible doesn't mean that a simple firmware update to make NAZA do it.
     
  5. TuT

    TuT

    Joined:
    Nov 30, 2014
    Messages:
    130
    Likes Received:
    3
    Location:
    Sydney, Australia
    It could only hope to spiral down to the ground with 3 motors due to the fact that 1 motor cannot counteract the torque from the other 2.

    With a hexacopter or octacopter it can be done, and many systems do have the capacity to recover, but it is almost impossible for a quad.
     
  6. CapnBob

    Joined:
    Aug 17, 2014
    Messages:
    297
    Likes Received:
    26
    Location:
    Massachusetts
    It should be easy enough to cut the opposite motor, throttle up the remaining two, and have some semblance of slower controlled descent.
     
  7. jtludwig

    Joined:
    Jan 12, 2015
    Messages:
    35
    Likes Received:
    13
    And do all of this for a price that we can all afford?
     
  8. LuvMyTJ

    LuvMyTJ ADMINISTRATOR
    Staff Member

    Joined:
    Feb 9, 2014
    Messages:
    6,811
    Likes Received:
    4,376
    Location:
    Live! From New York!
    There is more to it than just adding some code to the firmware.
    The craft in the video is a much lighter and a better balanced quad than the Phantom with gear hanging off of it.
    Maybe in the future we will see this feature implemented.
     
  9. N017RW

    Joined:
    May 2, 2014
    Messages:
    6,257
    Likes Received:
    1,619
    Location:
    Palm Beach Co.- FL
    The quad configuration is under-actuated as it is.

    That is to say the 4 controllable DOF (Degrees Of Freedom), available as the speeds of the four motors, is fewer than the 6 DOF the quad, or any rigid body in space, can achieve.

    In Robotics it would be described as non-holonomic.

    It is not a trivial task to try to control an under-actuated system (quad) as it is let alone with the loss of a control input. The nonlinear dynamics are difficult to model and it is unlikely the Phantom will ever see such an upgrade.
     
  10. yorlik

    Joined:
    Feb 9, 2015
    Messages:
    547
    Likes Received:
    91
    I am glad a lot of you are not scientists or engineers! If so, we probably would not yet have gas stoves, dishwashers, or manned flight.

    Luckily there are enough of us who are obviously too dumb to know something is impossible, like you know.

    Wilbur and Orville Wright faced a lot of similar comments about it is impossible too in their day.

    Three props is not enough to support flight. -- Glad to know that, but in meantime, please watch the video above to see one fly this way...

    Why didn't Henry Ford put disc brakes on his Motel T? -- I was not there so cannot answer this, but I can assume from your comment that you mean in the beginning of a new-ish product, no one should innovate new ways...

    And have enough power in reserve to carry the weight of the quad -- Please see Three prop reply above to see how wrong you are...

    Just because it has been demonstrated to be possible doesn't mean that a simple firmware update to make NAZA do it -- so if it is not easy, don't do it? Actually, if you follow the subroutine algorith Mark Mueller, a lowly student in school, developed, it is indeed an algorithm that can simply to added to any flight program - read his pdf - he has already begun offering it.

    It could only hope to spiral down to the ground with 3 motors due to the fact that 1 motor cannot counteract the torque from the other 2 -- please tell that to these guys - they obviously are not controlling their quad with 3 rotors active:

    http://robohub.org/quadrocopter-failsaf ... ller-loss/

    And do all of this for a price that we can all afford? -- I think you are not an entrepreneur; of course for a price we can afford as it is already done and there is no cost adder for new hardware to the drone to implement it.

    There is more to it than just adding some code to the firmware -- I can see how you would have reacted to poor Wilbur and Orville....

    The craft in the video is a much lighter and a better balanced quad than the Phantom with gear hanging off of it -- Nonsense comment. The algorithm is just as applicable to a 500# quad as to a 10 oz one. Please see post above for a dji size if you still think all this is impossible.

    The quad configuration is under-actuated as it is. That is to say the 4 controllable DOF (Degrees Of Freedom), available as the speeds of the four motors, is fewer than the 6 DOF the quad, or any rigid body in space, can achieve. In Robotics it would be described as non-holonomic. It is not a trivial task to try to control an under-actuated system (quad) as it is let alone with the loss of a control input. The nonlinear dynamics are difficult to model and it is unlikely the Phantom will ever see such an upgrade. -- Just wow. Again, glad there are some of us too dumb to understand your self imposed limits, and dumb enough to believe some things some folks think are impossible are not.
     
  11. Meta4

    Meta4 Moderator
    Staff Member

    Joined:
    Jan 8, 2014
    Messages:
    7,767
    Likes Received:
    3,478
    Or perhaps too dumb to realise that people have answered the question you asked, which was:
    So 1 motor or ESC can fail in a flight. Why does the firmware (fw) not address this real possibility?
    Why on a $ 1300.00 sophisticated device like this is that not addressed?


    Maybe when you buy a Phantom 4 or 5 it will have this feature implemented but it's not going to happen for a Phantom 2 because it is very heavy and incapable of generating enough thrust to stay in the air with only 3 props running.
    The feature you're so excited about is a very recent development that has only been demonstrated and has not been implemented in any production quad.
    For it to work, the quad will have to be designed and built with an appropriate power to weight ratio.
    It's not something that can just be bolted on as an update to an existing quad.

    Of course you are welcome to come up with your own home redundancy proofing system if none of this makes sense to you.
     
  12. yorlik

    Joined:
    Feb 9, 2015
    Messages:
    547
    Likes Received:
    91
    Or perhaps too dumb -

    it is very heavy and incapable of generating enough thrust to stay in the air with only 3 props running.

    meta4, I am sorry, but a drone that can raise itself at over 6m/sec is certainly strong enough to prevent failing out of the sky with 75% power.

    I do not understand why you take this so personally, or why you feel a need to defend firmware from being improved. Unless YOU designed this unit, you do not have the knowledge to make such a limiting statement. Just admit you believe it is 'impossible.' Enough.
     
  13. LuvMyTJ

    LuvMyTJ ADMINISTRATOR
    Staff Member

    Joined:
    Feb 9, 2014
    Messages:
    6,811
    Likes Received:
    4,376
    Location:
    Live! From New York!
    Companies that produce products can implement what they want and when they want. Most companies do not release every option to a particular product until they test it and plan it's release to gain them the most profit. The feature you describe will most likely make it to market sooner or later. If it is a proven technology companies will implement it. It just hasn't happened yet. The same way there are no mass produced flying cars and such. The ideas are out there, sometimes the technology is not.

    So it seems the best thing to do here is to agree to agree I guess since we all agree it will be a feature sometime. So now that that is settled lets move on from the squabbling.

    Also - @yorlik, is there a reason you have multiple profiles on this forum?
     
  14. landmannnn

    Joined:
    Jan 18, 2015
    Messages:
    350
    Likes Received:
    62
    Location:
    Coventry and Limoges
    My thoughts..sorry

    The phantom isn't a sophisticated $1300 quad, it is a $500 quad with some expensive toys bolted on.
    Why try to solve a symptom, not the cause? Better to spend R&D money on improving the ESCs rather than building a recovery method.
    The trend for consumer electronics, including cars, is that the device performs a series of self diagnostics. What we would like to see is the phantom run through a series of checks and report any items outside of the normal parameters.
     
  15. Meta4

    Meta4 Moderator
    Staff Member

    Joined:
    Jan 8, 2014
    Messages:
    7,767
    Likes Received:
    3,478
    It's amusing to have someone tell you you are dumb because he's too dumb to understand why last year's Phantom doesn't have next year's features.

    Even more amusing when it's someone who understands Phantoms so well that they can write this:
     
  16. LuvMyTJ

    LuvMyTJ ADMINISTRATOR
    Staff Member

    Joined:
    Feb 9, 2014
    Messages:
    6,811
    Likes Received:
    4,376
    Location:
    Live! From New York!
    This is where this thread is heading.
     

    Attached Files:

  17. jtludwig

    Joined:
    Jan 12, 2015
    Messages:
    35
    Likes Received:
    13
    If you don't like the product then don't buy it.
     
  18. yorlik

    Joined:
    Feb 9, 2015
    Messages:
    547
    Likes Received:
    91
    My question: "why does firmware not appear to address 1 missing motor?"

    Your reply: "If you don't like the product then don't buy it."

    Very helpful. Thanks.
     
  19. jtludwig

    Joined:
    Jan 12, 2015
    Messages:
    35
    Likes Received:
    13
    About as helpful as your posts eh?
     
  20. max

    max

    Joined:
    Sep 19, 2014
    Messages:
    199
    Likes Received:
    6
    Maybe time to lock this thread........