[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
E-TESLA 5.50 final...
With my latest improvement, I now believe E-Tesla is more accurate than
one can measure. I get 0.7% average and 2-3% maximum error on my test
grids from measured values. At that level, I no longer know if the small
error is in the program or my measurements.
I improved the accuracy of the secondary voltage profile greatly. I
didn't use lumped or T-line theory or any think like that. I just simply
stress averaged the thing ;-)) Dumb and simple, but it works! A simple
block of code before line 33000.
So I can no longer think of any basic theoretical things to do to the
program. It seems to be basically as good as it can get. The speed will
not increase that much faster nor will the practical accuracy. Perhaps a
breakthrough in secondary voltage profile theory but that would have to
take into account the surrounding objects... I don't think that will
happen for awhile.
Both the BASIC and C version are fast and clean. They should be easily
portable and easy to understand. Others may go ahead and add fancy
graphics and such but that is beyond "my" reach.
Many thanks to everyone who has contributed and helped me with this 17
month long project! It was a tremendous advantage to have the code in
simple BASIC so many people could quickly look at it and find all my errors
and suggest many wonderful improvements "i" never would have thought of.
Sarah's C version allowed me to examine the sensitivity of the program to
probe for errors. That was a key step.
I do not "plan" on doing more work on the program unless there is an
obvious need. It certainly fulfills it's original purpose and can be a
building block of other similar endeavors.
So I guess the "final" version of the program is now E-Tesla5.50 at: