Anotace:
This paper presents utility for the design of the cross-coupled charge pump, which is used for supplying peripherals with low current consumption on the chip, as the EEPROM or FLASH memories. The article summarizes the knowledge in the field of the theoretical and practical analysis of the cross-coupled charge pump (design relationships and their connection with the pump parameters, as the threshold voltage, power supply voltage, clock signal frequency, etc.) that are applicated in the design algorithm. Optimal MOSFETs sizes (W, L) were find based on the construct of the time response characteristics of the pump sub-block and finding of the maximal voltage increase in the active interval of the clock signal and minimizing of the pump losses, as the switch reverse current, inverter cross current, etc. Synthesis process includes the design of the pump functional blocks with dominant real properties, which are described based on BSIM equations for long channel MOSFET. The pump stage complex model is applicated for estimation of the number of pump stages via state-space model description and using of the interpolation polynomial functions in the algorithm. It involves the construction of the time response characteristic due to the state variables and prediction of the number of the pump stages for the next cycle based on the previous data. Optimization of the pump area is based on the minimizing of the main capacitor in each of the pump stages (number of the pump stages must be increased to obtain the desired output voltage value.) Access is designed to stress the maximum pump voltage efficiency. The whole procedure is summarized in the practical example, in which the solution is shown both in terms of maximal voltage efficiency and the optimal pump area on a chip with respect to the clock signal frequency. Added functions of the design environment are explained, inclusive of the designed pump netlist generating for professional design environment Mentor Graphics including the real models of components that are available in library MGC Design Kit. The procedure gives designer credible results without long timeconsuming optimization process. In addition, the complex model allows the inclusion effects of higher-levels.