tube lens position on Nikon Ti2

*****
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.
*****

We are tinkering with a microscope build based on a Nikon Ti2.

For the purposes of beam alignment, we sent a collimated beam through the side port and were surprised to see that we are unable to collimate the output after the 60x water immersion.  Someone suggested that maybe Nikon does this on purpose to position the lens so that it is halfway between ideal for the standard one turret option and the stage up kit that adds a second turret. We tested this idea by raising the objective on a tube by about 1.5" (3.75cm) and we got an approximately collimated beam out.  It seems to have been a good tip (thanks Hazen).

I believe this means that in the typical one-turret configuration that we typically use,  the 20cm tube lens is perhaps ~16cm from the back focal plane of the objective lens.  Further, this slightly non-telecentric arrangement should lead to defocused images having a somewhat different magnification from the infocus image. We may be able to live with this for our application, but wanted to first entertain the idea of fixing the arrangement.

We are pondering options.

1) Raise the lens and stage and condenser. Then all the objectives would need to be on ~4cm risers. Not sure if perfect focus would still work well but easy to test. Anyway, this would be annoying.

2) Replace the tube lens with a tube lens positioned halfway between objective back focal plane and side port image plane. I'm guessing this would in principle mean an 18cm tube lens positioned slightly in a different place from the current tube lens. I guess this may be possible but going into the microscope guts is a little frightening.

3) Try to build around this design compromise we would rather not have. We will probably do this option but I wanted to see what people think about the options.


Thanks!
Josh

[Zdenek Svindrych-2]

*****
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 Josh,

I've seen this with other microscopes, too, e.g. Zeiss Axio Scope.A1.

But it's vital to realize the importance of the aperture stop in
telecentricity. Since most common objectives are designed so that the
aperture stop coincides with the back focal plane of the objective lens,
the system is object-side telecentric, and the apparent size of your sample
does not change with defocus. Of course, if the tube lens is closer than it
should be, the image-side is not telecentric and the image expands as you
move the camera further away.

I'm afraid this is quite common practice that helps save space and glass...
especially on the eypiece / camera ports (you don't want this with TIRF or
confocal, but those setups usually use separate tube lenses).

And don't forget that the 10 mm of focus stroke limits the precision with
which you can make the system doubly-telecentric!

Best, zdenek


On Thu, Jan 24, 2019 at 9:51 PM Joshua Vaughan <[hidden email]> 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.
> *****
>
> We are tinkering with a microscope build based on a Nikon Ti2.
>
> For the purposes of beam alignment, we sent a collimated beam through the
> side port and were surprised to see that we are unable to collimate the
> output after the 60x water immersion.  Someone suggested that maybe Nikon
> does this on purpose to position the lens so that it is halfway between
> ideal for the standard one turret option and the stage up kit that adds a
> second turret. We tested this idea by raising the objective on a tube by
> about 1.5" (3.75cm) and we got an approximately collimated beam out.  It
> seems to have been a good tip (thanks Hazen).
>
> I believe this means that in the typical one-turret configuration that we
> typically use,  the 20cm tube lens is perhaps ~16cm from the back focal
> plane of the objective lens.  Further, this slightly non-telecentric
> arrangement should lead to defocused images having a somewhat different
> magnification from the infocus image. We may be able to live with this for
> our application, but wanted to first entertain the idea of fixing the
> arrangement.
>
> We are pondering options.
>
> 1) Raise the lens and stage and condenser. Then all the objectives would
> need to be on ~4cm risers. Not sure if perfect focus would still work well
> but easy to test. Anyway, this would be annoying.
>
> 2) Replace the tube lens with a tube lens positioned halfway between
> objective back focal plane and side port image plane. I'm guessing this
> would in principle mean an 18cm tube lens positioned slightly in a
> different place from the current tube lens. I guess this may be possible
> but going into the microscope guts is a little frightening.
>
> 3) Try to build around this design compromise we would rather not have. We
> will probably do this option but I wanted to see what people think about
> the options.
>
>
> Thanks!
> Josh
>

--
--
Zdenek Svindrych, Ph.D.
Research Associate - Imaging Specialist
Department of Biochemistry and Cell Biology
Geisel School of Medicine at Dartmouth

[Andriy Chmyrov]

*****
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 Josh and Zdenek,

I can confirm that this is the case for Olympus IX83 as well.
I just defocused the light on the way to the stand in order to get a collimated output from the objective.

Funny thing is that if you move the objective, for example when you need to look further away from a coverslip, that degree of defocusing need to be changed, or you will get non-collimated output. Luckily for objectives with WD of 180 um this is not a dig deal, but it needs to be considered if you change your sample plane by raising a stage or having some kind of sample spacer...

Regards,
Andriy

[Lu Yan]

*****
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.
*****

I thought all those 4 brand microscopes are not actually telecentric or 4-f systems with their build-in tube lenses, so you cannot really expect to have a collimated beam output with a collimated input (?).


In fact, if you were to use their objective lenses and tube lenses in customized systems, they usually specify the upper limit for the the infinity space length beyond which the image quality, e.g. all aberration correction etc., are not guaranteed.


Maybe a simple solution to get a collimated beam is to insert the Bertrand lens into the beam path if  there is one that is meant to be used to image the backaperture of the obj onto the the side port camera. But usually Bertrand lens is in a separate beam path.





________________________________
From: Confocal Microscopy List <[hidden email]> on behalf of Andriy Chmyrov <[hidden email]>
Sent: Monday, January 28, 2019 8:19:08 AM
To: [hidden email]
Subject: Re: tube lens position on Nikon Ti2

*****
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 Josh and Zdenek,

I can confirm that this is the case for Olympus IX83 as well.
I just defocused the light on the way to the stand in order to get a collimated output from the objective.

Funny thing is that if you move the objective, for example when you need to look further away from a coverslip, that degree of defocusing need to be changed, or you will get non-collimated output. Luckily for objectives with WD of 180 um this is not a dig deal, but it needs to be considered if you change your sample plane by raising a stage or having some kind of sample spacer...

Regards,
Andriy

*****
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 to folks for responses. I have a few follow up points.

Zdenek mentioned objective-side telecentric or detection-side telecentric
and I'm not sure how that works. I can follow the rays for simple thin lens
systems regarding telecentric or not, but I'm not sure how to think about
it for objective lenses. Any suggestions for a reference or text that may
help explain this? I think this is remedial reading I should have done a
long time ago.

Zdenek and Andriy both mentioned, I think, that telecentricity will be
perturbed if the objective lens position moves from being perfectly
positioned and I think this is also right. Without any adjustment, the
Nikon Ti2 objective lens is is ~4cm too close to the tube lens. If we
repositioned the lens 4cm away but had to tune away from the position by
1-2 mm, the perturbations should be much smaller than we currently see. In
principle, one could position the objective "perfectly" and then move the
sample rather than the lens, though it might be a little tricky. I think
our ASI stage has a 500um piezo on it, but I just don't think the 1-2 mm
will matter much.

Regarding Lu's second point, I'd be interested to learn where those
specifications are listed. From what we can tell, the Ti2  (like the
earlier TiE/TiU/TiS) system can come with a second turret as part of the
stage-up kit. In that case the objective lens is raised perhaps 8 cm or so.
It seems that the default objective lens position is designed to fit about
halfway between perfect for the one-turret and the two-turret. I would
guess that the operation is good between both of those distances, including
at the 'perfect' position from the 20cm tube lens.

When we look through the window on the Ti2 it seems changing the tube lens
is a bit of a mess...for instance installing a 18cm tube lens ~2cm further
away from the objective lens than it is at present (with no stage up kit).
We tried raising the lens ~4cm and the stage, etc. Collimation is good, and
the perfect-focus system works still. We might end up just doing that, at
least for now. We are also looking at putting in something like the stage
up kit to raise objective and sample stage by 4cm, but we still can't tell
how hard that would be.


On Mon, Jan 28, 2019 at 8:09 PM Yan, Lu <[hidden email]> 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.
> *****
>
> I thought all those 4 brand microscopes are not actually telecentric or
> 4-f systems with their build-in tube lenses, so you cannot really expect to
> have a collimated beam output with a collimated input (?).
>
>
> In fact, if you were to use their objective lenses and tube lenses in
> customized systems, they usually specify the upper limit for the the
> infinity space length beyond which the image quality, e.g. all aberration
> correction etc., are not guaranteed.
>
>
> Maybe a simple solution to get a collimated beam is to insert the Bertrand
> lens into the beam path if  there is one that is meant to be used to image
> the backaperture of the obj onto the the side port camera. But usually
> Bertrand lens is in a separate beam path.
>
>
>
>
>
> ________________________________
> From: Confocal Microscopy List <[hidden email]> on
> behalf of Andriy Chmyrov <[hidden email]>
> Sent: Monday, January 28, 2019 8:19:08 AM
> To: [hidden email]
> Subject: Re: tube lens position on Nikon Ti2
>
> *****
> 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 Josh and Zdenek,
>
> I can confirm that this is the case for Olympus IX83 as well.
> I just defocused the light on the way to the stand in order to get a
> collimated output from the objective.
>
> Funny thing is that if you move the objective, for example when you need
> to look further away from a coverslip, that degree of defocusing need to be
> changed, or you will get non-collimated output. Luckily for objectives with
> WD of 180 um this is not a dig deal, but it needs to be considered if you
> change your sample plane by raising a stage or having some kind of sample
> spacer...
>
> Regards,
> Andriy
>

--
Joshua C. Vaughan
Assistant Professor
Department of Chemistry
Box 351700
University of Washington
Seattle, WA 98195
206-543-4644

[James Kerin]

*****
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.
*****

&nbsp;
*** COMMERCIAL POSTING ***



This is a most interesting issue, which is very familiar to us at Cairn, so we are very pleased to give some detailed answers.&nbsp; More, in fact, than you may ever wish to know!



First of all, microscopes generally aren't designed to be telecentric.&nbsp; For them to be so, the distance between the rear of the objective and the tube lens needs to be at least the focal length of the tube lens, which is 200mm for Nikon.


The problem here is that the optical beam diameter tends to increase through this "infinity distance".&nbsp; The primary purpose of this region isn't to make the microscope telecentric, but rather to allow various accessories such as epi-illuminators to be incoroporated into the optical pathway without introducing any optical aberrations (specifically, any tilted component, such as a partially reflective mirror, in a FOCUSSING pathway, will introduce astigmatism, whereas it being in an infinity region prevents that).&nbsp; For that, around a 100mm infinity path length should be sufficient, and that sort of distance is generally chosen in order for the beam diameter through this region not to increase too much.&nbsp; However, "double deck" microscopes will require about double that distance, which coincidentally (in our opinion) tends to make them more or less telecentric.&nbsp; That possibiliy requires a larger aperture tube lens, which we believe is one of the improvements made in
the Ti2, but even so we believe the maximum field diameter to be somewhat reduced compared with the single-deck version, so there is a tradeoff here.


Another factor to consider is that the Nikon tube lens is a multielement design, so although one can can model such a design in terms of a single lens at some position, that position isn't going to correspond to any of the individual elements.&nbsp; In practice it's generally (but not always if there are negative elements!) going to be somewhere in between them.&nbsp;



The problem with a nontelecentric system is indeed that the magnification will partly depend on the focus, although that may not always matter of course.&nbsp; However, we are very familiar with this issue at Cairn, as one of the supported applications for our image splitters and multicamera adapters is that of simultaneous imaging of samples at multiple depths, for which one ideally requires changes in focus WITHOUT changing the magnification. A technical paper on our website&nbsp; (https://www.cairn-research.co.uk/support/technical-papers/) describes how a subsequent optical relay can be configured to make the configuration telecentric OVERALL, without having to change the microscope in any way, which is clearly a much better solution.&nbsp; And which we just happen to support!&nbsp; But whether or not you want to get into the theory of it all, we also have a spreadsheet on our latest news item that will calculate everything for you so in practice things can be much simpler than
you may think!&nbsp; News item -&nbsp;https://www.cairn-research.co.uk/2019/01/29/the-multitalented-multisplit-goes-into-multidepth-mode/


Martin Thomas, Cairn Research Ltd
&nbsp;
Dr Martin V Thomas
Chairman

&nbsp;
&nbsp;

&nbsp;
 
&nbsp;
 
&nbsp;

&nbsp;
&nbsp;
&nbsp;
&nbsp;

&nbsp;




On  01/24/2019, 02:44am, Joshua Vaughan ([hidden email]) 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.
 
*****
 

 
We are tinkering with a microscope build based on a Nikon Ti2.
 

 
For the purposes of beam alignment, we sent a collimated beam through the side port and were surprised to see that we are unable to collimate the output after the 60x water immersion.  Someone suggested that maybe Nikon does this on purpose to position the lens so that it is halfway between ideal for the standard one turret option and the stage up kit that adds a second turret. We tested this idea by raising the objective on a tube by about 1.5" (3.75cm) and we got an approximately collimated beam out.  It seems to have been a good tip (thanks Hazen).
 

 
I believe this means that in the typical one-turret configuration that we typically use,  the 20cm tube lens is perhaps ~16cm from the back focal plane of the objective lens.  Further, this slightly non-telecentric arrangement should lead to defocused images having a somewhat different magnification from the infocus image. We may be able to live with this for our application, but wanted to first entertain the idea of fixing the arrangement.
 

 
We are pondering options.
 

 
1) Raise the lens and stage and condenser. Then all the objectives would need to be on ~4cm risers. Not sure if perfect focus would still work well but easy to test. Anyway, this would be annoying.
 

 
2) Replace the tube lens with a tube lens positioned halfway between objective back focal plane and side port image plane. I'm guessing this would in principle mean an 18cm tube lens positioned slightly in a different place from the current tube lens. I guess this may be possible but going into the microscope guts is a little frightening.
 

 
3) Try to build around this design compromise we would rather not have. We will probably do this option but I wanted to see what people think about the options.
 

 

 
Thanks!
 
Josh

[Craig Brideau]

*****
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.
*****

On Mon, Jan 28, 2019 at 10:53 PM Joshua Vaughan <[hidden email]> 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.
> In principle, one could position the objective "perfectly" and then move
> the
> sample rather than the lens, though it might be a little tricky. I think
> our ASI stage has a 500um piezo on it, but I just don't think the 1-2 mm
> will matter much.
>

I actually did this for my CARS microscope. Overlapping the two lasers was
picky, so I didn't want my objective lens to move. I have three really
large stepper motors moving my entire sample stage in XYZ. They have 50nm
step size, so I even use them for z-stacks. My system scans only 1 fps, so
any 'bounce' from the steppers doesn't seem to cause any issues with
vibration in my case. I think a system like this for coarse focusing and
then a small piezo for high-speed stacks would allow this to work on a
video-rate system.

Craig

[Zdenek Svindrych-2]

*****
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 Joshua,

see here (https://drive.google.com/open?id=12tZrXiMXrK6tmLVr6zo1-OdPrN6gzRyS)
simple paraxial ray-traces hat illustrate the telecentric concept. Hope it
clarifies the matter a bit (and yes, a stop is a simple yet important
optical element) :-).

As Martin (aka James?) mentioned, the tube lens can be smaller if it's
closer to the objective lens (my gut feeling is that the tube lens diameter
in the 4-f position has to be the diameter of the image filed /also called
the Field Number/ plus the max supported BFP stop diameter; but when it's
touching the objective lens it can be as small as the BFP stop alone).
Otherwise you'll get vignetting.

Anyway, if one bad lens can mess up the telecentric arrangement, another
lens should be able to fix it (look at my drawings mentioned above). The
question that remains is: Why do you need perfect 4-f system? And how far
are you willing to go? Most confocals / TIRFs can deal with the situation
in one way or another...

Btw, the way I solved this problem (in one particular case of a
point-scanning system) was the back port - I removed all the fluorescence
illumination stuff and put the (manufacturer-recommended) tube lens where I
needed it (at the possible cost of reduced field of view, but the design
was for 18 mm Field Number anyway).

Another 'btw', multi-element tube lens (Nikon, Olympus) is a good thing, it
suggests that the objective itself is well corrected. With Zeiss, part of
the chromatic correction is done with the (single element) tube lens, so
you can only use this particular tube lens with their objective lenses.

Still another 'btw', way back when I've created other raytraces that may be
helpful, see here:
https://drive.google.com/open?id=1a4RU1MAo3CeHG6DHw73Wx06ipkbB4kGn

Best, zdenek

On Tue, Jan 29, 2019 at 12:53 AM Joshua Vaughan <[hidden email]> 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.
> *****
>
> Thanks to folks for responses. I have a few follow up points.
>
> Zdenek mentioned objective-side telecentric or detection-side telecentric
> and I'm not sure how that works. I can follow the rays for simple thin lens
> systems regarding telecentric or not, but I'm not sure how to think about
> it for objective lenses. Any suggestions for a reference or text that may
> help explain this? I think this is remedial reading I should have done a
> long time ago.
>
> Zdenek and Andriy both mentioned, I think, that telecentricity will be
> perturbed if the objective lens position moves from being perfectly
> positioned and I think this is also right. Without any adjustment, the
> Nikon Ti2 objective lens is is ~4cm too close to the tube lens. If we
> repositioned the lens 4cm away but had to tune away from the position by
> 1-2 mm, the perturbations should be much smaller than we currently see. In
> principle, one could position the objective "perfectly" and then move the
> sample rather than the lens, though it might be a little tricky. I think
> our ASI stage has a 500um piezo on it, but I just don't think the 1-2 mm
> will matter much.
>
> Regarding Lu's second point, I'd be interested to learn where those
> specifications are listed. From what we can tell, the Ti2  (like the
> earlier TiE/TiU/TiS) system can come with a second turret as part of the
> stage-up kit. In that case the objective lens is raised perhaps 8 cm or so.
> It seems that the default objective lens position is designed to fit about
> halfway between perfect for the one-turret and the two-turret. I would
> guess that the operation is good between both of those distances, including
> at the 'perfect' position from the 20cm tube lens.
>
> When we look through the window on the Ti2 it seems changing the tube lens
> is a bit of a mess...for instance installing a 18cm tube lens ~2cm further
> away from the objective lens than it is at present (with no stage up kit).
> We tried raising the lens ~4cm and the stage, etc. Collimation is good, and
> the perfect-focus system works still. We might end up just doing that, at
> least for now. We are also looking at putting in something like the stage
> up kit to raise objective and sample stage by 4cm, but we still can't tell
> how hard that would be.
>
>
> On Mon, Jan 28, 2019 at 8:09 PM Yan, Lu <[hidden email]> 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.
> > *****
> >
> > I thought all those 4 brand microscopes are not actually telecentric or
> > 4-f systems with their build-in tube lenses, so you cannot really expect
> to
> > have a collimated beam output with a collimated input (?).
> >
> >
> > In fact, if you were to use their objective lenses and tube lenses in
> > customized systems, they usually specify the upper limit for the the
> > infinity space length beyond which the image quality, e.g. all aberration
> > correction etc., are not guaranteed.
> >
> >
> > Maybe a simple solution to get a collimated beam is to insert the
> Bertrand
> > lens into the beam path if  there is one that is meant to be used to
> image
> > the backaperture of the obj onto the the side port camera. But usually
> > Bertrand lens is in a separate beam path.
> >
> >
> >
> >
> >
> > ________________________________
> > From: Confocal Microscopy List <[hidden email]> on
> > behalf of Andriy Chmyrov <[hidden email]>
> > Sent: Monday, January 28, 2019 8:19:08 AM
> > To: [hidden email]
> > Subject: Re: tube lens position on Nikon Ti2
> >
> > *****
> > 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 Josh and Zdenek,
> >
> > I can confirm that this is the case for Olympus IX83 as well.
> > I just defocused the light on the way to the stand in order to get a
> > collimated output from the objective.
> >
> > Funny thing is that if you move the objective, for example when you need
> > to look further away from a coverslip, that degree of defocusing need to
> be
> > changed, or you will get non-collimated output. Luckily for objectives
> with
> > WD of 180 um this is not a dig deal, but it needs to be considered if you
> > change your sample plane by raising a stage or having some kind of sample
> > spacer...
> >
> > Regards,
> > Andriy
> >
>
>
> --
> Joshua C. Vaughan
> Assistant Professor
> Department of Chemistry
> Box 351700
> University of Washington
> Seattle, WA 98195
> 206-543-4644
>

--
--
Zdenek Svindrych, Ph.D.
Research Associate - Imaging Specialist
Department of Biochemistry and Cell Biology
Geisel School of Medicine at Dartmouth

[Lu Yan]

*****
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 Joshua,


I cannot find the recommended tube length on their website, but I have impression that I got this from engineering rep from several manufacturers, especially those who also do correction in tube lens.


Regarding the stage up kit/lens, I thought the stage up lens was for epi illlumination application through the backport of the lower tier. because the objective lens is lift up, you will need a lens to still focus the illumination beam through the backport, e.g. could be regular epi widefield, or TIRF, to the back focal plane of the objective lens. Note that the position of this plane also varies for different objective lens. For example, I think for a Nikon 60X/1.45 lens, it is a few mm inside of the lens.


If you use the upper tier turret/back port for illumination, I believe you would not need the stage up lens, even now your infinity space is longer.


Lu

________________________________
From: Confocal Microscopy List <[hidden email]> on behalf of Joshua Vaughan <[hidden email]>
Sent: Tuesday, January 29, 2019 12:42:01 AM
To: [hidden email]
Subject: Re: tube lens position on Nikon Ti2

*****
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 to folks for responses. I have a few follow up points.

Zdenek mentioned objective-side telecentric or detection-side telecentric
and I'm not sure how that works. I can follow the rays for simple thin lens
systems regarding telecentric or not, but I'm not sure how to think about
it for objective lenses. Any suggestions for a reference or text that may
help explain this? I think this is remedial reading I should have done a
long time ago.

Zdenek and Andriy both mentioned, I think, that telecentricity will be
perturbed if the objective lens position moves from being perfectly
positioned and I think this is also right. Without any adjustment, the
Nikon Ti2 objective lens is is ~4cm too close to the tube lens. If we
repositioned the lens 4cm away but had to tune away from the position by
1-2 mm, the perturbations should be much smaller than we currently see. In
principle, one could position the objective "perfectly" and then move the
sample rather than the lens, though it might be a little tricky. I think
our ASI stage has a 500um piezo on it, but I just don't think the 1-2 mm
will matter much.

Regarding Lu's second point, I'd be interested to learn where those
specifications are listed. From what we can tell, the Ti2  (like the
earlier TiE/TiU/TiS) system can come with a second turret as part of the
stage-up kit. In that case the objective lens is raised perhaps 8 cm or so.
It seems that the default objective lens position is designed to fit about
halfway between perfect for the one-turret and the two-turret. I would
guess that the operation is good between both of those distances, including
at the 'perfect' position from the 20cm tube lens.

When we look through the window on the Ti2 it seems changing the tube lens
is a bit of a mess...for instance installing a 18cm tube lens ~2cm further
away from the objective lens than it is at present (with no stage up kit).
We tried raising the lens ~4cm and the stage, etc. Collimation is good, and
the perfect-focus system works still. We might end up just doing that, at
least for now. We are also looking at putting in something like the stage
up kit to raise objective and sample stage by 4cm, but we still can't tell
how hard that would be.


On Mon, Jan 28, 2019 at 8:09 PM Yan, Lu <[hidden email]> 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.
> *****
>
> I thought all those 4 brand microscopes are not actually telecentric or
> 4-f systems with their build-in tube lenses, so you cannot really expect to
> have a collimated beam output with a collimated input (?).
>
>
> In fact, if you were to use their objective lenses and tube lenses in
> customized systems, they usually specify the upper limit for the the
> infinity space length beyond which the image quality, e.g. all aberration
> correction etc., are not guaranteed.
>
>
> Maybe a simple solution to get a collimated beam is to insert the Bertrand
> lens into the beam path if  there is one that is meant to be used to image
> the backaperture of the obj onto the the side port camera. But usually
> Bertrand lens is in a separate beam path.
>
>
>
>
>
> ________________________________
> From: Confocal Microscopy List <[hidden email]> on
> behalf of Andriy Chmyrov <[hidden email]>
> Sent: Monday, January 28, 2019 8:19:08 AM
> To: [hidden email]
> Subject: Re: tube lens position on Nikon Ti2
>
> *****
> 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 Josh and Zdenek,
>
> I can confirm that this is the case for Olympus IX83 as well.
> I just defocused the light on the way to the stand in order to get a
> collimated output from the objective.
>
> Funny thing is that if you move the objective, for example when you need
> to look further away from a coverslip, that degree of defocusing need to be
> changed, or you will get non-collimated output. Luckily for objectives with
> WD of 180 um this is not a dig deal, but it needs to be considered if you
> change your sample plane by raising a stage or having some kind of sample
> spacer...
>
> Regards,
> Andriy
>

--
Joshua C. Vaughan
Assistant Professor
Department of Chemistry
Box 351700
University of Washington
Seattle, WA 98195
206-543-4644

[Ian Dobbie]

*****
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.
*****

"Yan, Lu" <[hidden email]> writes:

> I thought all those 4 brand microscopes are not actually telecentric
> or 4-f systems with their build-in tube lenses, so you cannot really
> expect to have a collimated beam output with a collimated input (?).

One technique you really have to worry about this in is 3D-SIM. To get
nice, high contrast stripes, especially in Z, you need well focused
spots in your back focal plane and the easiest way to get this is have
collimated beams entering the tube lens. A true 4-f setup really helps
with this. It also means you cant move your objective, so sample based
focus all the way.

As to a later point about the location of back focal planes differing
with objective. This is definitely true. Typically low NA low mag
objective have BFP out the back, whereas high NA high mag objectives are
actually quite far in. Our Olympus 60x oil objective has its BFP 19.2mm
from the shoulder of the RMS mount, so roughly half way down the lens!

Ian