random yawing while take-off or landing: SOLVED

[Kisssys]

Hi Angel,

Looking at that video you have something major out of wack and messing with the code is not the way to go. I have never seen an AQ do that no matter how far out of wack I've had it.

I would download the latest repo and start over with no modifications to the code.

Look at you mix table close, I don't think a quad would fly if the sums were not all zero but I'd make sure.

At 800 grams, what is your throttle factor setting?

[afernan]

Hi Steve:
thanks for your comments and your time.

I´m using the default throttle factor (700) with a large thro-dead band (120) to let PH be in the middle of the stick.

The mix table I use is:

------T-----P-----R-----Y
M2 100 50 50 -100 front left
M5 100 -50 50 100 rear left
M9 100 50 -50 100 front right
M13 100 -50 -50 -100 rear right

In this case I use Flyduino Simon-K ESC. Maybe that could be the problem: they couldn´t react the same under same PWM generated by the code. I use to set limits individually for each ESC using the Top-Bottom stick method. I guess in between limits they are linear, and maybe is not true.

Could you imagine a way to check that ESC influence?

Angel

[Kisssys]

Well, that's a huge deadband but may be ok if 700 is really what you need. With a deadband that large, when you turn off PH you could being climbing or descending at a high rate if you were at the limits of the deadband window.

When you are in manual flight what does your throttle position read when you are in a hover?
You need LOG_RADIO_CHANNEL0''s average at a hover divided by 700 times your current throttle factor to get what you need. Their is more to the throttle factor than just making a place to stay level on the TX. It properly distributes the power available over the total flight range.

In manual mode when you give it Yaw does it travel laterally also?

Your mix table looks ok to me.

It is very important to see that each motor starts at the same time and reaches the same rpm at max throttle. Simon_k's are fine but but each one needs to be calibrated using an external receiver so that they match. Don't fly them if they don't match pretty close.

[bluuu]

small notice:

Today I reconstructed "yaw problem".

Some facts:
- FW r79
- it was third flight on second LiPo (After change LiPo)
- full check list (connections, radio TX, connect via BT to GCS, check voltage (full - 16,8V), wait for 3d fix
- all ok
- take off (manual mode)
- after 2-3 seconds copter was yawing left(rather fast), then I put yaw stick right - copter was yawing right fast ... no neutral point
- i do emergency landing (with success, no demage)

after reconnect LiPo next flight was succesful (and 3 more flights without any problems).

I have log
Mayby helpfull. Somebody look at this ?

[afernan]

I don´t think this is the issue I talk in this post.

The one I talk Is solved by adding the code lines I mentioned before (I always incorporate in my code version). Hopefully some day will be added to the trunk.

Your problem looks like for some reason the board didn´t made a correct initialisation. I´ve never experienced such proble.

Best regards

Angel

[sandmen]

[skew]

Sandman, I just tried your code worked great for me. I did a landing from about 40m and it came down straight as an arrow no twisting around like it always has. This code is going in every compile of mine now.

[afernan]

@sandmen:
thanks for you code. I think your solution is better than mine since you can select the target heading for landing, in addition

I´ll give a try this WE, although it looks that it works.

Angel

[skew]

@sandmen, any suggestions on reducing or fixing the random yawing at long loitering WP or descent to a waypoint during a mission? here is a video that shows coming to a way point but the alt has not been meet. As it descends it will yaw until WP has been reached. It starts at 6:04 until it hits WP at 6:20 and your new landing code kicks in and goes into landing stage

[sandmen]

Can you send me your mission file?