Thread: Discussion ECILOP Easy
View Single Post
Old Nov 18, 2012, 05:55 AM
stanleyy is offline
Find More Posts by stanleyy
Registered User
Slovenská republika, Bratislavský kraj, Bratislava - Podunajské Biskupice
Joined Feb 2011
836 Posts
Originally Posted by Zaitsevsky View Post
One month ago I send my suggestions about Naza to DJI , but it may be useful for another controllers:

1. While setting a quadrocopter output M5 and M6 are not used. I offer to make an option which allows to give to these outputs the signals for a gyro gain control. Signal GAIN for the gyro TILT may have one meaning chosen by the user. Signal GAIN for the gyro axel ROLL has to have two meanings: one for vertical position and another in case of camera skewed forward or backward on the axel TILT. The case is when the cameras gimbal skewed forward or backward it is preferable to lower the GAIN for the gyro of axel ROLL. These three values are to be chosen by user experimentally. For the proper work of all types of the gyro it is necessary that PWM impulses would be submitted in the same order as on the helicopter: first the signal of position, then GAIN signal, then a gap (not less than 10 msec).

2. Add a possibility of more precise linearization for setting automatic compensation. Now we have three parameters set: neutral position, two end points, ratio. I suggest taking a table with 9 meanings for different angles: -40, -20, -10, -5, 0, +5, +10, +20, +40 degrees. So user has to skew the gear frame to the indicated angle and choose the meaning of compensation signal for each position. Program makes a more precise compensation graphic based on the chosen meanings. This function could be as a separate menu point “Exact settings”

3. Produce own external MEMS gyros, which are attached to NAZA controller. That is two little sensors with flexible wires of necessary length set on the gimbals axels. In this case servos are attached directly to controller avoiding using extra gyros.
Your point 2) I have implemented by myself using external arduino connected between Naza and gyros. Was not worth the effort, the output was not much better (I did software that measured a few points in the interval and created a 3rd order polynomial to compensate the shift).
But anyway, it could be interesting option to play with
stanleyy is offline Find More Posts by stanleyy
Reply With Quote