It seems that I can no longer use a 8/20 microseconds standard surge in FFTSES in the 2002 version of the CDEGS software. When I do this, an error message appears. What has changed?
Answer The message that you get in FFTSES indicates that an 8/20 us wave cannot be represented exactly with a double-exponential form. In the past, we would use an approximate approach to find the parameters of the double exponential that would best represent a given wave form. This approximate approach would give reasonable results in most cases, but there were some cases where the results were not very good. This happened, in particular, when the fraction time was short compared to the rise-time.
For the 2002 release, we have implemented an exact solution for the parameters of the double-exponential that gives precisely the requested values for the amplitude, rise-time, and fraction time. We have found that this exact solution can exist only when the fraction time is sufficiently large compared to the rise-time. This is the reason behind the error message.
We may re-introduce the approximate approach in the future, once we understand better its domain of (approximate) validity. In the meantime, you can follow the suggestion in the error message, and use a 8/22 us waveform (the difference in the results should be minimal).