New code broken?

Posted by Gustav Kuhn on June 27, 2011 at 10:39am

Hey, I'm a noob, but here goes...

Up to the middle of last week, I could quite quickly and accurately use Mission Planner to install the code it auto downloads onto a stock quad.

Since then, after upgrading, almost nothing works for me.

If there is no magnetometer connected, the board doesn't come alive at all, and hangs Mission Planner.

Even after calibrating the ESC's on the RX itself, the closest I get to functionality is three of the motors arm, the last one just sits and beeps at me.

Mostly two motors beep at me.

The CLI ESC calibrate doesn't work at all.

These problems crop up on three different systems, two of my own, one of a friend.

YET, when I load ArduPiratesNG, all three systems work.

 

So what am I doing wrong ?

Please?

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

Join diydrones

Email me when people reply –
Follow

Replies

  • Hans Kulker July 1, 2011 at 10:36am

    I have now updated to APM is 1.0.22 and firmware AC 2.0.31 beta.

    I have now the same problem

    The USB communicatie with Mavlink is not fixed.

  • Hans Kulker June 29, 2011 at 11:34pm
    In response to the problem I have done some further research.
    I noted the following:

    Mavlink the communication between the USB and APMboard APMplanner is not built
    I watched this with the program Portmon
    Seems that the firmware gives a wrong answer
    So is there anything on the Mavlink resent communication has changed.

    This may help to solve the problem.
  • Gustav Kuhn June 29, 2011 at 11:02am

    Well, as of yesterday, the two systems I had problems with over the weekend, now seems to be working fine.

    Loads code, connects to APM, and the one installed in the quad flies.

    The third one from Wessie I have now confirmed actually DOES have a hardware problem on the sensor board, not the CPU board. Sorry for the waste of bandwidth on that one.

    I have confirmed that a barebones system without gps or compass will connect and program, thanks Michael, Jason :-)

    (Should have kept my trap shut, and mind open, but Why does the faulty one program and show info with PirateNG and Configurator ????? )

    Built a 4th unit tonight, works fine, I'll try to drop it off tomorrow Wessie, confirmed it is working with fixed wing code.

  • Gustav Kuhn June 29, 2011 at 1:54am

    Mm......, very close to be totally confused now.

    Loaded code with Arduino, confirmed that it connects to APM 1016.

    Upgraded APM to 1020, confirmed that the brain still connects.

    Loaded code with APM firmware updater, NOW the brain connects properly.

    Same notebook (Toshiba), cables, brain.

    Time for tea, methinks.

  • Gustav Kuhn June 29, 2011 at 12:37am

    "Cat's away, mice will play", configured modes using APM 1019 terminal, still working.

    Oh yes, I configured the speed controllers manually, directly from the RX.

    Now just waiting for the latest storm to pass, to test the quad.

    Testing of the latest code will be done on a separate notebook, and IMU..........

  • Gustav Kuhn June 28, 2011 at 11:46am

    Ok loaded 2029 using Arduino, setup in Arduino, it talks to APM 1019, and the quad flies :-)

    (Using my Notebook)

    ..........

    ..........

    ..........

    Right, I officially give up now, quad NOW connects to APM 1020 without problems, GPS and Horizon working.

    (Using my desktop, that didn't work earlier)

     

    The only thing I still haven't setup yet, is the mode switch, leaving that to tomorrow, need to show a working quad then....

    Still love all you all !

     

  • Gustav Kuhn June 28, 2011 at 11:26am

    Hey, I learn all these new catchwords at the saltmine lately (except the place is coming to a grinding halt), so let us "touch base" here.

    My equipment is as follows:

    Standard quad as supplied by jDrones, gps, mag, and sonar.

    Radio is a Spektrum DX8, Mode 1, working faultlessly on other aircraft.

    Three different computers for programming, all running XP.

    ArduPiratesNG works.

     

    And the latest Planner/Code I tried tonight can't even connect to the quad.

    Will load older code and use older planner to load, and will report back.

  • Gustav Kuhn June 27, 2011 at 12:21pm
    Ouch, thanks, not wasted effort though, one of these is supposed to be a plane, eventually !
  • Gustav Kuhn June 27, 2011 at 12:08pm

    Oh my, just compiled and loaded APM2.12, and it seems to be compiling for something other than a quad, "as is"

    A few pointers, please ?

    It loaded ok :-)

  • Gustav Kuhn June 27, 2011 at 11:35am

    Hi Jason, I've just confirmed that the Hotel  1.0 board, without the magnetometer hangs two different CPU boards in APM Planner.

    Might be a fault on it, except it runs PiratesNG.

    Thanks for the reply Michael, but I did also calibrate the controllers individually, same problem, and they run properly, once again, in NG.

    I'll try Jason's suggestion, program and configure out of Arduino.

    Had a 14 Hr day, just a bit tired to swop out anymore ArduCopters now :-)

This reply was deleted.