Home Forums Everything about everything else Testing – Triflight rebased on 3.2.2 Betaflight

Viewing 15 posts - 31 through 45 (of 196 total)
  • Author
  • #40663

    Ok up and running 3.2.2!

    Should be able to log a flight and provide logs tomorrow. That’s assuming the weather remains “ok”. Expecting 11 kph gusts and a chance of light rain.

    Question does 3.2.2 add support for the Runcam ?
    I ask because the option is listed under Betaflights ports menu for camera control.



    Was able to perform the first flight after the tail calibration. With Dynamic Filtering active (all others disabled) the responsiveness was silky smooth as if it was on rails. The tail however was wagging like crazy so I am not sure what to make of this right off. (Yaw P gain was 40)

    Black box did not record a damn thing so I’ll have to do a bit of in-house testing to see why it’s misbehaving. However I did recover the video of the flight:

    As you can see tail-wag is pretty bad and it yaws counter-clockwise during punchouts.


    Looks like my OPenlog is dead, can’t get it to power up. Kinda shitty really, as I was really wanting to pull a log for 3.2.2’s hex.


    It seems like either the P or I gain is too high. A log would help :).



    Previously with earlier versions of Triflight I’ve used the Servo low pass filter (LPF) and Yaw filtering. Normally I set both to 100hz and along with the other notch filters it seams to fly ok. However With Triflight 3.2.2, there is no longer a need for the filters, which for my first flight, I hadn’t yet realized. Obviously this is far superior to anything I’ve flow thus far. It’s extremely responsive and smooth. It’s to the point that for me It feels a bit unnatural and is taking me a while to get used to.

    With just the Dynamic filters enabled and the stock PID’s, here’s what it looks like:

    # diff
    # Triflight on Betaflight 3.2.2 v2 / RCEXPLORERF3 (REF3) 3.2.2 Dec 29 2017 / 14:29:45 (30f68c8) MSP API: 1.36
    name Tricopter Mini
    servo 5 1018 1905 1436 100 -1
    feature DYNAMIC_FILTER
    map TAER1234
    serial 1 128 115200 57600 0 250000
    serial 2 1 115200 57600 0 115200
    aux 0 0 0 1700 2100
    aux 1 1 2 900 1300
    aux 2 2 2 1300 1700
    aux 3 37 1 1800 2100
    rxfail 6 s 1000
    set gyro_sync_denom = 2
    set gyro_notch1_hz = 0
    set gyro_notch2_hz = 0
    set rssi_channel = 16
    set min_throttle = 1100
    set max_throttle = 1900
    set failsafe_delay = 5
    set failsafe_off_delay = 30
    set failsafe_throttle = 1200
    set failsafe_kill_switch = ON
    set failsafe_procedure = AUTO-LAND
    set align_board_pitch = 180
    set bat_capacity = 1800
    set current_meter = ADC
    set ibata_scale = 336
    set tri_tail_motor_thrustfactor = 111
    set tri_tail_servo_speed = 267
    set pid_process_denom = 2
    profile 0
    set dterm_lowpass_type = PT1
    set dterm_notch_hz = 0
    rateprofile 0


    @Kevin_Erik: So what’s the difference in the latter video compared to the first one? Seems like you got rid of the tail wag. The copter still seems to be bouncing a bit on roll axis (or yaw, hard to tell).


    We need to remember that the stock PIDs are for quadcopter, I haven’t yet done any tuning.



    I was flying with Horizion mode active. The roll “bounce” is how the auto-leveling reacts during turns. I’ve yet to turn down the P-Gain for leveling down so its reacting very strongly. Once I can travel to the farm area i fly at, I’ll give acro a spin. Just a bit worried about control so close to somebodies home so i keep Horizon active.


    Did a lot of flying this morning with the PID’s and auto leveling turned down.

    Roll: 30 (50)
    Pitch: 50 (58)
    Yaw:50 (70)
    Horz-Leveling P:25 (50)
    No Notch filtering, only Dynamic, Original values shown in brackets.
    Overall it felt really good, with maybe only the Pitch being too low.

    This is a flight in Horizon mode: https://youtu.be/25Sk6dlhNZY

    Got a bit too aggressive (was fun thought) and tapped a wire taking off maybe 1/8th of on inch from a propeller blade: https://youtu.be/cr7tFil7frE

    Did some Acro mode flying as well but notice the tail wag has returned after the wire strike. Guessing that the Tailtune is now out of alignment. This is one of my first flights using acro mode and as such I’m using allot more throttle. Maybe after I get some more Acro time in I can manage the throttle better and keep the altitude more even: https://youtu.be/9BLpdWW-7lo




    Would you build it for omnibus F4 sd?




    As samp38 said 🙂 Omni F4 sd would be nice so I could test out and maybe help with my babytri, also Betaflight F4 maby 😛 soon ready with an new custom mini tri, have taken all of davids parts and put in an BFF4 instead of the stock f3, should be nice together with the ikon 32 esc.


    Here is the flight log from the Tri-Mini build using the new triflight_bf_3.2.2_v2_MATEKF405.hex_.zip firmware.

    This is a very short test flight in-house as the weather outside is nasty. Please note that this flight is with PID’s that are 1/2 the default and that only Dynamic filtering is active.


    1. LOG00015-1.zip

    Did a full flight just to get the feel of it.

    Blackbox log via Google Drive link: https://drive.google.com/file/d/1rpcDI8fHmVFRat7kc6ikWKrAvAo1cxn5/view?usp=sharing

    Be aware that it was OMG windy and as such I wasnt able to complete a tailtune. (I did try)




    I’ve added some hooks to the upcoming betaflight 3.3 release to make it easier for you to keep triflight up to date with the latest betaflight, see PR https://github.com/betaflight/betaflight/pull/4999

    I’ve added (mostly empty) mixer_triflight.h, mixer_triflight.c and servos_triflight.c files, so that you could use those for the triflight extensions. Betaflight will try to avoid changing anything in those files, so that it should be relatively straightforward for you to keep triflight up to date.

    If there is anything else that would be helpful to you, please say so in a comment to the PR.

    Note also, if you would prefer to merge triflight into betaflight, then that would be possible as well. But if you prefer to keep triflight separate that is also fine.


    Ive been having thoughts along the lines in orientating an F4orF7 FC like OdinnFPV has, but was thinking that a hole in top plate to pass a ribbon through & mount gyro on top plate closest to totally central CoG inside the Mini-Tri cage for protection….
    One of those boards with the separate ‘floating’ gyro’s would be required though, of course 😉

    BTW, out of curiosity, what is the thickness of the tri-copter’s carbon base-plate please?
    Wondering if screws in kit are long enough for 2 of the carbon plates or would longer ones be needed…

Viewing 15 posts - 31 through 45 (of 196 total)
  • The forum ‘Everything about everything else’ is closed to new topics and replies.