MSL port current probe size.

[dsa-t]

When running the MSL_Losses.m MatLab demo, I can see that the MSL current probes are rectangles placed around the conductor line (1.5 cells bigger on all sides):

Then, looking at ports.py file in the Python interface I see:

        i_prope_pos = u_prope_pos[0:2] + np.diff(u_prope_pos)/2.0
        self.I_filenames = []
        self.I_delta = np.diff(i_prope_pos)
        i_start = np.array(self.start)
        i_stop  = np.array(self.stop)
        i_stop[self.exc_ny] = self.start[self.exc_ny]
        for n in range(len(i_prope_pos)):
            i_start[self.prop_ny] = i_prope_pos[n]
            i_stop[self.prop_ny]  = i_prope_pos[n]
            i_name = self.lbl_temp.format('it') + suffix[n]
            self.I_filenames.append(i_name)
            i_probe = CSX.AddProbe(i_name, p_type=1, weight=self.direction, norm_dir=self.prop_ny)
            i_probe.AddBox(i_start, i_stop)
            self.port_props.append(i_probe)

Because of i_stop[self.exc_ny] = self.start[self.exc_ny],
it looks like this would create a thin line segment probe with length = conductor line width instead?

• Why there is this difference between Matlab and Python code?
• Which one is better?

(unless I'm misunderstanding something)

[thliebig]

Well I would have to look if python or Matlab are even doing it really differently, but at the end of the day it should not make a difference as openEMS will extend the current probe box to a half cell oversize anyways, because this is were the magnetic fields are located to integrate along. I guess running openEMS with the verbose option will tell you if you look at the snapped coordinates for the probes as reported by openEMS.