Long time reader, first time poster and first time project for me.
Let's get cracking, I've got a fair bit of R/C plane experience and I'm an aerospace engineer for defence electronics company, but this is my first project so I still have a few questions for you guys!
I'm aiming to build a fixed wing fully autonomous aircraft with a decent payload capability, ideally up to 750g - 1kg.
I'm based in the UK, and I've started looking at http://www.coolcomponents.co.uk/ as a source for my avionics.
Questions are....there seem to be a number of different IMU solutions for the Ardupilot board, so, straightforward question: which is the most popular? and which is the cheapest that gets the job done?
Magnetometers? does anyone use them?
Piezoelectric pressure transducers for airspeed and static pressure measurements?
looking to use xbee for telemetry, any good?
I'm aiming to use a GoPro HD as the main camera with a 2.4 or 5.8ghz video link, FPV footage using GoPro is simply stunning.
This is a long term project that I am willing to invest a significant amount of time in, but I understand the importance of a solid system specification from the start!
Any advice / feedback would be appreciated, and once the design gets underway I'll start a thread.
Thanks guys,
Mark.
Replies
At the moment Ardupilot Mega + Oilpan IMU is the mainstream way to go. You do need to know which end of a soldering iron to pick up to assemble them, but not much more, provided you read the manual.
APM does altitude to throttle feedback so airspeed sensing isn't absolutely necessary. GPS altitude accuracy isn't wonderful, so a barometric sensor can help. Magnetometers come into their own when hovering. Otherwise getting direction from GPS will be fine.
I'm flying more or less OK without any sensors other than what's on the oilpan. Any issues are down to the FW and my setup.
With a few years RC experience I was able to install the APM system and have it operate first time out. The HIL simulation is very useful for practicing flying with the APM ahead of time. You can see what the modes do for you how you expect the model to fly (depending on how close your airframe is to the simulator-modeled airframe).