This paper quantitatively analyses the Joule heating in surface micromachined Greek cross test structures with different arm lengths and degrees of arm undercut. Analysis has been performed using Finite Element (FE) simulations and sheet resistance measurements, which indicated that most Joule heating occurs in fully suspended (undercut) test structures with long arms. Simulations and measurements have shown that the Joule heating effect can be dramatically reduced if cross arms are kept short and the test structure is undercut by no more than 50% of its arm width. Using both guidelines, the sheet resistance measurement error associated with Joule heating can be reduced by 2 orders of magnitude.
展开▼
