In Reply to: The PSUD step function posted by tweakydee on July 2, 2009 at 20:02:17:
The stepped load can be programmed to change (increase or decrease) after
a desired delay in seconds.
I usually set up PSUD to run for 2 or 3 seconds of simulated time, allowing
plenty of time for the voltage to stabilize from initial start-up before
the current step, then another 1-2 seconds after. I don't bother with the
initial delay, and usually my designs settle so fast that 1 second before
and one second after (2 seconds total) is long enough run time. You need to
see a straight line with ripple.
I set the scale at full size first, then zoom in around the flat line part
of the trace to check ripple, and zoom in around the current step to check
transient response.
The current steps I program for class A amplifiers are about 10-20% of
idle current, i.e. the idle current is "value 1" and idle * 1.1 is
"value 2". For class AB, value 2 might be 3X or 4X value 1.
I look for a good well-damped response with quick settling time ( <20mS)
and little ringing or overshoot. The change in voltage divided by change in
current is an estimate of the dynamic impedance. BTW, the word "dynamic"
can also mean the parameter (slope of the line, dv/dt) at some place on a
curve, i.e. "dynamic transconductance"... In a power supply the dynamic
impedance might change when the load is changed.
There are a lot of posts in the archives with examples of this. Look for JLH posts on PSUD from about a year ago and some of my own as well.
Michael
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- RE: The PSUD step function - Michael Koster 07/3/0908:11:03 07/3/09 (2)
- great, thanks! nt - tweakydee 07:10:41 07/4/09 (0)
- RE: The PSUD step function - PakProtector 09:04:50 07/3/09 (0)