Building a prop guard / cage, thing wont fly

Discussions and questions about the Crazyflie Nano Quadcopter
Post Reply
r0b0tn1k
Beginner
Posts: 17
Joined: Fri Aug 30, 2019 9:53 pm

Building a prop guard / cage, thing wont fly

Post by r0b0tn1k » Wed Nov 20, 2019 2:05 pm

Ok, so i've been trying to build my own cage for like a month now.

I have done a sphere, which was too heavy.

Then I tried to reduce the form factor and got to this:

Image

Image

Image

This is video of what happens when it tries to fly with bottom and top shell:
https://i.imgur.com/2c7w7Tc.mp4

This is video of what happens when it tries to fly with bottom half only:
https://i.imgur.com/Ii607bu.mp4

It did take off one time, flew all the way to the ceiling before crashing down.
So few questions:
- Seems like it does have a bit of power to take off with bottom guard only. But the flight is not stable. Any ideas why?
- Are there any settings to pimp the power to the motors so it's able to fly with a prop guard?

I've been trying different printing materials and width, but anything below 2x2mm width/height for the lines is just too flimsy and does not really work.
I have tried PLA, PETG and PC Plus.
Any ideas / suggestions welcome.

Btw, I've seen all things crazyflie related, but I wanted to have something protecting the multiranger dock, hence the top part.



arnaud
Bitcraze
Posts: 2275
Joined: Tue Feb 06, 2007 12:36 pm

Re: Building a prop guard / cage, thing wont fly

Post by arnaud » Fri Nov 22, 2019 8:16 am

When only the bottom part is attached, it seems that the controller is not able to control the position. One thing to try might be to increase the PID controller base-thrust: it should be set to about the hover thrust and since you added a lot of weight it should be higher: http://github.com/bitcraze/crazyflie-fi ... pid.c#L298.

Hovever maybe it is just still a bit too heavy. The controller needs thrust margin foe each of the motors to be able to control the attitude. One way to see this thrust margin is to fly manually and try to look in the client for the amount of thrust on each motors that is required to take off. Motors should have roughly the same thrust when hovering and no motors should touch 100% thrust. It is possible that in your case one motor or one propeller is a little bit damaged or tilted which reduces the thrust margin by pushing one motor closer to the top.

r0b0tn1k
Beginner
Posts: 17
Joined: Fri Aug 30, 2019 9:53 pm

Re: Building a prop guard / cage, thing wont fly

Post by r0b0tn1k » Fri Nov 22, 2019 6:44 pm

Thanks for that. In the mean time I read some papers, and found that the payload weight is 15G. Considering that I have the multiranger and flow decks, plus some extra solder to link to big connectors, I'm wondering how much remains.

In the meantime I just got an SLA printer, so I can reduce some of the parts in weight.

The primary reason for the cage is that crashes damage the multiranger deck. I re-soldered the front sensor since it was reporting as bad, and it started working. One crash later, left sensor reports distance of none, so I re-soldered that. Another crash, and it reports a none as distance again.
Considering the multirager is around 80 eur, I want to give it all the protection I can.
Oh and it bends the connector pins on crashes like it's nothing.

In regards to your answer:
> It is possible that in your case one motor or one propeller is a little bit damaged or tilted which reduces the thrust margin by pushing one motor closer to the top.
What should I do to recognize the damaged prop?

Oh and another thing, one time with just the lower cage it just took off all the way to the ceiling? So perhaps it's not that heavy but there's something else at play? Any types of logs I can provide to tell us more?

Btw, more than happy to share the STL's after I'm done, but I have to get them working in the first place.

arnaud
Bitcraze
Posts: 2275
Joined: Tue Feb 06, 2007 12:36 pm

Re: Building a prop guard / cage, thing wont fly

Post by arnaud » Mon Nov 25, 2019 1:49 pm

Yes, 15g is the specified max payload for the Crazyflie. It is not a hard limit but it is the point from which you start loosing a lot of the agility.

For the propeller/motor margin, I was propoging to look at the motor thrust bars in the client: if you set the flight control in hover mode and look at the M1/M2/M3/M4 thrust bars, all bars should be roughly at the same level when hovering. If one is much higher than the other, you can try to change the propeller on that motor.

Just a random though about protecting the multiranger: if you install the battery bellow the Crazyflie, between the Crazyflie PCB and the flow deck, you will be able to mount the multiranger deck much lower and so it will be protected by the motors and propellers.

Post Reply