I built two CVD systems when setting up my lab. They were basically clones of a system I had used previously, so the design was already in my mind and mostly down on paper. However, I severely underestimated the time involved in selecting all the proper pieces and parts. By the time things finally came together and the first system came on-line, I found myself wishing I had simply purchased complete solutions from a company that had already invested the time to engineer a reliable system. As a new faculty member, funds are a limited resource, but time is even more so. Quickly getting the ball rolling on research is worth more than what one might save with a home-built approach.