Pixhawk & Helicopter tuning

Hello, I used the tuning procedure in the traditional helicopter wiki pages.

The results are very interesting, in stabilize the helicopter fly well

HELICOPTER = TREX450

PixHawk with HELICOPTER 3.2.1 with GPS & external magnetometer

So, I have some ask to reach a better tune.

- Switching from STABILIZE to ALTHOLD the thrttle stick position to have an altitude fixed is 25% of the stick travel, instead in stabilize is 50%.

So the problem occour when I swtich from ALTHOLD to STABILIZE, I read about THROTTLE_HOVER parameter but I don't understood the behavior.

- After a accelerometer calibration the articial horizon in the mission planner is as I expect for a leveled helicopter on the bench.

Then I fly the helicopter, I trim it (without wind) the I save the trim with the option for CH7, the I center the trims, efectively the helicopter is trimmed, but the horizon in the mission planner, for a leveled helicopter on the bench (se same as previous) is not horizontal, the PITCH is OK the ROLL NOT

3691229562?profile=original

- Flying a mission, please check at this link "Mission".

I used SPLINE WAYPOINT, they work very well with multicopter, but for helicopter the tail doesn't follow the line of the curve, it seems that the YAW command is done but with great delay

- RATE_ROLL_D or RATE_ROLL_P, are both unusable, with very little values like 0,005 the shaking is un maneagable.

There is a reason ?

- RATE_ROLL or PITCH _IMAX is the time to dampen the I term, correct, I raised it to 150 to have an effect as I want, please could detail me better the meaning ?

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

Join diydrones

Email me when people reply –

Replies

  • Thanks.
    It's enough clear about PI and FF so I can do better tuning knowing the rational behind.
    But, I apologize, for the trim, this affect offsets for the control loop, the artificial horizon is for the pilot for this reason I'm stressing ☺.
    In any case this issue isn't important for me it is a nice to have.

    Finally about the yaw behavior during the mission, the command seems delayed, the helicopter does the turn but the coordination between helicopter direction and yaw are asynchronous, the yaw complete the maneuver after 60 70 % of the full turn.

    Did you experience similar behavior?

    Thanks Paolo
    • Paolo, I'm sorry, I have never had issues on T/R control with my helis so far. Better asking Robert or Randy, who have designed the system!

      I agree with you regarding artificial horizon, but, I assume you are not looking at that instrument while piloting...or at least if doing FPV, not rely only on that. 

      Few time ago, I made some experiments introducing a variable offset (a sort of mixing curve with collective_out signal) put straight into the error signal (bypassing then trim things and so on...) on the roll axis. In that case you didn't see the horizon moving of course, was a sort of variable trim being introduced in the controller of roll axis. So instead of self leveling to the horizon, went to horizon plus a "certain" (parametrized, so tunable) amount. 

      At the end of the story it worked good, but honestly I think it is much much easier using save_trim, not caring too much of what you see on the tablet :-) .

  • "

    - After a accelerometer calibration the articial horizon in the mission planner is as I expect for a leveled helicopter on the bench.

    Then I fly the helicopter, I trim it (without wind) the I save the trim with the option for CH7, the I center the trims, efectively the helicopter is trimmed, but the horizon in the mission planner, for a leveled helicopter on the bench (se same as previous) is not horizontal, the PITCH is OK the ROLL NOT

    "

    That's pretty normal, with the save trim function you are actually introducing an offset in the AHRS. The helicopter has to lean on the right to stop drifitng sideways (due to T/R thrust). So the horizon level is offset to the left, when the helicopter is straight level.

    "- RATE_ROLL or PITCH _IMAX is the time to dampen the I term, correct, I raised it to 150 to have an effect as I want, please could detail me better the meaning ?"

    Imax values should be the maximum output the controller is giving as feedback for an external disturbance. If you change them you should probably re-adjust you P-I values as well. 

    • Hello, I'm sorry but I disagree with helicopter leveled the artificial horizon shall show the correct position, I agree that when the helicopter is flyng it isn't leveled like at bench the torque doen't permit it.

      So, the accelerometer offsets don't modify the horizon level, they are only trims.

      The asks for 

      - Flying a mission, please check at this link "Mission".

      I used SPLINE WAYPOINT, they work very well with multicopter, but for helicopter the tail doesn't follow the line of the curve, it seems that the YAW command is done but with great delay

      - RATE_ROLL_D or RATE_ROLL_P, are both unusable, with very little values like 0,005 the shaking is un maneagable.

      There is a reason ?

      please could you give me an answer ?

      Thanks

      Paolo

      • So,regarding trims, make this simple test:

        go to full parameter list, change AHRS_TRIM_X to 0,1745 for example. Go back to flight data and, without moving the helicopter, see what happened to the horizon...

        Moreover it's not the torque itself, it's the horizontal (sideways) thrust generated by the tail rotor. Hencet he tip path plane of the main rotor, during a stable hover, has to lean on one side to give a proper horizontal component of the lifting force to counteract this thrust.

        I assume you are referring to RATE_ROLL_D and RATE_PITCH_D... I agree they are not that usable in heli firmware, but, after having setup many helicopters, It's just fine like that, proper tuning of P-I-FF makes thing working perfectly.

        have a look here, it's very well explained how the control loop works http://diydrones.com/forum/topics/arducopter-2-9-for-tradheli-discu...

  • The TLOG file with the mission if the link doesn't work

    Thanks

    Urka Durka

    2015-07-25 18-39-42.tlog

    https://storage.ning.com/topology/rest/1.0/file/get/3702054921?profile=original
This reply was deleted.

Activity

Neville Rodrigues liked Neville Rodrigues's profile
Jun 30
Santiago Perez liked Santiago Perez's profile
Jun 21
More…