Home Forums Everything about the Tricopter V4 Naze32 Luxfloat Pointers!

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #12875
    Jay.Ing
    Participant

    Just tried luxfloat over the Rewite i have been using and WOW so much more dialled in! Apart from the yaw that is…

    In forward flight it has a tendency to yaw left (counter clockwise), and on rearward flight the opposite occurs (clockwise yaw).

    Has anyone also experienced this? Ive changes the props and this has no effect, ive messed with the yaw PID settings and ave recall’d the accelerometers!?

    Any help greatly appreciated!

    #12881
    rue
    Participant

    There’s a dedicated thread regarding LuxFloat YAW issues here: http://rcexplorer.se/forums/topic/yaw-drift-on-naze32flip32-anyone-having-issues/

    In addition, there have been changes to Luxfloat’s YAW Integral term in Cleanflight 1.8.0 – they basically multiplied the setting by 10 to increase the weight of the Integral.

    As for my early findings: on Cleanflight/Naze32 I found that the Harakiri PID worked best for me and I hope I can spend some time to narrow down the sweetspot for the PID settings. Which settings are you using?

    #12886
    Jay.Ing
    Participant

    Rue, Just posted them in the Naze32 setting thread.

    Have you had chance to fly the updated luxfloat?

    #12889
    rue
    Participant

    Hi Jay – thank you for posting your settings!! I found YAW “better behaved” with lower P settings (around 3, but again, on Harakiri).

    Have not checked the 1.8 version of Luxfloat yet since I try to establish some kind of acceptable baseline on my initial system to be able to compare when I update. Looking at the commits for 1.8 on GitHub I have not found other changes to the Luxfloat controller compared to 1.7.x except the factor of 10 for the YAW I term, so unless you were inclined to go beyond the current upper I limit of 0,255 there should be no difference in behaviour (browse here: https://github.com/cleanflight/cleanflight/commits/master)

    #12893
    Jay.Ing
    Participant

    I will have a read. Flashed 1.8 this morning. One this i have noticed is that they have implemented the same behaviour that the KK2 board exhibits when arming. The tail servo does not operate until the throttle is above the minimum setting. Thought i’d fried a servo at first!!! 🙁 but after 30 minutes had it figured out. If it clears up today going to try and fly a couple of packs, very misty at the moment, and will report back. I have set up Harakiri on profile 3 so will have a play at that again. I did try it yesterday but had issues with it as it kept wandering backwards towards me at centre stick??? So how i get on today.

    #12900
    rue
    Participant

    >> kept wandering backwards towards me… << Maybe this is why they call it Harakiri after all 🙂 Just checked – I have nick trimmed slightly off center towards front to prevent the backwards drift.

    Best of luck!

    #12901
    Jay.Ing
    Participant

    So just burned a battery, tried the Harakiri, WOW much better, dialled in some settings while i was hovering around, need to get down to the field to check the handling during FPV but so far so good IT IS ROCK SOLID in the hover. Very impressed!!! Onwards and upwards!

    #12931
    rue
    Participant

    One issue that generally puzzles me a bit: no matter which PID algorithm I try I always have to trim rudder/yaw substantially off-center to avoid an immediate turn around yaw after start. Why is that? The yaw control loop is feed by the gyros (and the receiver’s rudder value, which is the default 1500 w/o trim). Shouldn’t the yaw PID take this as a “please don’t turn on yaw” and compensate based on the error detected by the gyro?

    #12935
    Terje
    Moderator

    @rue the FC does not know how much to compensate until the legs have lost their grip on the ground. If you were to take off from a turn table it would be able to compensate on the ground 🙂

    #12939
    rue
    Participant

    Thanks, Terje, for the hint! I’d understand if there’d be a transient compensation right after the yaw gyro gives a non-zero reading in the moment of the take-off. What I still don’t get is that I need to feed the controller an off-center stick position (using the trim) during flight to indicate a target turn rate of ~zero.

Viewing 10 posts - 1 through 10 (of 10 total)
  • The forum ‘Everything about the Tricopter V4’ is closed to new topics and replies.