David's Tricopter problem

[DoctorKira]

Hello,
so i've built davids tricopter 2.65, but with a 3s battery(2800mah,30c).
The problem is, when I arm my tricopter I need to apply ~3/4 of throttle to get in even moving from the ground, but when it lifts (~1cm) it just goes forward and to the left and almost flips, without me touching anything. All motors are spinning the correct directions, Im using CW SF 9x47 props. Any help?
Thanks

 

[Tactical Ex]

Have you calibrated the ESC to the motors?

 

[DoctorKira]

yes, I did calibrate my motors, also when I try to counter the roll , it doesnt do anything(when i move my aileron stick to the right), but when I move to the left, it just instantly flips

 

[DoctorKira]

Update: I added more points to stick scaling on throttle, now at 150, now lifts with 1/2 throttle, but still flyes to the left

 

[Tactical Ex]

What happens when you input left and right aileron control at no throttle? Do motors 1 and 2 behave as expected?

 

[DoctorKira]

What happens when you input left and right aileron control at no throttle? Do motors 1 and 2 behave as expected?

You mean 0 throttle? They dont do anything.

 

[DoctorKira]

It seems like M1 (left motor) isnt generating any lift, could changing the prop to CCW help? I read somewhere that it would, but it doesnt make sense, the motor is rotating CW.

 

[Tactical Ex]

Very odd to me, it's worth double checking that m1 is spinning the right way, it would explain a lot of the behaviors. If that's not it hopefully someone more knowledgeable with multirotors will chime in. Other than that I'd start checking to make sure that m1 is configured correctly with your control board, it might be plugged into the wrong servo pin or the multi rotor type isn't set up properly on the board.

 

[DoctorKira]

Very odd to me, it's worth double checking that m1 is spinning the right way, it would explain a lot of the behaviors. If that's not it hopefully someone more knowledgeable with multirotors will chime in. Other than that I'd start checking to make sure that m1 is configured correctly with your control board, it might be plugged into the wrong servo pin or the multi rotor type isn't set up properly on the board.

Connections are double checked, directions also. All Im thinking now is to change M1 prop to CCW. Because other motors are generating lift spinning same direction with same props, and M1 is spinning other direction with same prop as others, and not generating lift. Ill try some CCW props on this. Anyway, thanks

 

[RoyBro]

Connections are double checked, directions also. All Im thinking now is to change M1 prop to CCW. Because other motors are generating lift spinning same direction with same props, and M1 is spinning other direction with same prop as others, and not generating lift. Ill try some CCW props on this. Anyway, thanks

I may be misunderstanding what you just typed here. It sounds like you are using the same prop on all rotors, but just have M1 spinning the opposite direction. For clarification, which directions is each rotor spinning? And what prop size is on each rotor. Look on the prop and it should have some indication like 9x4.7 or 9x4.7-R. or there may be a CW or CCW. Of course you have to have the right type of prop for either CW or CCW rotation.

 

[Balu]

Are the props mounted the right way (numbers up)?

Your props need to be the right ones for the motors. For CW rotating motors you need CW props and vice versa. Flipping a CW prop to be mounted on a CCW motor does not work.

 

[DoctorKira]

I may be misunderstanding what you just typed here. It sounds like you are using the same prop on all rotors, but just have M1 spinning the opposite direction. For clarification, which directions is each rotor spinning? And what prop size is on each rotor. Look on the prop and it should have some indication like 9x4.7 or 9x4.7-R. or there may be a CW or CCW. Of course you have to have the right type of prop for either CW or CCW rotation.

So, motors are spinning directions like kk2 told, M1 is CW, M2 and M3 CCW. All props are the same, 9x4.7 CW. But the thing is, CCW rotating motors are spinning CW props and generating lift, while CW motor with CW prop isnt.

 

[DoctorKira]

Are the props mounted the right way (numbers up)?

Your props need to be the right ones for the motors. For CW rotating motors you need CW props and vice versa. Flipping a CW prop to be mounted on a CCW motor does not work.

Props are right way up. I know, but David said, that you can use same props on a tricopter. Now I know. But the thing is CCW motors with CW props are generating lift, while CW with CW isnt

 

[Craftydan]

DK,

First off, welcome to the forum!

Second, do you have any video or pictures of your setup? it's amazing how much that can help.

What you're describing is an "insta-flip" -- the control board/ motors are not laid out properly in some way and the board *thinks* it's making the board more stable, when in fact it's corrections are only making things worse. this can be seen by manually testing boom drops (with the props off), but it's hard to see if you don't know what you're looking at. take the pros off and throttle up to half. hold the frame in your hand and tilt the platform. The motors should oppose the motion -- the boom that drops should briefly speed up in response (it may stay a that speed if you've got Self Level on). **until this is right, there is NO SENSE in putting the props back on**

the issue is the board and motors are not laid out in the correct alignment. you've not mentioned which control board you have, so hard to say what the correct assignments for ESCs are. For the board, it should be face up and if it's got a direction arrow, it needs toward the front of the tri. Triple check your ESC connections to the board against your board's documentation -- if they're out of order, it will flip. every time.

Motor direction isn't as critical on a tri as a quad, but if you can match the recommended as best as possible, it'll fly better . . . but since most of the yaw correction comes from the servo, it'll still fly without spinning wildly, so long as the servo doesn't need to be reversed. get the first issue fixed as that might not be a problem.

 

[DoctorKira]

DK,

First off, welcome to the forum!

Second, do you have any video or pictures of your setup? it's amazing how much that can help.

What you're describing is an "insta-flip" -- the control board/ motors are not laid out properly in some way and the board *thinks* it's making the board more stable, when in fact it's corrections are only making things worse. this can be seen by manually testing boom drops (with the props off), but it's hard to see if you don't know what you're looking at. take the pros off and throttle up to half. hold the frame in your hand and tilt the platform. The motors should oppose the motion -- the boom that drops should briefly speed up in response (it may stay a that speed if you've got Self Level on). **until this is right, there is NO SENSE in putting the props back on**

the issue is the board and motors are not laid out in the correct alignment. you've not mentioned which control board you have, so hard to say what the correct assignments for ESCs are. For the board, it should be face up and if it's got a direction arrow, it needs toward the front of the tri. Triple check your ESC connections to the board against your board's documentation -- if they're out of order, it will flip. every time.

Motor direction isn't as critical on a tri as a quad, but if you can match the recommended as best as possible, it'll fly better . . . but since most of the yaw correction comes from the servo, it'll still fly without spinning wildly, so long as the servo doesn't need to be reversed. get the first issue fixed as that might not be a problem.

Thanks for the reply, I'm using KK2.1.5, here's a video and some photos:

Sorry for VERY unprofessional video, you can just barely see whats happening, and also don't know why 360p...
Edit: this is me just giving it throttle

 

[Craftydan]

Ok . . . looks like orientation of the board is fine, and the cables appear to run to the right ports. (the twisted one is going to the rear ESC, right?)

I see you're using all the same rotation props (I was getting confused by your description of the setup) -- in this case, ignore the direction from the motor layout screen -- it's really talking about the preferred direction for props -- as I've said before, critical for a quad, but not so much here.

All three props should blow air down and if you set M1 according to the layout, it will not . . . swap any two ESC leads to M1 and it should behave better for you.

 

[DoctorKira]

Ok . . . looks like orientation of the board is fine, and the cables appear to run to the right ports. (the twisted one is going to the rear ESC, right?)

I see you're using all the same rotation props (I was getting confused by your description of the setup) -- in this case, ignore the direction from the motor layout screen -- it's really talking about the preferred direction for props -- as I've said before, critical for a quad, but not so much here.

All three props should blow air down and if you set M1 according to the layout, it will not . . . swap any two ESC leads to M1 and it should behave better for you.

OMG, I love you! I can call it a flying tricopter now! Thank you so much!

Another thing, you say it will fly better if I hook M1 according to the FC, what would I need to change to make it fly? Put a CCW prop?

 

[Tactical Ex]

OMG, I love you! I can call it a flying tricopter now! Thank you so much!

Another thing, you say it will fly better if I hook M1 according to the FC, what would I need to change to make it fly? Put a CCW prop?

Its really up to you. By changing the direction the motors spin you have the opportunity to reduce negative torque movements. In tricopters you have an odd number of rotors which makes it virtually impossible to completely eliminate which is why the rotation of the rotors in less important.


Using this image as a reference


M1 and M2 counter act each other but M3 still will pull to one side, requiring input from the tail servo to negate it.

In this particular example the front rotors rotate inward which I think is not optimal. If my tri hits an object like a tree branch I don't want the props pulling the object toward my camera and other electronics, I'd rather have them rotating outward and push objects away from the inner body and increase the chance I just bounce off the object.

 

[DoctorKira]

Its really up to you. By changing the direction the motors spin you have the opportunity to reduce negative torque movements. In tricopters you have an odd number of rotors which makes it virtually impossible to completely eliminate which is why the rotation of the rotors in less important.


Using this image as a reference
View attachment 29726

M1 and M2 counter act each other but M3 still will pull to one side, requiring input from the tail servo to negate it.

In this particular example the front rotors rotate inward which I think is not optimal. If my tri hits an object like a tree branch I don't want the props pulling the object toward my camera and other electronics, I'd rather have them rotating outward and push objects away from the inner body and increase the chance I just bounce off the object.

That's true, about the pushing away thing. But if I do change the motor spinning direction, I will get more rudder control to one side, right? Because now it's using quite a bit of it to counter all 3 motors.
As a side note, just crashed my tri :/ It looked cool with a lot of yaw speed, but it didn't level itself out when I stopped, and crashed to the ground. Lesson learnt: don't do too quick movements in auto-level or else KK2 will forget where level is.

 

[Craftydan]

Glad you're up and flying

Have you updated the ROM yet?

The stock ROM for the kk2.1.x boards is garbage, and their port of 1.6 is just as bad. It's good enough to test the hardware, but I wouldn't trust it in flying. Look into flashing either the latest Steveis or RC911 ROMs. Not only will it improve the behavior, it'll make it fly much nicer . . . even more than getting opposing prop rotations.