Tricopter yaws uncontrollably

[cwhummel]

So I just put together a tricopter with a hk multiwii pro, and whenever I try to takeoff, the servo just all of a sudden goes full left and the tricopter spins before it even gets off the ground.

Running mw 2.3
Hk 6ch tx/rx

Any ideas what's going on? I'll be happy to provide any more details, just not sure what to add. I didn't modify the mw code any other than necessary for frame, gps, and mag.

[waltr]

Try searching the forums here or a Google search.
I have read about this issue a number of times but since I haven't build tri-copters didn't play attention to what the solutions are.

Here is a thread on RCG that has a similar problem:
http://www.rcgroups.com/forums/showthread.php?t=2054587

[cwhummel]

i searched a little bit but couldnt find anything similar to my issue. the thread you posted isnt quite what im experiencing.

I have full control of the servo, and even when arm the tri and just start spinning up, everything is fine and the servo is holding the motor straight. it isnt until I just barely start getting off the ground that the servo goes full tilt left and the tri just starts spinning uncontrollably. I can hold the stick full right to try and counter, but that doesn't really help, or solve anything. The trims don't seem to have anything to do with it either.

[waltr]

Ok,
Sounds like the FC is trying to YAW the copter so re-check:
Are the motor/ESCs plugged into the correct pins?
IS the Servo plugged into the correct pin? (I did see aa RCG thread where this was the problem)
Are the three motors turning in the correct directs? Do they have the correct hand props (CW vs CCW).
Did you calibrate the Acc in MultiWii Config?
Are your tx/rc channels at 1500 when sticks are centered?
Does the servo properly respond to the Yaw stick?
Does the servo get enough power?
Is the FC in ANGLE mode or some other mode like MAG?

When you power on the FC is it motionless while it calibrates the GRYO?

[cwhummel]

Motors are plugged in as: front left D6, front right d5, rear d3, servo d2.

I'll look at the code later and try to find the servo pin in question (though im not sure it should matter as i have full control of the servo with the yaw stick)

motors and props are correct.

acc has been cal'd (a few times)

yes, channels are at 1500 when centered.

yes, servo responds to yaw stick.

yes, servo has enough power. it has its own bec.

have only tried acro and angle mode. neither make a difference.

when powered on, i let it sit until all the lights stop flashing.

I'll try to look into the code when i get a chance later.

[Hamburger]

Wrong servo direction? Change yaw direction. Read config.h and wiki how to

[cwhummel]

I'm not sure that would matter. Like I said, I have full control over the servo with the yaw stick. When powered on, the servo immediately centers, and even before I arm it, I can move it back and forth with the yaw stick. Even once I arm it and start spinning up, everything is fine, servo pointing the motor straight up. It isn't until I'm just about to get off the ground that it just goes full left for some reason.

[greenmen]

Hi, I'm facing exactly same problem.

Did You solve this issue? Could You tell us what was causing this?

[Magik30]

I also same problem.
but it is a little better with my settings on my radio position E.POINT RUD 120% / 107%
It is true that when accelerates without lifting Yan tilts a little worm right.

trying to do with # define MAG_DECLINATION 0.0f
now works better my Yaw

[bajacondor]

I also have this problem. Much searching shows man yaw problems, but not this exact thing. Has there been any progress by anyone?

[bajacondor]

After further troubleshooting, I can visually verify that changing the YAW_DIRECTION variable has no effect on the how the tricopter corrects for yaw in 2.3. In 2.2 changing between 1 and -1 causes the servo to tilt either opposite or the same direction as the induced yaw.

Is the YAW_DIRECTION define ignored in 2.3?

Thanks

[krebbi]

After further troubleshooting, I can visually verify that changing the YAW_DIRECTION variable has no effect on the how the tricopter corrects for yaw in 2.3. In 2.2 changing between 1 and -1 causes the servo to tilt either opposite or the same direction as the induced yaw.

Is the YAW_DIRECTION define ignored in 2.3?

Thanks

yes, it is ignored. use the GUI to reverse yaw-direction.

[AkaMrBill]

I had the exact same issue when I replaced my failed Turnigy TGY-5009OM with a HiTec HS-5055MG.

As it turns out, HiTec servos rotate backward than most servos. It took me a bit of digging to find this out.
So, here is what is needed in MW 2.3 to make things work.

In Config.h, the entry below ONLY sets the direction of movement of the tail servo on a Tri based on the Rud input from the receiver. Set it to 1 or -1 to suit your flying preference.
In "SECTION 2 - COPTER TYPE SPECIFIC OPTIONS " find the following and set to your preference.

Code: Select all

/* NEW: not used anymore for servo coptertypes  <== NEEDS FIXING - MOVE TO WIKI */
    //#define YAW_DIRECTION 1
    #define YAW_DIRECTION -1 // if you want to reverse the yaw correction direction

Now, here is what you are looking for to fix the "correction" based on Acc/Gyro.

In the Output.cpp file, locate the following in the "main Mix Table" section:

Code: Select all

#elif defined( TRI )
    motor[0] = PIDMIX( 0,+4/3, 0); //REAR
    motor[1] = PIDMIX(-1,-2/3, 0); //RIGHT
    motor[2] = PIDMIX(+1,-2/3, 0); //LEFT
    servo[5] = (SERVODIR(5, 1) * axisPID[YAW]) + get_middle(5); //REAR

In the last line, set the "1" to "-1". So, it should look like this:

Code: Select all

#elif defined( TRI )
    motor[0] = PIDMIX( 0,+4/3, 0); //REAR
    motor[1] = PIDMIX(-1,-2/3, 0); //RIGHT
    motor[2] = PIDMIX(+1,-2/3, 0); //LEFT
    servo[5] = (SERVODIR(5, -1) * axisPID[YAW]) + get_middle(5); //REAR

That will cause the correction from the Acc/Gyro to work with your setup.

[jrb]

You do not need to change Output.cpp to reverse the servo. As krebbi says above you can do this in MultiWiiConf23 (the GUI). However the GUI is a bit odd in this regard but this sequence works reliably for me:

Start MultiWiiConf23
Select the serial port
Click "start"
At the top of the window "SERVO" will appear between "MULTIWII.COM" and "SETTINGS", click that
Now in the servo screen you have to click the red button labeled "SERVO", it should turn green and more things will appear
One of the things that will appear is a red "GO LIVE" button. Click it, it will also turn green
Another thing that appeared above the "GO LIVE" button is a box labeled "Change Gyro/Acc Direction", click that, it should go solid white
Now click the green "SAVE" button
Click "CLOSE COMM"

Your servo is now reversed and your tri will stop spinning like a top Hope that helps someone.

[Hamburger]

You can also reverse servo direction via config.h.
But to find it you must read the comments in config.h and maybe even follow the link to the wiki.

Please, do not lead others to follow your path in editing code unneccessarily. It will only lead to more erraneous situations and irritated questions.

[AkaMrBill]

You can also reverse servo direction via config.h.
But to find it you must read the comments in config.h and maybe even follow the link to the wiki.

Please, do not lead others to follow your path in editing code unneccessarily. It will only lead to more erraneous situations and irritated questions.

You are partially correct. You can edit this in "config.h", BUT, as I said in my post, doing so ONLY changes the direction of the servo for the input from your transmitter! It does NOT change the direction of correction by the Acc/Gyro.
I agree that changing code without understanding what you are doing is a dangerous thing to do. That is why I specifically spelled out in detail what to do and why. Please re-read my post carefully and you will see, not only did I read the comments to find the section, but I quoted them in my post.

If the info I gave was false, please show us why and give specific details.

[fidelp11]

I have the same exact issue, it only happens when I am about to take off and have perfect control of servo before disarming.

Can anyone confirm that reversing servo will fix this or anyything else in this page?

[Ddpemberton]

Reversing the servo from within the flight controller should fix the problem. If you reverse the channel in the remote it won't do anything. I've only had this problem on CC3D boards, but all you have to do there is click a yaw reverse button that fixes the problem.