Design constraints correlation analysis for the automatic design of a Miller OTA

This paper presents a design constraints correlation analysis for an analog integrated circuit automatic sizing tool to synthesize the optimization process constraints. The specifications of an operation amplifier can be very correlated, making that a correlated pair force each other to a value even if one of them is not a constraint in the optimization process. To evaluate how correlate the specifications of the circuit are the data must be correlated by ranking because of the non-linear space for the evolution of the specifications. For this purpose, the Spearman ranking correlation is used. This type of correlation produces a coefficient which indicates how correlated a pair of data is in a positive or negative way. To analyze this correlation a Miller OTA in 0.25 μm technology is designed for a set of specifications using an automatic sizing tool. Results show that the design considering the correlation analysis contribute to better solutions in terms of power dissipation.