An interesting hardware bug

Firmware/software/electronics/mechanics
Post Reply
liki
Beginner
Posts: 16
Joined: Thu Nov 13, 2014 5:54 am

An interesting hardware bug

Post by liki »

I was debugging my new attitude controller with CF2 on my table and removed the propellers for safety reason. But the communication between the CF2.0 and the client is very bad, and the link quality is less than 50% when the motors were running without propellers on it. At beginning, I though this may be due to the fact that my new attitude controller consumes too much CPU resources and could not spare any more for the communication.

However, when I put back the propellers and ran the controller with propeller spinning, the communication quality goes back to normal again. I did not expect that whether the propellers are mounted or not when the motors are spinning can impact the communication quality. No sure if this is related to motor back emf which might impact the circuit.
tobias
Bitcraze
Posts: 2339
Joined: Mon Jan 28, 2013 7:17 pm
Location: Sweden

Re: An interesting hardware bug

Post by tobias »

This is also something we have seen and investigated. The noise seem to be generated by the brushes in the motor, it varies a lot from motor to motor, is related to the RPM and strongest for M1 and M4. As it doesn't have any noticeable impact on lower RPM, like when the propellers are mounted, we have not done anything about it. We found out that it is possible to shield it away by wrapping the motor cables in piece of aluminium foil which should connect to the motor can. Though it is a very strange problem as it only effect TX and not RX. Also the effect is way less in 2mbit mode, probably because the packet is shorter and less likely to hit a "brush-noise-spike".
liki
Beginner
Posts: 16
Joined: Thu Nov 13, 2014 5:54 am

Re: An interesting hardware bug

Post by liki »

Thanks for the info. Good to know that you guys notice this problem as well. :D
Post Reply