Learning Forum

[Joshua Chen]

I have a simple question. How come when I run a Lumerical sweep, when I use too few number of sweep points (2-4 values), the results for a certain sweep iteration will be inaccurate compared to when I use more sweep points (e.g. > 5)? 

For example, I ran a sweep with 3 or 4 points here: 

The value for the x-axis "Radius" = 0.15 will return a phase of ~1.7. However, when I run the sweep with 5, 6, 7, or more points, Radius = 0.15 will return a phase of ~7.8? 

In the parameter sweep, aren't all of the simulation sweep steps run independently? Why is it dependent on the number of sweep points? Thanks so much.

[Amrita Pati]

Hi Joshua,

You are correct, the sweeps should run independently. Are these single sweeps (that is are you only sweeping the radius)? Is there any chance that another parameter is being swept along with the radius? Also I believe, the last point is the one breaking the trend? All the other points look accurate, is that so?

Regards,
Amrita

[Joshua Chen]

Hi Amrita,

Yes, only the last point is breaking the trend. This is a 2D parameter sweep ("sweeping" height as well, though for only one value of height = 1.3um). The only variable I'm changing in between sweeps is the number of points in the sweep.

When I run it for three points, the last point is again a different value, different from when there were 4 points or even more points. 

Thanks so much!

Best,

Josh

[Amrita Pati]

Hi Josh,

This is unusual. What happens if you do a single simulation for this last sweep paramater? That is change the radius to 0.15 um, and run a single simulation. Also, would you be able to share a screenshot of the Sweep Utility, where the parameters are set up.

Regards,
Amrita