The general strategy we have for implementing a good PID control solution is as follows:

1. Learn enough about PID control, at an intuitive level with a minimum of math, to understand what makes a closed loop PID control system tick.
2. Perform measurements on the process you want to control, say a heating system, and analyse the numbers to determine its major characteristics.
3. Simulate the process and a PID controller using the PIDassist program, and tune the simulated PID control for best performance.
4. Use a SPLat controller, running our pre-written PID program, to control the real process. This is done using the settings found in the previous step, and uses PIDassist to fine tune the settings. Preferably you will use an MMi200DK216 (or its predecessor the MMi99DK216), as this whole tutorial is written around it. Later you can substitute your own choice of SPLat, maybe a custom unit we are making for you, as long as it has the right hardware resources.
5. Once the system is tuned up, cut the connection with the PC and run the program standalone in the SPLat.
6. If applicable, move the program over into the SPLat you intend to use for production, and integrate the PID function into your overall application (it is perfectly possible for the PID to be just one small aspect of a larger program).

However, before we get going, you need to be aware that not all closed loop PID implementations turn out to be straightforward, and without experience it may not be easy to tell whether your project will be straightforward or not. If your application is one where success is critical, you would be well advised to get someone with suitable expertise to check out your initial concept, and maybe your final design.