Discussions

Copter jumping Up-Down constantly without input in Alt-Hold (with sonar + baro)

Posted by Shyam Balasubramanian on November 1, 2013 at 3:39pm

Hi Everyone,

There are a couple of people I know who have already faced this in RC5. In 3.0.1 I see that I do not get issues, so posting it as a separate forum.

I am using sonar (XL-EZ0, MB1200) + baro on my APM 2.6. Using 3.0.1 RC5, I have attached the logs for perusal. "This did not happen before" and happens more with slight winds.

Note: There were slight traces of wind, but this Up-Down movement is very consistent. Goes up and down by same distance every time.

The only thing that I can think of is 'Thr`Mid' set at 596 for hovering, which I have tried to vary in earlier tests with similar results.

Can somebody have a look and suggest why this happens, from the log?


IMU: Vibrations controlled fantastically on my setup.

CTUN:

SonAlt Vs BarAlt: Reading correct values as it jumps and tries to correct.

3691074382?profile=original

ThrIn Vs ThrOut, you can see sudden dips. ThrIn (red) remains constant. See the green bars dipping? That is the problem.

3691074244?profile=original

Something with 3.0.1 RC5?

2013-11-01 17-40 1.log

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

Join diydrones

Tags: 2x, arducopter
Email me when people reply –
Follow

Replies

  • Andre Kotze December 9, 2013 at 10:19pm

    Randy,

    Am I correct in saying that in the current firmware the sonar and baro readings are mixed to maintain the height?

  • Andre Kotze December 9, 2013 at 4:37am

    Hi Randy,

    I see that you've announced that 3.1-rc8 is out and that it would probably become the official 3.1 in a week or so.

    I guess the official release won't have anything in it to fix the sonar issues then...?

  • Andre Kotze December 7, 2013 at 12:54am

    Hi Randy,

    Here is a comparison I did between 3.0.0-rc2 and 3.1-rc5.  With 3.0.0-rc2 it was quite windy yesterday but with 3.1-rc5 this morning there was almost no wind...

    https://vimeo.com/81257936

  • Developer
    Randy December 1, 2013 at 8:46pm

    This is the test I did of the sonar.  Not perfect but seems to work!

  • Developer
    Randy December 1, 2013 at 5:01pm

    I tried -rc7 yesterday and I needed to reduce the SONAR_GAIN to 0.8 but then it worked fine.  I'm using a 3dr quad with an Maxbotix MB1260 (aka XL-EZL0) sonar.

    It's reaction time is a little slow but that's perhaps better than it was in AC3.0.1 where it was far too fast.

  • Andre Kotze November 25, 2013 at 3:45am

    Randy,

    Is there perhaps anything in 3.1-rc7 to improve the sonar alt_hold or should we wait a bit more for you to fault find?

    Thanks,

  • Jorge Fernández November 20, 2013 at 6:38am

    Today I went for a flight, not the best day, too windy but fine to test the sonar... When the copter is above the sonar range, it maintains altitude very well, but when I descend under 7m (Sonar start working), It starts to jump, cut the motors etc...

    3692884161?profile=originalSeen this graph the only thing that I could think is a lot of noise in sonar... It has the R-C filter installed, and this didn't happened a few weeks ago... :(  

    I attach a TLOG where I loiter at 20meters the most time but I descend sometimes to see the problem.

    This evening I will record some videos showing you the problem...
    What do you think? Thank you very much.

    2013-11-20 13-08-53.tlog

    https://storage.ning.com/topology/rest/1.0/file/get/3692884127?profile=original
  • Warren Purnell November 20, 2013 at 12:25am
    On earlier versions than 3.01 Sonar for me was absolutely perfect. Once using 3.01 Sonar exhibited issues as described in the post.
    On the bench I am getting accurate readings. In flight up & down with motors cutting out etc.
    I've added all the recommended filtering. Tried two different (same model) Sonar modules.
    For now the Sonar is unplugged & disabled.
  • Shyam Balasubramanian November 19, 2013 at 11:39am
    Jorge,
    AFAIK, 3.01 main release for sonar was perfect.
    Stable release was stable as the name suggests!
  • Jorge Fernández November 19, 2013 at 4:53am

    Hello again, I'm still having problems with my sonar_gain... I can change the value in the full Parameters list to for example 0,2 or 0.1 but the copter is doing the same jumpings and cutting the motors the same way than before... I think that the mission planner is not writting the correct value... I tried to disconnect the sonar, and the copter is smooth and it maintains the altitud very well (Baro precision...)

    I am the only one with that problem? I think it could be an important BUG.

This reply was deleted.

Activity