Attempt to fix #4559 by increasing RonDiode and GoffDiode

[casella]

This attempts to fix #4559 by making the system less brutally nonlinear when approaching zero voltage and current on the load towards the end of the transient.

[maltelenz]

As described in my comment this seems good from a "simulate nicely in WSM" standpoint.

I want to make clear that I have no expertise on the model itself or the domain.

[casella]

I want to make clear that I have no expertise on the model itself or the domain.

Sure. For that, we wait for @AHaumer's or @christiankral's input.

[casella]

I also have a (slightly rusty by now) degree in Electronic Engineering cum laude 😄

[AHaumer]

@casella the whole discussing was a little bit misleading, there are 2 examples of SMPM braking using a rectifier:

• Modelica.Electrical.Machines.Examples.SynchronousMachines.SMPM_Braking
• Modelica.Magnetic.FundamentalWave.Examples.BasicMachines.SynchronousMachines.SMPM_Braking
  Should we apply the same solution to the other example, too?

[AHaumer]

LGTM
but IMHO the same solution should be applied to Modelica.Electrical.Machines.Examples.SynchronousMachines.SMPM_Braking

[Esther-Devakirubai]

Backported by #4570

[casella]

@casella the whole discussing was a little bit misleading, there are 2 examples of SMPM braking using a rectifier:

* Modelica.Electrical.Machines.Examples.SynchronousMachines.SMPM_Braking
* Modelica.Magnetic.FundamentalWave.Examples.BasicMachines.SynchronousMachines.SMPM_Braking

Good point. I got a regression on the second one, but if the two examples are similar it may make sense to also fix the other one. @maltelenz wrote in another ticket that they also had problems in WSL with the first one, so I'm fine with PR #4563.

In fact, as I discussed here, we could even think of changing the default values for the underlying switching supply models. However, I think it's best to fix the regressions locally in 4.1.0 where they show up, and then consider changing the default in MSL 4.2.0. That requires some analysis of side effects and is not something I'd do in a rush.

@AHaumer do you agree?

[AHaumer]

@casella these two examples are pretty much the same except for the machine models which are 100% compatible speaking if connectros and parameters (only the internal representation is different).