Very bad vertical axis scale with log Y

[ceraolo]

Consider the enclosed Filter.mo model. Run it and plot Rl.abs_v against time on a linear scale: we see that the span of abs_v is between 0.00026 and 4. When we switch to logY scale therefore we expect the y axis span to be between 10^{-4 and 10}1.
Instead, it goes from 1e-100 to 1e-20, which is very bad.