Which would be more beneficial?

[Adam Y.]

College graduation is two weeks away and I figured that I would want to start a new project. Namely I want to write a white paper on a technical subject. The problem is that I can't make up my mind as to what I want to write the white paper on. I can't figure out whether or not I want to write a white paper on either control theory or FPGA/VHDL. Control theory would be the more difficult subject because of the fact that my skills involving descrete time systems (What every team is typically using to implement their control algorithums) are non exisistant. FPGAs and VHDL would be the other topic. Yes I know the FGPAs are configurable via Labview but I am sure someone would be interested in learning how to implement designs with written code.

[FRC4ME]

I would go with the VHDL paper because the FIRST community already contains a vast knowledge base of control theory, but FPGAs and VHDL are new to FRC, so no one has done much work with them.

[JesseK]

I wouldn't say that the FIRST community contains a *vast* amount of control theory. Most teams use PID since it's "good enough" for first-order feedback, but precision control is usually derived from multiple sensors and proper pole/zero selection from the imaginary plane. How you set your constraints determines what regions you can select the poles/zeros from, and then some math will give you the 2nd or 3rd order equations (based upon the # of sensors) with exact coefficients. Running the driver control input will result in the desired motor output with behavior that is easily observable and reconfigurable by looking at the z-plane.

Ok, so maybe we know a bit about it but there aren't any white papers in CD-media . I only took an intro course in college on it.

[Adam Y.]

Originally Posted by JesseK

I wouldn't say that the FIRST community contains a *vast* amount of control theory. Most teams use PID since it's "good enough" for first-order feedback, but precision control is usually derived from multiple sensors and proper pole/zero selection from the imaginary plane. How you set your constraints determines what regions you can select the poles/zeros from, and then some math will give you the 2nd or 3rd order equations (based upon the # of sensors) with exact coefficients. Running the driver control input will result in the desired motor output with behavior that is easily observable and reconfigurable by looking at the z-plane.

Ok, so maybe we know a bit about it but there aren't any white papers in CD-media . I only took an intro course in college on it.

I did too but even that class only scratched the surface. I know there are four different ways to implement a controller with the math being different for each case. These methods are transfer functions which are the z and Laplace transforms, differential equations, state space models which models a system with a series of vectors, and h-infinity which I have no clue what is going on. Also, the subject also does reach into other areas that would be beneficial. My control system class derived the reason why the motor constants are the way they are.