Risley Prisms Alignment
Farsang
Hi:
There is a ray coming at angle (~ 3 mrad) twoards an axis (Optics
Axis), we'd like to bend this ray so that lies right
on the Optics Axis. With a Risley Prisms Pair we can make the ray
coming out of the Risleys parallel to the
Optics Axis but the ray dose not coencide with Optics Axis and because
of the Risleys it has an offset from it.
Is there a way to get rid of this offset without using a window (a
piece of glass with parallel faces)
as a beam shifter after the Risleys?
Can adding another prism (a Risley wedge) and adjusting the spacings
between these wedges work? I tried this
in ZEMAX but still couldn't get rid of the output ray offset.
Appreciate your suggestions.
Dave Schaack
"Farsang" <youzp...@netscape.net> wrote in message
news:8183c145-8e8e-4f44...@e6g2000prf.googlegroups.com...
>
>
> Can adding another prism (a Risley wedge) and adjusting the spacings
> between these wedges work? I tried this
> in ZEMAX but still couldn't get rid of the output ray offset.
>
for doing the optimization if I were you. I suspect that you have
inadvertently asked ZEMAX to do two inconsistent things at once, so that it
effectively does nothing.
Farsang
> "Farsang" <youzpal...@netscape.net> wrote in message
Which one should work, two wedges or three wedges?
The meritfunction is very simple, the variables are wedges' rotation
angles and their spacing from each other. Optimization is to make
Angle Of Incident = 0 and X,Y local coordinate =0,0 on the surface
following the Risleys. This surface is always perpendicular to the
established optics axis (i.e. position of this surface is independent
of Risleys).
The input ray might not hit the first Risley Prism at its center
(which lies on the optics axis) and this causes the output ray from
the Risleys
to have more offset from the optics axis when Risley are adjusted to
make this output ray parallel to optics axis
Dave Schaack
>
>
> Which one should work, two wedges or three wedges?
Three wedges.
> The meritfunction is very simple, the variables are wedges' rotation
> angles and their spacing from each other.
Sounds like too many variables. That can cause the problem you are seeing.
For instance, if you intend to keep the chief ray confined to the y-z plane,
then you want to adjust the orientation of only one of the Risley prisms
(and set the other one's orientation using a pickup with negative sign).
Adjust only a single distance; that between the single prism and the Risley
pair.
> Optimization is to make
> Angle Of Incident = 0 and X,Y local coordinate =0,0 on the surface
> following the Risleys. This surface is always perpendicular to the
> established optics axis (i.e. position of this surface is independent
> of Risleys).
>
> The input ray might not hit the first Risley Prism at its center
> (which lies on the optics axis) and this causes the output ray from
> the Risleys
> to have more offset from the optics axis when Risley are adjusted to
> make this output ray parallel to optics axis
In order to get the output to coincide with the optical axis, the input must
be offset from the optical axis. It seems to me that what you should be
targeting is a particular offset between the input and the output. Whether
it is the input or the output that is considered to be the optical axis is
immaterial.