Some time ago I wrote something about getting a cheap GPS receiver, a cheap AHRS, and how to read values from a RC receiver. All those tools can be put together to control an autonomous vehicle of some sorts. If the vehicle is a small ground robot nothing more is needed, except the vehicle itself, of course.I do not own anything that is capable of driving around outdoors, but I do own a small airplane.

Its the Multiplex EasyStar II, a RTF model meant for beginners and it basically flies on its own. Well, not really, but its very easy to fly and you do not necessarily need a motor to land it. The plan is to equipped this plane with a set of electronics capable of stabilizing the flight, tracking a trajectory or following a route defined by way points.

Only one sensor is missing, a pitot tube. This device calculates the velocity of the plane relative to the wind surrounding the plane. This value is different from the ground speed given by the GPS device as the plane might be flying in a wind stream, so the velocities add up. The device and the math behind it is rather simple. All you need is a relative pressure sensor and the actual tube.

I decided to use a relative pressure sensor that directly yields the pressure value needed for the equation. The MPXV5100DP built by Freescale Semiconductors has an operating range of 0 kPa - 100kPa and is rather small and only needs 10 mA of current at 5V.

[caption id=”attachment_118” align=”alignleft” width=”300”]Pitot tube Pitot tube prototype printed by a 3D printer[/caption]

Building the actual tube was a bit more complicated. I designed the first model in Autodesk Inventor and ordered a 3D print at Shapeways. The quality is really nice and I received it a week later after ordering. Those guys are crazy fast!

But when I received the model I realized that getting rid of the plastic powder within the 3mm internal tubes is quite troublesome and the tube for the dynamic pressure input is changing from 5mm to 3mm along the insides. I have no idea how this affects the performance as the equations assume a constant radius along the way. In addition, the whole thing is quite heavy and bigger than I thought. I guess I will have to redesign it to make it smaller. Nonetheless I will use this as a testing platform. If it works I can still use it with my bike or car ;-P

So far I haven’t come around coding anything, but I will update this post ASAP. Maybe I will write a new post for the whole development/testing process.

Stay tuned and happy hacking!