Browse Prior Art Database

Using Equivalent Values to Improve Convergence in Power Electronics Simulations

IP.com Disclosure Number: IPCOM000101342D
Original Publication Date: 1990-Aug-01
Included in the Prior Art Database: 2005-Mar-16
Document File: 2 page(s) / 75K

Publishing Venue

IBM

Related People

Kelkar, S: AUTHOR

Abstract

Device level simulation of switching converters is a technique used to obtain valuable design information, such as component voltage and current stresses and power dissipations. A major problem in reaching steady-state convergence in these ASTAP (*)- or SPICE-based simulations is caused by the wide range of time constants. The switching frequencies are usually of the order of 200 KHz, while the dominant output filter frequency is about 250 Hz. In order to reach steady- state convergence, it is necessary to use a trial-and-error procedure to optimize the initial conditions, as well as reset the maximum number of passes and the minimum step size. Simulation for a longer period of time may also be needed.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Using Equivalent Values to Improve Convergence in Power Electronics Simulations

       Device level simulation of switching converters is a
technique used to obtain valuable design information, such as
component voltage and current stresses and power dissipations.  A
major problem in reaching steady-state convergence in these ASTAP
(*)- or SPICE-based simulations is caused by the wide range of time
constants.  The switching frequencies are usually of the order of 200
KHz, while the dominant output filter frequency is about 250 Hz. In
order to reach steady- state convergence, it is necessary to use a
trial-and-error procedure to optimize the initial conditions, as well
as reset the maximum number of passes and the minimum step size.
Simulation for a longer period of time may also be needed.  Using
equivalent component values, as shown in this disclosure, it is
possible to attain steady-state convergence with speed and accuracy.

      The key contribution in this disclosure is a procedure to pick
a set of equivalent values for the output filter capacitor and its
ESR, such that steady-state convergence is easily reached while
preserving the key contributions of the output filter.  The
equivalent values produced will be, in general, not physically
realizable, but that is of little concern for a simulation tool as
long as no errors are introduced.  Another key point to be noted is
that the equivalent values are to be used in a steady-state
simulation only; transient response, cross-regulation and stability
analysis all require the actual values.

      The main purpose of the output L-C filter is to provide
attenuation at the ripple frequency and its harmonics.  The output
filter capacitor also carries the ripple current which shows up as
the ripple voltage at the output.  A three-step procedure is used to
pick a set of equivalent R,C values for the output filter capacitor
and its ESR....