Calculating Centroid Shifts

[dp]

I have a single lens that I am using to focus a collimated light
source into a glass subtrate. The lens is at about 15 deg aoi to the
substrate in the Y axis. The image plane of the lens is a few hunded
um below the surface of the substsrate.

I am trying to model the location of the centroid of the beam relative
to the refracted chief ray. I am using two different methos to
calculate this and I get two different results

Method 1. Recipolar Spot diagram
This methods calculates the centroid shift to be about +10um

Method 2. Line Spread Function
It appears that most of the energy is centered around the -10um
region.

Is there difference related to how, I am comparing the two
calculations, or are the results roughly the same, and is it just that
the sign convention is incorrect.

Don

[Brian]

On Dec 5, 4:39 pm, dp <dfperra...@yahoo.com> wrote:
<snip>

> I am trying to model the location of the centroid of the beam relative
> to the refracted chief ray. I am using two different methos to
> calculate this and I get two different results
>

<snip>
> Don

To get the line spread function (LSF) and knife edge distribution
(KED) using the geometric approximation, use the commands lke (to
print) and pgl (to plot). But this is only an approximation which
takles no account of diffraction. To get the exact values you should
use the diffraction calculation commands dlk (to print) or pdl (to
plot). Also you need to set a large number of aperture divisions (see
under Setup in the Surface Data Spread Sheet - e.g. 97.1) if such
small errors are important to you. Also do make sure the spectral
weights are correct if the light is polychromatic.

Brian
Ancient and Modern Optics