Final Thoughts

I set out to demonstrate MIMO converters as a viable solution for new power architectures, but experienced difficulty accomplishing my second objective of developing arbitrary NxN MIMO power loss model.In retrospect I hindered my progress by relying too much on Rodriguezes approach. While there is nothing wrong with his solution, his methodology considered only a single MOSFET, whereas my 2×2 MIMO converter incorporated 11 MOSFETs. This additional complexity proved to be too advanced for me. 

However, I did manage to successfully complete the first objective: power loss simulations in spice proving this 2×2 MIMO converter has comparable performance to traditional converters. The design can be optimized further though: sizing components, testing duty cycles and other SC topology (e.g. ladder). Although this experiment was limited in scope I believe it points out the merit in a MIMO converter based power architecture. Integrating control techniques with bidirectional ports could exceed traditional power delivery methods.


With respect to my difficulties in designing an analytic power loss model, the advice given to me my Professor Zhang is to apply further simplifications and more closely follow Seeman’s model: examine the two switching limits and only pay attention to the contributing power loss factors of impulsive transfer or switching loss.

An alternative approach to designing an analytic model could be further simulation tests. This was certainly where I saw the most improvement, as such a strategy could be to gather data on different simulations to develop an understanding of individual component power loss contribution. The disadvantage of this approach is that it lacks the intuition and speed with which a simple equation model could provide.