Is it possible to mount an APM or something similar off of the CG? If you know of any boards that are capable of this I would really appreciate the info. My current copter has an extra prop in the middle and I was wondering if it would be possible to mount the controller in front of it
Thanks
Replies
Development team ... can you pipe in here.
Thomas - IMUs consist of one or more sets of x/y/z gyros and accelerometers. Place the Flight Controller (FC) onto the EP (Electronics Platform). Then tilt the ship.
- the x, y, and z gyros tilt the same amount in degrees relative to gravity no matter where they are located on the EP plane
- the accelerometers, however, are affected by locations. if the ship is tilted up along the y axis:
... if located at the center of tilt, theoretically these IMUs sense no acceleration
... if located along the leading side of upward tilt, sense accelerations of +z and -y but not x
... if located along the trailing side of upward tilt, sense acceleration of -z and +y but not x
... the magnitudes of which depend on how far the accelerometer IMUs are located away from the center of tilt
Thus the accelerometer error terms are a function of the gyro readings and correctable. But i'm not sure if that is how it works in all FCs, some, or none. what i do know is that i've put the FC away from center and it has no effect.
So try it. Let me know if you see unusual behavior. Maybe someone from the development team will pipe in and tell us how it works and any limits.
The only limit I see is making sure that the center section is stiff so tilt is tilt, not frame flex.
The equation for the relative angular acceleration at our flight controller (I'll call it AFC) is:
AFC = AIC + a×r +w^2*r
Where AIC is the acceleration of the body's center of gravity, which is assumed to be the center of rotation. This would need to be done for all axis and the results biased against the actual sensor readings
This is what I'm assuming is the case, I was hoping someone with more knowledge could chime in - I'm glad you've had success mounting the fc off of center. I guess I need to test it
so many unknows.
what i thought happens is that there are three gyros (x, y, & z) in the IMU that measure angle. [there are also three accelerometers] Mathematically, one can then put the gyro angles against the clock and you can calculate the first derivative (angular speed) or second derivative (angular acceleration).
as far as i know you can put the FCU a km from the center and it will fly OK. just make sure that the Electronics Platform is stiff so what the FCU sees is what is actually happening at the center. that is the biggest issue to my knowledge.
You are right, I did a little research and it seems like that's how it works.
I will try this this week, your right we can use the derivatives to offset the accelerometers input if we need too, hopefully that is not too much of an issue though - I am not familiar enough with the code to attempt to even begin to implement something like this
hopefully that math is all ready in the FCU code. it really has to be since CG is such a fickle and theoretical place.
This will be an interesting discussion because it's all about the math. So let me start this off by saying that any controller that does not allow offset mounting isn't doing the math because there isn't a location on a quad that is at CG of which there are two:
1) Mass CG of the ship that changes as the ship flexes
2) Lift CG of the ship that also changes as the ship deals with air and ship variations.
.