4f system alignment with fluorescent light

[Kyle Michael Douglass]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Hi all,
I'm currently mentoring a student who will be designing, building, and aligning a 4f system at the output port of a widefield fluorescence microscope. What general tips and advice can I give her when she does the alignment of the 4f system? I don't want to do the work for her, but I also don't want her to be left without any help.

When I align such a system, I image a coverslip sparsely covered in sub-resolution fluorescent beads and jitter the objective up and down rapidly. I then adjust the x-y positions of the two Fourier lenses until the PSF spot size is minimized and symmetric. This approach requires a bit of "natural touch," which I find often frustrates students since they would like to first see a list of principles and rules. She hasn't done much alignment, neither with laser light nor incoherent light.

The microscope is custom built and uses a 1.49 NA/100x Nikon TIRF objective with the corresponding tube lens. The 4f system will consist of two identical 2" diameter achromats mounted in x-y translation stages. She will decide on the appropriate focal length, though I've suggested choosing such lenses with focal lengths as long as possible.

Many thanks, and happy New Year!

Kyle Douglass, PhD
Post-doctoral researcher
The Laboratory of Experimental Biophysics
EPFL, Lausanne, Switzerland
http://leb.epfl.ch/

[Reto Fiolka]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Hi Kyle

For such alignments, I recommend the use of the following tools and
procedure:

1) a shear plate

2) a laser diode in a mount that can be screwed into the nosepiece in place
of the objective.

Procedure:
With the shear plate, I make sure that a collimated beam that goes into the
4f is still collimated when it exits the two lenses. Further I determine the
position, relative to the lens housing, of the focal planes (for collimated
input). This helps to find the relative distances between the lenses and the
distance from the first lens to the intermediate image plane, as well as the
distance to the camera for the second lens. It is important to know that the
principal plane can lie outside of the lens, hence those distances are not
necessarily equal.

The laser diode helps to align the lenses downstream. It shoots a small
diameter laser beam through the downstream optics of the microscope to
the camera port, onto which I assume you will add the 4f. It is crucial that
the mount of the diode has some adjustability to make sure that the beam
runs perfectly centered and parallel to the optical axis. This can be
achieved with six fine screws that hold the diode in a cylinder, which then
has the appropriate threading for the nosepiece at one end.

 If the beam is sufficiently small in diameter, it will not change much in
divergence when it passes the tubelens and other components (i.e. its
divergence is already larger than the curvature it "sees" when passing a
lens). The lenses of the 4f system are ligned up such that their individual
backreflections are concentric to the incoming beam ( a card with a small
hole can help to find all backreflections of one element). This helps to find
the correct lateral position AND rotation of each lens.

Please also remind your student of the correct orientation of an achromatic
doublet: the flat side should face the side on which light is brought to a
focus, otherwise additional aberrations will occur.

Following these steps has allowed me in the past to maintain diffraction
limited performance when relaying an image. Aligning optics can take an
amazingly long time, but it is most of the time worth it.

Best,
Reto

[Kyle Michael Douglass]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Thanks Reto and to those who responded offline to my question.

I have a follow-up question:
Reto's suggestion involves using coherent light from a laser diode to do the alignment; are there any ways or general tips for doing the alignment with incoherent or fluorescent light beyond that mentioned in my original post? Should I make the suggestion that microscope and accessory alignment should always be done with coherent light?

Thanks in advance!
Kyle

P.S. A 4f system is a system of two identical lenses for doing optical Fourier processing on an image. The input image plane lies at one lens focal length behind the first lens, the lenses are separated by two focal lengths, and the output image is at one focal length in front of the second lens; hence, the total length of the system is four focal lengths, or '4f.'
________________________________________
From: Confocal Microscopy List [[hidden email]] on behalf of Reto Fiolka [[hidden email]]
Sent: Saturday, January 03, 2015 11:39 PM
To: [hidden email]
Subject: Re: 4f system alignment with fluorescent light

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Hi Kyle

For such alignments, I recommend the use of the following tools and
procedure:

1) a shear plate

2) a laser diode in a mount that can be screwed into the nosepiece in place
of the objective.

Procedure:
With the shear plate, I make sure that a collimated beam that goes into the
4f is still collimated when it exits the two lenses. Further I determine the
position, relative to the lens housing, of the focal planes (for collimated
input). This helps to find the relative distances between the lenses and the
distance from the first lens to the intermediate image plane, as well as the
distance to the camera for the second lens. It is important to know that the
principal plane can lie outside of the lens, hence those distances are not
necessarily equal.

The laser diode helps to align the lenses downstream. It shoots a small
diameter laser beam through the downstream optics of the microscope to
the camera port, onto which I assume you will add the 4f. It is crucial that
the mount of the diode has some adjustability to make sure that the beam
runs perfectly centered and parallel to the optical axis. This can be
achieved with six fine screws that hold the diode in a cylinder, which then
has the appropriate threading for the nosepiece at one end.

 If the beam is sufficiently small in diameter, it will not change much in
divergence when it passes the tubelens and other components (i.e. its
divergence is already larger than the curvature it "sees" when passing a
lens). The lenses of the 4f system are ligned up such that their individual
backreflections are concentric to the incoming beam ( a card with a small
hole can help to find all backreflections of one element). This helps to find
the correct lateral position AND rotation of each lens.

Please also remind your student of the correct orientation of an achromatic
doublet: the flat side should face the side on which light is brought to a
focus, otherwise additional aberrations will occur.

Following these steps has allowed me in the past to maintain diffraction
limited performance when relaying an image. Aligning optics can take an
amazingly long time, but it is most of the time worth it.

Best,
Reto

[Reto Fiolka]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Hi Kyle

I highly recommend the use of coherent light, as you can get a lot of light
into a single optical mode (here a collimated beam).

One lens has 5 degrees of freedom (three translation and two rotation)
that need to be aligned. The only method that I know is tracing the back-
reflections from a small collimated laser beam to get them perfect.

An achromatic doublet generates three reflections from each surface.
Rotating the lens will move all three together, translating the lens will
move them at different rates. The backreflections are weak, as the surfaces
are normally AR coated, but still visible by the naked eye.

I see no way how to do a proper alignement with incoherent light. As an
example, you can steer the beam by rotating or translating the lens. Thus
just observing if the beam stays on the optical axis will not tell you if the
lens is correctly aligned. Further with incoherent light you can not perform
the beam collimation test, which I find very helpful ( e.g. collimated in all
image planes and focused in all Fourier planes). Thus I also highly
advocate buying a shear plate and in addition, a small Webcam for beam
inspection.

Good luck with the optical Fourier transform project!

Best,
Reto

[John Oreopoulos]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

A similar question was raised about half a year ago:

http://lists.umn.edu/cgi-bin/wa?A2=ind1405&L=confocalmicroscopy&D=0&P=14281

In addition to the comprehensive answer that Reto has already given, you can find further details with diagrams in this book chapter by Rainer Heintzmann:

http://onlinelibrary.wiley.com/doi/10.1002/9783527671595.app1/summary

Cheers,


John Oreopoulos
Staff Scientist
Spectral Applied Research Inc.
A Division of Andor Technology
Richmond Hill, Ontario
Canada
www.spectral.ca



On 2015-01-05, at 11:03 AM, Reto Fiolka wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> Post images on http://www.imgur.com and include the link in your posting.
> *****
>
> Hi Kyle
>
> I highly recommend the use of coherent light, as you can get a lot of light
> into a single optical mode (here a collimated beam).
>
> One lens has 5 degrees of freedom (three translation and two rotation)
> that need to be aligned. The only method that I know is tracing the back-
> reflections from a small collimated laser beam to get them perfect.
>
> An achromatic doublet generates three reflections from each surface.
> Rotating the lens will move all three together, translating the lens will
> move them at different rates. The backreflections are weak, as the surfaces
> are normally AR coated, but still visible by the naked eye.
>
> I see no way how to do a proper alignement with incoherent light. As an
> example, you can steer the beam by rotating or translating the lens. Thus
> just observing if the beam stays on the optical axis will not tell you if the
> lens is correctly aligned. Further with incoherent light you can not perform
> the beam collimation test, which I find very helpful ( e.g. collimated in all
> image planes and focused in all Fourier planes). Thus I also highly
> advocate buying a shear plate and in addition, a small Webcam for beam
> inspection.
>
> Good luck with the optical Fourier transform project!
>
> Best,
> Reto