Boat zig zaging in AUTO

Posted by Richard Sterling on January 3, 2016 at 12:45pm

I have just set up this 740mm power cat, AMP 2.5, NEO 7 GPS with compass.

Everything works great in manual but when I switch to auto the boat zig zags across the lay line, it makes the course but the course it steers looks more like a sail boat tacking up wind. The course is a simple HM plus 3 way points, a simple rectangle.

Could I be getting interference from the motor? the APM is mounted 60mm in front of the motor and the GPS is mounted 120mm in front of the motor?

See logs attached

Thank you in advance,

RS

2015-12-31 13-05-03.tlog

2015-12-31 13-05-03.rlog

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

Join diydrones

Email me when people reply –
Follow

Replies

  • John Easton July 8, 2016 at 9:11pm

    Had the same thing but got it sorted out and she ran perfectly, then did a Mission Planner update and it is now worse than ever and I can't remember what I changed.

    I will try some of these settings and let you know.

  • Richard Sterling January 10, 2016 at 6:29am

    David;

    That's brilliant, I'll use that.

    At the moment its breeze here so I won't be able to get on the water for the next few days.

    I'll post as soon as I do.

    Thanks again,

    RS

    • David Boulanger > Richard Sterling January 10, 2016 at 11:16am

      It's not brilliant.  Just using all the tools that are available from the people who have given us this very versatile product.  It's amazing to me really.

      Regards,

      David R. Boulanger

  • Richard Sterling January 8, 2016 at 2:12am

    Guy's thank you so much for the feed back. It's been a bit breezy over the last week and I only got out on the water yesterday for some testing:

    Before I went out on the water I downloaded David's params and compared them to mine. I didn't want to touch too much just wanted to see where I was compared to David.

    I followed in info provided in this post and adjusted the NAVL1_PERIOD 24 (have a thing about odd numbers)

    This improved the tracking, was no longer tacking up the course but was a bit drunk in the way it corrected and especially when it over shot the way point.

    I then saw the param above NAVL1_PERIOD. NAVL1_DAMPING and started to adjust that.

    At the end of the day I had the NAVL1_DAMPING at max 1.0 and the NAVL1_PERIOD at 10 (I think this will end up being 8).

    I want to do some more testing and play with the STEER2SRV_I .

    The boat can now steer the course, but needs to be improved. 

    Yes the GPS refresh rate is 5Hz.

    Thanks again guys,

    RS

    • David Boulanger > Richard Sterling January 8, 2016 at 9:13am

      Richard;

      Glad you are making progress!  Sounds like your boat is a bit more nimble than mine.  Actually, a turtle is more nimble.  Maybe with a faster boat the steering P and I are a bit more important.  No experience with that.  One tool I did use when tuning the boat was the Tlogs.  I looked at my channel one out that steers the rudders.  Ran the same course over and over with different parameters and looked at the PWM output to my rudders.  When I saw my PWM only fluctuate +- 15 on my rudders heading to the next waypoint I considered the tuning to be good.  I do have a bit of a over steer problem when making a sharp turn,but, when the boat is going straight the rudders are not very "busy".

      Regards,

      David R. Boulanger

  • Harald Molle January 6, 2016 at 11:31am

    I agree with Chris,

    first try to reduce the P-Value. 

    If this does not help, the construction of the boat should be reviewed.

    I already had this problems with a monohull construction and a bad "center of gravity" due to 
    wrong placement of the lead in the keel.

    • David Boulanger > Harald Molle January 6, 2016 at 11:38am

      Harald;

      I disagree.  Tune it first with the P being the turning radius.  Boats don't have tires firmly planted on the ground.  The NAVL1_PERIOD has a huge effect on boats and may have values much higher than Rovers.  Probably has nothing to do with the construction of a Cat.

      Regards,

      David R. Boulanger

    • Harald Molle > David Boulanger January 6, 2016 at 11:53am

      David;

      you may be right. I used my own autopilot software based on the very first implementation on ArduPilot. 

      I don not know about the NAVL1_PERIOD. I assume it has to to with the update rates.

      My problems with zigzaging reduced drastically by:

      1. Moving from monohull to catamarane

      2. reducing P

      3. increasing the update rate from gps from 1Hz to 2 Hz.

         (my complete control loop is based on GPS and not on inertial data)

    • David Boulanger > Harald Molle January 6, 2016 at 12:11pm

      Harald,

      I think the GPS update is 5Hz but I'm not sure.  We also have a STEER2SRV_TCONST Param that did not help me much.  I have not yet played around with my STEER2SRV_I which is very low compared to my P.  I have a bit of a over steer condition when I make a turn greater than 90 degrees but the boat then locks into the next waypoint nicely.  The problem is when the rudders turn opposite of the direction the boat is going it takes a bit of time before the boat starts turning in the proper direction.  If you were to turn a rover to the left it would go left immediately more or less.

      Regards,

      David R. Boulanger

    • Admin
      Thomas J Coyle III > David Boulanger January 6, 2016 at 1:13pm

      Hi David,

      Good observations concerning the difference between a model boat turning and a rover turning in relation to the steering tuning parameters.

      Regards,

      Tom C AVD

This reply was deleted.