Nikon Diaphot TMD w/ DIC -- Help!

[René]

Hi, René -

Well, higher magnifications are OK on the .55 NA DIC condenser for the Diaphot, but the results are inevitably limited and there isn't much point in using the scope as configured for that, to my way of thinking. If nothing else, using high NA objectives on the Diaphot requires super thick immersion oil, and it's a pain to clean given the configuration of the scope.


In terms of LED fluorescence light sources, I really don't think we are there yet - at least not at any reasonable price point. I haven't looked into it in a few years. If you can point me to an LED UV source that has been shown to have a broad enough spectrum to be useful with even a small range of standard fluorochromes (TRITC, FITC, DAPI equivalents at a minimum), I would love to know about it!
Leonard

Hi Leonard, just try it, you might be surprised. Remember, the ubiquitous Abbe yields an aplanatic cone of just 0.65, and is still mounted on many a microscope, and used with 100x objectives; a 60x is more forgiving in terms of image quality. Our quality mark (iso 17025) was resolving Frustulia saxonica on a testslide in brightfield, no problem.


Concerning led fluorescence, I've limited the use to a UV led for DAPI and Calcofluor white (for determination of dinoflagellates). A Nichia UV LED (385nm works well) is not that expensive anymore. The thing with the Oly IMT-2 and your Diaphot is that you have plenty of room in front of the cube. Critcally was the use of a small plastic fresnel lens in front of the excitation filter, and the LED mounted at focal distance (a cm or so) in front of the cube (fresnel). No collimators or lamphouses necessary. The LED was dimmed to a great extent, so a large heat sink is not necessary. Got the fresnels for a tenner each or so from https://www.fresneltech.com, just remember to take the UVT acrylic ones for UV transmittance.
Biggest expense was the cubes. I'm sure you can think of a way to mount different LEDs in a row on a slider , if you want flexibilty in excitation wavelengths.

Good luck, René

[Pau]

I suspect that this arrangement with separate sliders is probably an afterthought because it adds 21.5mm in height to the base of the objective turret, as noted above. In effect, it is converting a microscope of 160mm mechanical tube length to a 181.5mm mechanical tube length scope. There is no corrective lens in the accessory piece for the prism slider. I see no indication of a corrective lens coupled with the prism, either. The eyepieces are standard Nikon CFWN 10x/20s. The stage accommodates the additional turret height by the simple expedient of being clamped 20mm or so higher than normal.

It may be that this tube length increase is not very important, especially at magnifications of maybe 20x and up? The well-known Leitz memo from the mid 1970s regarding their shift from 170mm TL to 160mm TL objectives suggests that the difference in objective performance when using a 160TL objective on a 170TL scope is minimal with objective powers of 16x and above. And it notes that there is +/- 10mm of play in tube length on any binocular scope having an adjustable inter-ocular distance for the eyepieces.

Here the difference in the position of the intermediate image created by the tube length discrepancy is probably more because the objective focal length discrepancy is greater (181/160 instead of 170/160). I certainly do not know enough to say how much more the difference is. At a total guess, I am wondering whether magnifications of 20x and below might suffer from the increased tube length, so I may do some comparative tests.

Leonard,
21.5mm is quite a change in the tube length, it will affect the image with high NA objectives (and of course change the magnification) but the easier way to spot it is the loss of parfocality between the objectives. If you use the original objectives and parfocality is good you can be sure that some kind of TL compensating system is present, likely a very weak negative lens or a different lens than in the original model and this could not be evident.

Actually the difference of optical tube length between Leitz 170 and 160 objectives is only 2mm, the other 8mm are just mechanical: the eyepieces for 170 microscopes protrude 8mm deeper into the tube, 18mm vs the 10mm DIN standard. When mounted together in the same stand the parfocality mismatch is very small (barely noticeable) and the spherical aberration issue just theoretical.

[Pau]

LEDs as fluorescence excitation source is a real alternative both in expensive off the self devices and DIY systems

It's what many amateurs use, including myself
https://www.photomacrography.net/forum/ ... 937#201937
(now improved with some modifications)
The main idea is simple: put the LED at the same position of the HBO lamp arc and use the same collimating optics

[Lou Jost]

The main idea is simple: put the LED at the same position of the HBO lamp arc and use the same collimating optics

For those who are trying to do this with an ordinary epi-illuminator rather than one designed for fluorescence, be aware that the collimating optics are thick and will greatly reduce UV transmission. I had to remove the collector lens and replace it with a plastic fresnel lens that I cut out of an old flash.

[ldflan]

"...the easier way to spot it is the loss of parfocality between the objectives."

This is very good to know! Thank you! I will look for that.

"...some kind of TL compensating system [may be] present, likely a very weak negative lens or a different lens than in the original model and this could not be evident."

The scope itself is not modified. The supplementary DIC slider attachment has no lens in it. I suppose the correction lens in the turret might be different, but it looks like the normal model.

"Actually the difference of optical tube length between Leitz 170 and 160 objectives is only 2mm, the other 8mm are just mechanical: the eyepieces for 170 microscopes protrude 8mm deeper into the tube, 18mm vs the 10mm DIN standard. When mounted together in the same stand the parfocality mismatch is very small (barely noticeable) and the spherical aberration issue just theoretical."

If I understand correctly, that's actually a much better explanation than the original memo, which discusses rather confusingly the 2mm "displacement of the intermediary image..." It will be interesting to see the impact the change in height has. The whole setup is designed for the LWD DIC objectives, so I will try with those. Need to get a camera mount attached to the front of this scope first. I am not sure why so many of these have the F mount missing when offered for sale - it does not appear to be a case of damage...

Leonard

[ldflan]

LEDs as fluorescence excitation source is a real alternative both in expensive off the self devices and DIY systems

It's what many amateurs use, including myself
https://www.photomacrography.net/forum/ ... 937#201937
(now improved with some modifications)
The main idea is simple: put the LED at the same position of the HBO lamp arc and use the same collimating optics

Impressive!! I especially like the slider / heat sink arrangement. I have a bag full of LEDs of various output wavelengths from a few years ago. Perhaps I will renew my attempts...

If I were working at home, there is no question I would go with LED as my epifluorecence light source for reasons of safety alone. I'm fortunate to have space where I can use an HBO burner in relative safety.

The issue I noted in my post was the narrow output spectrum of any given LED. I haven't seen a good solution. Are there any LED-based systems, DIY or otherwise, that enable dual or triple wavelength excitation (e.g. combined FITC/TRITC or FITC/TRITC/DAPI)? Perhaps some kind of light combining box? It would suck light, but one could perhaps use two or three LEDs of each wavelength... Or some kind of off-axis collimating system. Parabolic mirror? Fresnel? If it were easy, people smarter than me would probably have done it by now, I figure!

Leonard

[ldflan]

Critcally was the use of a small plastic fresnel lens in front of the excitation filter, and the LED mounted at focal distance (a cm or so) in front of the cube (fresnel). No collimators or lamphouses necessary.

René -

Thanks. Pretty sure I can visualize what you are describing. There should be room for a sliding bar similar to what Pau did under the turret. I'll take a look later today.

Leonard

[Lou Jost]

Idflan, for multiband fluorescence, the excitation of the all the desired bands needs to be simultaneous, so you'd need multiple LEDs in very close proximity, or an optical solution for combining more widely separated lamps into a single beam, perhaps with an offset reflecting parabola fed by several different beams of different colors.

[ldflan]

Idflan, for multiband fluorescence, the excitation of the all the desired bands needs to be simultaneous, so you'd need multiple LEDs in very close proximity, or an optical solution for combining more widely separated lamps into a single beam, perhaps with an offset reflecting parabola fed by several different beams of different colors.

Right... say three off-axis collimating mirrors (parabolic sections) and three different LEDs - with matching focal lengths at their particular wavelengths, yes? Alignment of the mirrors or LEDs, or both, would have to be adjustable. Would it not all also have to be packed in a very tight space?

Leonard

[Charles Krebs]

Late to the party here but a few items...

I seriously doubt that Nikon just ignored the increase in distance, My TMD came set up for DIC and fluorescence (I don't use it) so I have that ~20mm increase. I would be curious for those who do not have this to look at their nosepieces. My nosepiece (and I suspect all TMD nosepieces do, but maybe not) has an optic built into the "active" aperture. If only the fluorescence set-up scopes have this, then it might correct for the length increase. If all have this optic then I suspect Nikon has made this scope (with optics) to essentially convert to an infinity type system of some sort. If you just look at the distance between the objective shoulder and the intermediate image presented in the camera port or the eyepiece tube it is far more than 150-180mm. It is certain that there are optics providing the needed corrections.

I've also found that on a full frame camera, the front photoport, with the built-in 2.5X, gives me darkened corners. If I use the side port with a Nikon PL 2.5X I get much more even exposures across the frame into the corners.

[ldflan]

... My nosepiece (and I suspect all TMD nosepieces do, but maybe not) has an optic built into the "active" aperture. If only the fluorescence set-up scopes have this, then it might correct for the length increase. If all have this optic then I suspect Nikon has made this scope (with optics) to essentially convert to an infinity type system of some sort.

Charles -

Yes, all the turrets have that corrective lens in the active aperture. Your suspicion is an interesting one. Like you I can't imagine Nikon ignoring the problem, but I don't have enough optics knowledge to opine meaningfully on your suggestion. I haven't directly compared the results from the two setups yet, but a quick look at images coming from this new (to me) system with the separate prism and analyzer all looks very good using the usual LWD 40x/0.55. I would be surprised if the combined prism and analyzer system does appreciably better with that objective. Will report back later on results using higher NA objectives once I get a camera mount on the front of this rig.

Leonard