3D Robotics

ArduPilot 2.0 Beta Code Released!

I'm delighted to announce the release of the ArduPilot 2.0 software, now in public beta. This software uses the current ArduPilot hardware but handles the FMA Co-Pilot stabilization function itself, without the need for a separate stabilization unit. It is, in short, a complete integrated autopilot. That little ArduPilot board can do a lot! 2.0 Features:
  • All the ArduPilot 1.0 functions: programmable 3D waypoints; return-to-launch mode, in-flight reset ability, fully programmable actions at waypoint and totally expandable board.
  • Now integrates the stabilization and navigation functions, eliminating the need for a FMA Co-Pilot and vastly improving waypoint tracking
  • Controls elevator and ailerons/rudder. Throttle control coming in 2.1
  • "Fly-by-wire" mode stabilizes the aircraft in RC mode, duplicating the function of the FMA Co-Pilot
  • Stores home waypoints and sensor calibration in EEPROM, so they are retained even in the case of a system restart
  • Controls altitude with the elevator, vastly improving performance in wind
  • Currently optimized for the three-channel EasyStar. Versions have also been tested on Funjet and four-channel Superstar (code coming soon)
  • Simple in-field calibration process
  • Can use any thermopile XY sensors (default settings are for the FMA sensor, but Paparazzi, AttoPilot and custom sensors can also be used)
  • Currently supports the EM406 GPS. Support for Locosys and Ublox GPS modules coming soon
  • Uses "chained PID loops" to combine the stabilization and navigation functions seamlessly
What you need (if you already have an ArduPilot and FMA Co-Pilot you don't need anything else): Instructions for modifying the FMA sensor cable to use it with ArduPilot are here. Instructions for in-field calibration and setup of ArduPilot 2.0 are here. (Note: ArduPilot 2.0 put the EM406 GPS module into binary mode, which disables its onboard LED. That means that if you try to use the same GPS module with ArduPilot 1.0, it won't work (unless you program back into NMEA mode with a utility such as this one). However, if you let it sit for a few days without power, the onboard capacitor will run out and it will return to its default NMEA mode and will work with ArduPilot 1.0 again) The next version, 2.1, is now in private alpha. It will include the following:
  • Uses Z sensor for self-calibration in the field
  • Uses small desktop app to load ArduPilot settings and waypoints, without having to work with code
  • Adds throttle to altitude control, improving accuracy and wind performance
E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • I thinking that the change to binary GPS communication, and the higher baud rates will impact Xbee telemetry? For example, No more NEMA sentences sent to google Earth. I could just move the Xbee over to the OSD which has it's own GPS Module

    Thoughts, and discussion?
  • Thanks Jordi, im trying to change this part of code for V1.x but all parameters are changed and have a lot of errors in compile.

    We are waiting your new version and Much Thanks
  • Moderator
    Thanks, I like the tweaking but can only do some of the basic stuff, so will stick to the v1.0 if the v2.0 is very complicated to tweak. Are there new functionalities that the v2.0 code has that might benefit the v1.0 code? I see the waypoints tab uses a fixed distance from the home position/last waypoint rather than an actual waypoint to navigate to.
  • Developer
    Hello everybody, many thanks for the complements,

    To Pumuky:
    Yes we going to update version 1 for you people.

    Now just to clarify the PID chain or P and PI chain to be exactly, is different from EasyStar to Funjet. Funjet uses ailerons to make turns (that was easy to code), now, to control the heading vs bank angle with the rudder (in this case easystar) was trickier than funjet... I had to use a different PI system that also includes an absolute position. Whatever it works, and works like this: The heading error, output the rudder/ailerons SetPoint of the roll PID, Example Ardupilot says: OMG i want to go 50 degrees to the right, so it will output more or less 30 degrees to the roll set point, so the Roll PI will try to bank and maintain 30 degrees, then the airplane will start to turn right without roll over. As soon the airplane get close to the desired heading the PI will start reducing the bank angle to 0, and fly forward... Is similar with the altitude.
  • Developer
    To Shane:

    Yes the messages change to binary protocol and the speed to 57600 bps. You can use you module with ardupilot and also you can use the GS405 with helical antenna (SF).

    Now if the unit is already in binary protocol it will ignore the commands. The problem is with EM405, it switch back to NMEA after some time (when the battery/capacitor discharge)...
  • 3D Robotics
    Graham, In your case, I'd stick with the 1.0 code. Less to tweak. Same for the boat.
  • Moderator
    Excellent work Chris and Jordi, a quick question - I have a plane that uses high wing dihedral to maintain stability and thus does not really require the stabilization provided by the FMA sensor. Could I use Ardupilot v2.0 code and what if anything would need to be changed or would it be better to use continue using the ArduPilot v1.0 code. I also would like to try the navigation in an rc boat and the same question applies?
  • 3D Robotics
    William. Yes, that's correct. This is basically the functionality of a >$1,000 autopilot for around $100. Go open source!
  • I'm I understanding correctly that the Superstar version no longer uses the rudder for navigation, but rather banks the wings? If so, I think thats a first for basic inexpensive autopilots!
  • Thanks Chris and sorry for the mistake above, I meant Chris (you) and not Dean of course (have too long to do with Dean Goedde :-) !

    Erwin
This reply was deleted.