QuadCOP - More construction Pictures

Good post.

It reminded me of the time I drove a helicopter FLIR camera based upon its flight vectors.

I used five degree steps to keep the movement smooth and then one degree steps once I had the FLIR looking down the flight vector.

Those were good days, especially during flight test with the UH53.

DAB

For my purpose, I wanted 360 degree rotation but I would need another Magnometer but I only have 1 that works on hand.  My original specs say the quad will do a 360 degree sensor sweep, I changed that to be a 180 degree. The quad wont rotate, just the head.  The head will only moved when stopped but I may add some "bling" movements so it looks around while moving to the next waypoint and while being flown manually.

the head will have a sonic sonsor and motion detector on it as well.

Basically the same algorithm I used on my B-25 bomber turret but for that I used Arduino, but the ChipKit can handle it!  It will have a smooth motion with variable speeds.  I am considering putting my Luxeon LED setup on it too so it acts like a search light.

You don't have permission to edit metadata of this video.

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

Comment on QuadCOP - More construction Pictures My other related hobby:  RC Jets and Planes

Poster

Upload Preview

Ok

Cancel

Joey,

this is fine but I have experienced in past with drones a potential issue - that probably can be solved before with a proper workaround. The sonic sensor detects the height from ground supposing a variable but continuous ground surface. I mean for example the drone while flying at a fixed height of 5 meters from ground while going over a hill.

The problem unfortunately occur if you try adopting the same method flying the drone on a top terrace at the third floor of a house. As the drone goes outside the terrace, the sonic sensor detect a ground height sort of out-of-scale Depending on the software can try to compensate with not perfectly predictable fly-down or lose the control because the sensor is really in a range of unexpected values or something else. Just a warning to take in account.

Hope this can be useful. Enrico

For altitude, I will be using the GPS readings.  This is to keep me 10 foot or so up in the air.  The ground sonic sensor is for landing or obstacle avoidance.  Its really on good for 2 feet or so.

The assumption is the take off and landing areas are clear.

BTW, if I am over a house there is a serious issue!

For altitude, I will be using the GPS readings

Are you sure it will be accurate enough ??

Barometric pressure (corrected at takeoff) could be much more accurate, but you do have airflow disturbance that it would need to be protected from.

The sensors sent with the Robot kit are good for distance to 5.5m ( 18 feet).

http://www.element14.com/community/groups/stemacademy/blog/2015/08/13/virtual-eye-system--post-3--testing-and-calibration

Mark

I think it will be accurate enough for 10+ feet, but if not then the position itself is not accurate I guess?  I did attempt to use the pressure sensor on the xtrinisic board but no compatible with RPi 2 at this time.  I did get it working but not very reliably.  On the RPI B+ it works great! 

The promised GPS vertical accuracy is supposedly quite poor however most devices that claim reasonable accuracy on their "z" axis  actually combine the barometric sensor data to give a more accurate final "GPS" value .

The promised GPS vertical accuracy is supposedly quite poor however most devices that claim reasonable accuracy on their "z" axis  actually combine the barometric sensor data to give a more accurate final "GPS" value .

As a matter of fact, I have serious suspects about the reliability of the GPS vertical accuracy and the best solution maybe an altimetric sensor. The problem of the height stabilisation was one of the bigger issues in all the generations of the AR-Drone devices as I have explained in a post above.

Some other high profile (and very high price) drones, including the DJI Phantom if I am not wrong, consider also the parameters returned by a very reactive accelerometers that is used to compensate the z-axis offset of a given altitude based on the small movements detected and their relative speed.

Enrico

Yeah the inputs from all the sensors are used to compensate for the GPS inaccuracy. Also they have an elaborate calibration procedure for those sensors at the beginning of the flight . finally of course  the craft doesn't need to to really know where it is in absolute terms but relative to the start so if it's wrong it doesn't matter as long as it's consistently wrong in it's measurements.

What I have experienced until now with drones, is that the aspect that seems more complex, to make them flying, as a matter it is the less problematic and simple. True it is easier than making fly a traditional aero model. Then when the aero model flies 95% of the effort is done while when a drone flies (almost any thing included a rick can fly as drones) it is only done 10%-20% of the job.

Enrico

Well since I have a bit more time before I fly I may give the xtrinsic board one more shot.  I was playing with the pressure sensor.  Had a hard time not getting it to lock up, but I was using the code they provide in Python.

I agree GPS is not going to be perfect but its what I got.  The sonic sensors will keep it from hitting the ground.