sonar sensors affected by propeller background noise

http://prism2.mem.drexel.edu/~billgreen/Videos/3D_flying/fullAutonomousHover/fullAutonomousHover.wmv

Looks like a fun project.


I'd guess the airflow from the propeller is bending the sonar beam as
its emitted; this doesn't affect the tail sensor since its beam is
parallel to the flow.

Could you mount the new sensor out near a wingtip?  Any attempt to
shield the fuselage-mounted sensor will result in drag...

- Daniel

Very cool.  How does it regulate spinning about the vertical when in the
hover?  Does it use accelerometers?


Well, just a thought.  Maybe it's not just the background noise in general,
but a direct reflection off the wall from the prop to the sensor?  You
could try blocking that direct path by either putting a colling around the
prop to make it a ducted fan or just adding a surface to the wing that
extended out in front of the wing to the side of the prop.  To test the
idea, I guess you could hold a sheet of something to the side of the prob
(above the wing) as it hovered to block the path.

I would assume the noise was creating problems because one of the harmonics
was in the frequency range of the sonar.  Maybe a different prop will allow
it to run at a different speed which might move the harmonics out of the
range of the sonar.  A larger prop that spins slower I would think would
help to reduce the harmonics at the high frequency used by the sonar (40
KH?)

Or maybe the noise is coming more from the motor than the prop???  Or maybe
the vibration of the body itself is producing the noise and the fact that
you seem to have a body which is just a vertical surface causes it to
radiate out maximal noise to the sides?  A different engine mount might
help to isolate the motor vibrations from the body and stop the body form
acting like a speaker cone (if that is what is happening?).  Or some
shielding to keep the noise radiating from the motor might help?

I also wonder if there might be vibrations traveling through the body of
the plane that are being picked up by the sensor.  Better vibration
isolation mounting of the sensor might solve that if that is the case.

--
Curt Welch                                            http://CurtWelch.Com/
curt@kcwc.com                                        http://NewsReader.Com/

I mounted two miniature DC motors on each wingtip, facing in opposite
directions.  Thus, when both motors are activated, it creates a
rotational force which counters the torque roll effect.  I measure the
roll angle of the aircraft and implement PD control to generate the
control signal to the motors.

Ah, I didn't spot them in the video.  I assumed you must have been using
either special close in ailerons to create the torque (since normal
placement of ailerons on the tips would do no good in the hover) or special
independent controlled elevators.


What type of sensor do you use to measure the roll angle?

BTW, I like the answer from the other poster better than the answers I
thought up for your question - that the air rushing past the sensor is the
cause of the problem.

--
Curt Welch                                            http://CurtWelch.Com/
curt@kcwc.com                                        http://NewsReader.Com/

I'm using a 6 DOF IMU from Microstrain.

-weg