Replies to This Discussion

DeveloperPermalink Reply by Olivier ADLER on July 3, 2011 at 10:08am

 

Did you check that all motors are horizontal ?

• ▶ Reply

Permalink Reply by Jose Angel on July 3, 2011 at 10:37am

Visually are ok... I checked with some friends and they also think they are ok.

I don't know if I would be able to distinguish if the arm is twisted 1 or 2 degrees.
Any trick to check it with accuracy?

• ▶ Reply

DeveloperPermalink Reply by jasonshort on July 3, 2011 at 10:51am

I looked at the log you sent. Your copter is yawing left very hard and is out of balance. It's trying to hold a 104° yaw but it's hanging around 89° That means your copter is using 60% of it's Yaw power to try and maintain it's position. That's pretty bad and needs to be fixed. You'll see all kinds of odd behaviors, even on pitch and roll with that issue. Look at all your motors and find the torqed ones.
Jason

• ▶ Reply

Permalink Reply by Jose Angel on July 3, 2011 at 11:28am

Jason, thanks for taking your time reviewing my log.
I will check again all the structure.

Maybe you could add some guidance to the troubleshoot section of the wiki regarding the logs. For instance, howto check if your frame is well balanced from the logs?

Everyone should learn howto to debug a copter based on a log.

Thanks.

• ▶ Reply

Permalink Reply by geir on July 3, 2011 at 12:03pm

It would be grate if we could get a quick whats what in the log, especially on the CTUN

• ▶ Reply

DeveloperPermalink Reply by jasonshort on July 3, 2011 at 12:10pm

Until I update the wiki, check the log.pde file. You can see clearly the values written for CTUN.

• ▶ Reply

DeveloperPermalink Reply by Olivier ADLER on July 3, 2011 at 12:07pm

 

There are many solutions for testing motor orientation.

 

I think that a good one (very precise and not expensive) is to remove propellers and mount instead test cylinders. You could use aluminium RC car wheels for this.

 

When all motors have their test cylinders reverse the multicopter on a very flat surface, and check that all cylinders do not have angles.

 

Some users are using precision levels. It's certainly longer and not as precise as the first solution.

 

This is for motor orientation.

 

After this you need to check that all motors have approximately the same thrust. Can be done with a tachometer. If all motors have the same rotation speed at full throtle (+/- 5%) then you should be ok.

 

If not, replace the faulty motors ESC.

 

Do not forget that if you have a crash, some components inside motors (or ESCs) can be distroyed. So a complete check should be done after a crash. Vibrations can appear after a crash because of bent shaft or destroyed roller balls. In this case the motor need to be replaced or serviced. Propellers need to be checked as well for correct balancing.

 

• ▶ Reply

Permalink Reply by geir on July 3, 2011 at 12:28pm

Just to be clear i meant like which value is the yaw set point (nav_yaw i guess), whats yaw_debug and so on. That's not so easy to read from the log file.

• ▶ Reply

Permalink Reply by Jose Angel on July 3, 2011 at 2:06pm

This is what i understand for the log screen (after reading log.pde), but maybe I'm wrong:

Rudder In: your yaw stick position

Rudder Out: yaw output to the motors

Yaw Debug: only applicable to yaw rate control (YAW_OPTION = 0)

Yaw: yaw sensor

Nav Yaw: only for navigation, target yaw angle

Yaw Error: how off are we pointed

Omega.z, Z from gyros

 

Not really sure about the differences between Yaw and Omega.Z..... maybe someone can elaborate more...

 

I'm sure that when devs can take a break from solving the last bugs, the wiki can be completed with this info.

 

Note:

reviewing the current troubleshooting section in the wiki, the key parragraph for me is:

"This is usually due to an airframe that it out of tune, either with one motor tilted slightly or the weight balance not centered. Typically, the bad behavior is that the quad will always yaw by a certain amount, typically around 30-45 degrees on takeoff, but will stay in that direction."

 

I should pay attention to this yaw on the takeoff!

• ▶ Reply

DeveloperPermalink Reply by jasonshort on July 3, 2011 at 7:07pm

Yeah, that's pretty close:

 

Rudder In: your yaw stick position

Rudder Out: yaw output to the motors

Yaw Debug: only applicable to yaw rate control (YAW_OPTION = 0)

Yaw: yaw sensor

Nav Yaw: target yaw angle - for heading hold or navigation

Yaw Error: how off are we pointed from heading hold

Omega.z, Z from gyros - rate of rotation of Yaw gyro

 

What happens is your quad would normally just spin around if the frame were out of whack. The yaw error builds up till your about 15-20° and the forces to return it to center balance out. This is what is called a steady state error. The i-term of a PID loop would normally solve this, but you loose so much control authority it's not worth it.

 

I'll add this to the wiki,
Jason 

• ▶ Reply

Permalink Reply by Rob. S on July 4, 2011 at 4:30am

hey

guys can someone review my log pls
it would b a great help
i have the same issue as Jose

 

Attachments:

• 04-07-11 09-09 1.log, 56 KB
• 04-07-11 09-09 3.log, 5 KB

• ▶ Reply