Blogs

Developer

Arducopter Beta update.

Posted by Jason Short on May 9, 2011 at 10:46am

3689405810?profile=original

Found some good bugs already. An issue with auto trimming being activated due to unusual Radio configurations fixed. Mavlink support is now up to par with APM. and A compass bug in the library squashed.

 

I revved the firmware's EERPOM fingerprint. You must redo the setup completely. (except ESC calibration.)

 

Thanks for the feedback!

Jason

E-mail me when people leave their comments –
Follow

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

Join diydrones

Comments

  • Developer
    Jason Short May 10, 2011 at 1:35pm

    You need to pick your kP based on the power of your copter.

    the purple engines Jani sells work well with kP of .75  but many kP values would work. kP is not that big of a deal. It's merely how much power you want to go to the engines for rotating. Faster engines might want a lower kP. 

    I suggest setting this in your AP_Config.h

     

    #define CHANNEL_6_TUNING CH6_STABLIZE_KD

     

    This will let you fly and interactivly change kD.

     

    A low kD value will be very sloppy

    A high kD will be very sluggish

    A kD that is way too high will oscillate.

     

    There is a test called test_tuning in the CLI that will output the correct value once you've confirmed it flies well.

     

    Jason

     

  • emile May 10, 2011 at 1:25pm
    @Jason based on your experience, what is the best method to tune these PIDs?
    Shall we concentrate on the Stabilize Damp as suggested in the wiki?
    It would be nice if you can give us a little tutorial so we can "spread the word" :)
    Cheers, and again impressive work!
    Emile
  • Vincent Mees May 10, 2011 at 1:22pm
    I'm having the same issue. Even with GPS lock no data.
  • Peter Seddon May 10, 2011 at 1:14pm
    Have you got GPS lock Nigel. On the ArdupilotMega you don't get GPS co-ordinates translated until you've got lock.

    Peter
  • Developer
    Jason Short May 10, 2011 at 1:08pm

    Our standard PID loop doesn't take into account the Sensor data for D term. 

    This was Jose's original approach. Technically, it's a standard PID loop. Rate can be derived from a saved angle, or directly from the sensors. The reality is that we could use the saved rate and make it 100% standard. But it's not worth changing because we also use some nice constrains in there too.

  • emile May 10, 2011 at 12:49pm

    @Magellan, thanks for the answer. I was used to the always there "408" value. So I bet this is a 1G pulling down just a matter of showing it in a different way (maybe makes more sense now :))

    Ok, so here is another Q.

    In the PIDs D is a real D or is it the P-Rate as for the former AC? 

    Cheers,

    Emile

  • Magellan May 10, 2011 at 12:34pm
    It shows -1000 since the device is sitting on a level surface and gravity is pulling down on it. Gravity is always there.
  • emile May 10, 2011 at 12:30pm
    Just a studid question....
    Why is the Z Accel now showing -1000?
    Is this the normal behaviour?
    If I move the copter everything seems fine, the data moves accordingly.
  • Nigel May 10, 2011 at 10:56am

    I know Jason is aware of Mavlink issues but for the record, with today's current code I get different data from the USB port than from the UART.  The USB data seems to be complete with full telemetry and GPS whilst the data from the UART only seems to contain servo data...

    3692209299?profile=original(This is just for info - I know this is still being worked on)

  • John C. May 9, 2011 at 9:11pm

    Pitch and Roll - gyro stabilized with control mixing

     

    Quick and dirty stabilization

This reply was deleted.