Compensation design of power converters

1. A method comprising:

computing, by a power converter design system in a server, values for compensation components in a compensation network by minimizing an objective function given by O(c)=Se (c)+T RC (c)+S PM (c)+P(c) where c is a vector of values of the compensation components, Se(c) is a sum of weighted square error of a loop specification, T RC (c) is a sum of weighted resistor-capacitor (RC) time constants of the compensation network, S PM (c) is a weighted slope of a phase curve at a gain cross-over frequency, and P(c) is a penalty function;

selecting, by the power converter design system in the server, passive components for the compensation components in the compensation network based on the computed values.

2. The method of claim 1 , wherein the loop specification comprises crossover frequency (Fco), phase margin (PM), Gain Margin (GM), and low frequency gain (LFG).

3. The method of claim 1 , wherein the objective function comprises in part a loop-gain model of the power converter design to compute loop specification values for candidate values of the components, the derivation of the loop-gain model including power stage parasitics and non-ideal behavior of a feedback amplifier in the power converter design.

4. The method of claim 3 , wherein the loop-gain model is a unified model.

5. The method of claim 1 , wherein minimizing an objective function comprises using a hybrid optimization algorithm to minimize the objective function, the hybrid optimization algorithm combining a differential evolution algorithm and a conjugate gradient algorithm.

6. The method of claim 1 , further comprising:

identifying the compensation components from the power converter design; and

determining the minimum, maximum, and target values of the loop specification.

7. The method of claim 6 , further comprising receiving user specified changes to at least one of the minimum, maximum, and target values.

8. The method of claim 1 , further comprising receiving user specified optimization tuning indicating relative importance of performance and robustness, wherein one or more weights in the objective function are changed based on the user specified optimization tuning.

9. A system comprising:

a component database in a server, the component database configured to store data regarding passive components;

a power converter design tool in the server, the power converter design tool is communicatively coupled to the component database and configured to:

compute values for compensation components in a compensation network by minimizing an objective function given by O(c)=Se(c)+T RC (c)+S PM (c)+P(c) where c is a vector of values of the compensation components, Se(c) is a sum of weighted square error of a loop specification, T RC (c) is a sum of weighted resistor-capacitor (RC) time constants of the compensation network, S PM (c) is a weighted slope of a phase curve at a gain cross-over frequency, and P(c) is a penalty function;

select passive components for the compensation components in the compensation network based on the computed values and the stored data in the component database.

10. The system of claim 9 , wherein loop specification comprises crossover frequency (Fco), phase margin (PM), Gain Margin (GM), and low frequency gain (LFG).

11. The system of claim 9 , wherein the minimization is performed by execution of a hybrid optimization algorithm combining a differential evolution algorithm and a conjugate gradient algorithm.

12. The system of claim 9 , wherein the values are further computed based on power stage parasitics and non-ideal behavior of a feedback amplifier in the power converter design.

13. The system of claim 9 , wherein the values are further computed based on poles and zeros placed on a Bode plot by a user of the compensation design tool, the Bode plot displayed in a graphical user interface of the compensation design tool.

14. The system of claim 9 , wherein selection of passive components is further based on a user specified optimization factor.

15. The system of claim 14 , wherein the user specified optimization factor indicates relative importance of a footprint of a passive component, cost of a passive component, and efficiency of a passive component.

16. The system of claim 9 , wherein the objective function comprises in part a loop-gain model of the power converter design to compute loop specification values for candidate values of the components, the derivation of the loop-gain model including power stage parasitics and non-ideal behavior of a feedback amplifier in the power converter design.

17. The system of claim 16 , wherein the loop-gain model is a unified model.