Ardupilot hardware: appropriate for a UAV not using the Ardupilot software?
Ok, so the more I've been thinking about it, the more I've realized that the current Ardupilot software would take more effort to change for my particular needs than just patching together my own code. Maybe I'll integrate the two at some point, but that's down the road.
I do plan on sticking with the Arduino IDE for simplicity. That leaves me with a toss-up between using an Ardupilot board and something like the Arduino Pro Mini. The failsafe system wont do me much good as the UAV wont have a traditional RC setup onboard. The only real advantage I can see is that the Ardupilot board has power and ground conveniently placed next to output pins for attaching to servos and escs. But perhaps there's some aspect to it that I'm missing. Any thoughts?
Thanks!
I do plan on sticking with the Arduino IDE for simplicity. That leaves me with a toss-up between using an Ardupilot board and something like the Arduino Pro Mini. The failsafe system wont do me much good as the UAV wont have a traditional RC setup onboard. The only real advantage I can see is that the Ardupilot board has power and ground conveniently placed next to output pins for attaching to servos and escs. But perhaps there's some aspect to it that I'm missing. Any thoughts?
Thanks!
Replies to This Discussion
-
Permalink Reply by Reto on -
In my opinion, the only important thing you are missing at this point starting a discussion is being a little more explicit about your project goals. How could we else contribute? Unless I am missing some aspect, though.
-
Permalink Reply by Xander on -
Fair enough. I didnt want to unnecessarily complicate the question, but perhaps that was a bit too brief.
The project is a quadrotor. Version 1 will basically consist of a 6dof IMU controlling 4 brushless motors. Xbee modules will be used for remote control and telemetry. That's what I'm trying to find the best board for right now.
GPS, compass, waypoints, etc will come later and version 2 may very well be a more powerful uC. Right now it's mostly the IMU and control algorithms that I want to fine tune.
-
Permalink Reply by Joseph Glanville on -
Quads are fun. There is no such thing as a complicated question imo. Only a convoluted one, the complexity goes away when the question is worded correctly I think.
However moving on:
The ardupilot board can do what you want it to do.. the atmel 328 should be enough power to deal with the 6DOF IMU calcs. You might want to take a look at 9DOF and AHRS for V2 though.
Are you building your own sensor head?
You have probably already seen them all but there are many purpose built boards for controlling quads, what is your rationale behind not using one of those?
You are indeed correct though, the bonus of using the Ardupilot over a stock Ardunio board is the pinout.
Interested to hear your reasoning.
Goodluck and don't hesitate to ask more specific questions.
-
Permalink Reply by Xander on
-
Cost, flexibility, and performance. At least one of these is lacking from every available quad board I've come across. Perhaps I'm being cocky, but I think I can do better, and I enjoy the challenge.
The price tags in particular irk me. Some of those boards cost $500 or more for basic stability control. I can do that for about $60. It's not so much about a lack of funds, just the principal of the thing.
-
Permalink Reply by Mike Bakula on -
Don't underestimate the advantages of the failsafe system. Having a way to cut power to the motor controllers will save you some propellors. (We use the pro mini at school for our IARC project, and have had a few runaways due to programming errors.) It would be wise to have some sort of datalink, and learn how to use both the hardware watchdog timer and software timeouts on critical functions.
-
Permalink Reply by Xander on
-
Thanks Mike.
I do intend to have a (two way) datalink through xbee modules, but there wont be conventional RC gear aboard. So the Ardupilot failsafe wouldnt do me much good. No PWM signals coming from a reciever that can be patched directly through to the motors. For that matter, quads are so inherently unstable they pretty much can't be flown without electronic assistance.
So I'll just have to code carefully and have my failsafe(s) in the arduino software. The watchdog timer and timeouts are definitely part of the plan. Hopefully I can get proficient enough at using them.
© 2010 Created by Chris Anderson
