Having recently published Graphic demonstration of loss under standing waves, I have received a stream of emails asking whether some online calculator or another is subject to the errors discussed in the article.
If we look at the graph from the article showing the example of a load of 500+j0Ω on 5m of RG58A/U, test the left hand data point at 1MHz and see what you get. If the loss is greater than the matched line loss (around 0.08dB), the calculator probably uses one of the flawed formulas discussed in the article… in any event it is wrong.
Here is an example from KV5R.com.
The calculated loss is around 4 times what it should be.
A lot of these calculators are copies of one original calculator since withdrawn. If imitation is the sincerest form of flattery, these copyists think more of the original calculator than its author.
Essentially, any calculator where you specify VSWR to determine the loss under mismatch is exposed to this problem.
My own calculators that provide that option are just as exposed… but they warn the user of the accuracy issue.