究竟有没有明白人？？？
引用：
楼主，很幸运，几天前我也和一个国外朋友讨论过这个问题，mail转帖如下，不是和你炫耀英文，只是懒得翻译回来了，你就将就着看看吧，谢谢理解，看得出你很爱动脑筋。祝你好运！
I was checking my Hiribo swash plate setting this weekend, there are some thinking and feeling want share with you.
Problem found:
I was so surprised to find my swash plate was swing in the period of pitch changing from 0%~ 100%.
Problem analysis:
Actually ccpm swash plate is driving by 3 units servo together in same step,and the servo is turning in round, but the swash plate is moving in linear motion, so the transfer function is L=k*r*sin(θ+α). here L is the swash plate travel distance, k is a constant ratio, r is the radius of the servo arm,θ is the angel between the servo arm and the horizontal line,αis the center angel of the θ when the pitch is 50%.
Cause the assemble tolerance, and the tolerance of servo arm output axis or servo's teeth position, it is almost impossible to get the exactly horizontal servo arm when the input is 1520us for each swash plate driving servo. this tolerance will be relected in the valus of α. and each arm's moving speed will be .
For each servo's α ,for sure, we will get the different of L when we are changing the pitch,also it will get the different transient speed at each point, this is the reason when my Hirobo's swash palte is swing when I am changing the pitch.
Solution:
Here we assum these 3 servo is exactly same but we only have different center angle α. What I am thinking is we need well cancel this tolerance (α) or difference for each servo, we can accept the value of this α, but we can't accept we will have different α for each servo.
Precedure:
I will check the 50% position (center angle) for each servo, then find the value when the servo arm is at the exactly horizontal position. check the reciever 's output when the output is 50%, of course, the digital meter which you give me last time will be the useful tool this time. and subtrim will be right parameter to cancel this α.
Extension:
If we are using different servo to drive the swash plate, that will be another story, because the gain (L VS output from RCV or v VS output from RCV) will be different. so I think we should fine tune the subtrim to well center the servo, also we need well setting the travel range for each servo in separately, ideally, it will get the same static state swash plate position. but the speed of each servo will creat some transient speed difference issue.
Do you agree? waiting for your input.
Jimmy
[ 本帖最后由 NO4LY 于 2010-6-7 00:29 编辑 ] |
/2 
发表于 2010-6-6 22:24
收藏
楼主
