Experimenting with Dark Field and Polarized microscopy with Modularized condenser (no LED grid)

[ashish_kumar_4]

Disclaimer: I am new to microscopy and optics etc. Just making all this as hobby cuz I am really interested (since I was a child). Made everything on a resin printer so might or might not work directly on FDM

Hey! So I finally made my openflexure microscope v7 (beta1) with high res optics (though cheap 40x objective). I then learned about other kinds of microscopy including dark field, polarizing, phase contrast etc, so curious and wanting to try them out, I dug a little deep as how to upgrade my current scope for these. I am starting with Dark field and Polarizing microscopy for now, and wanted to share my experiments and experiences with the forum.

Openflexure project provides an LED grid solution for dark field, but I couldn’t source the original LED grid, and didnt wanted to port the code for alternatives, so I decided to make modular versions of the condenser module, one in which the Lens gripper is seperate from the condenser body, and add a condenser diaphragm/annulus in between the space (or anything else for that matter). Afterall, for dark field, the light needs to be indirect, and many traditional solutions are similar to this. But the original condenser in Openflexure is only for 5mm LED or an LED board (which I dont have), which won’t give enough illumination for an anulus based design. So some designing and lots of failed prints later, I designed an 8mm compatible illumination system, with detachable Lens Gripper, and a 3d printed annulus to go in between.
I also tried adding a linear polarizer instead of the annulus, and another linear polarizer in the gap between the objective and the tube lens in the unmodified optics module. Rotating the gripper allows to tweak how much light passes. When at 90 degrees, no light can pass unless its ‘unpolarized’ by something in between (the sample), thus giving us polarized microscopy. Inspired by this video:
(240) Upgrading a Cheap Microscope Lets You See Rainbows! - Polarized Light Mod - YouTube (240) Upgrading a Cheap Microscope Lets You See Rainbows! - Polarized Light Mod - YouTube

The results I think are quite good (dark field) (1st is my blood clotted, 2nd is my hair):

The polarized light results looked similar at polarizers@90 (some random glass debris):

However at angles != 90, it made everything colorful, here is a sperm slide (not mine, pre-prepared off of amazon):

Couldn’t capture it at 90 degrees as the cells weren’t visible properly.
Infact, I am unable to test the polarized version with any of the cheap prepared ones properly (maybe the dye used or something); I guess the samples aren’t bifringent enough. And I cant prepare my own for it, as don’t have 0.17 cover slips, and plastic tape interferes with the polarization, causing everything to appear white.

I calibrated the scope with no lens gripper (and no lens/annulus/polarizer) attached, with just the 8mm LED shining on the objective, As there was no way of calibrating it with the proper setup, cuz ideally everything is black and the software tries to make things Not-black.

Here is my scope and the parts:

Not including the modified Lid and the LED holder, they are trivial. Also, made an 8mm version of the aperture and modified the cutout

Let me know if I am doing things right and suggestions and advice. And If someone wants, I can put the SCAD files in a fork or something too. I have taken care to not modify any core SCADs to alter existing standard parts, and have taken care with the code quality, so maybe if admins permit, I can put it in a separate branch and make a PR too?

[ashish_kumar_4]

Added photos of my setup and the new parts

[j.stirling]

Nice work! The last time we did cross polarising microscopy I am pretty sure it involved taking a hacksaw to the microscope illumination!

If you have SCAD files it would be awesome to see them. If you fork the project on GitLab and put them in a branch, you can open a merge request to the main project (it is fine to mark it as a draft if you are still working on it, just having the draft merge request means people can find the work easily). Hopefully in time we can add any extra parts as accessories to the main repository.

[ashish_kumar_4]

Sure, would love to!

[ashish_kumar_4]

Here is the PR: Draft: Additional illumination components for Dark Field, Cross Polarized Microscopy, Modular Condenser (!332) · Merge requests · OpenFlexure / openflexure-microscope · GitLab

[nethead]

Great article. I would like to build a microscope for (blood) dark field imaging too. Which lense do you recommend with your current experience?

[ashish_kumar_4]

Sorry for replying late, was on a vacation.The exact same lenses recommended for the standard microscope. But because I couldn’t find the 5mm FL PMMA 13mm diameter lens recommended, I bought and used a 7mm focal length 13mm lens from edmund. The main thing to use is the 8mm compatible modular illumination modules with the condenser annulus

[tomvh]

That looks aweomse! Great work. I don’t really understand the gitlab commit. Are the .stl files somewhere available?

[WilliamW]

The STL files which are built under that merge request will be temporarily available under the View App button (right hand side of the central pane of the overview). This is not a stable link, it will change if any changes are made to the merge request, and I think there is a time limit on retaining the files, but I have checked that they are there now. The link will give you a full set of instructions that look like the normal released version, but contain all of the suggested changes. Built STLs are in customisations and alternatives, where you can download all of the STLs.
The Merge request does build the STLs but does not add any instructions to help with using them, but what @ashish_kumar_4 has posted above should be enough to get started.

The new STLs seem to have the names:
condenser_with_inbuilt_annulus.stl
condenser_annulus_standalone.stl
condenser_body_dovemount.stl
condenser_lens_gripper.stl
condenser_8mm.stl`` condenser_8mm_with_inbuilt_annulus.stl
condenser_lid_8mm.stl
condenser_led_holder_8mm.stl
condenser_8mm_body_dovemount.stl
condenser_aperture_8mm.stl

[tomvh]

Got the stl’s! Many thanks for the explaining!

[mzedp]

Cool stuff! I’d been wondering how to experiment with polarization/darkfield. I’m printing the new illumination now, and will try it later with the SMD LED.

Edit: Prints great! You’re right that the SMD LED is too weak for this dark-field annulus - I’m going to try modifying the annulus to see if I can see something using a smaller one.

Adding a smaller annulus that worked for me with the SMD led.
TestAnnulus_3p3.stl (24.3 KB)

Focus stacked image of a birds feather using the first method (calibrate without lens holder/annulus, then add lens holder).

This is with a 20x objective, maybe with a higher magnification the darkfield effect could be better.

Bringing the light source closer to the sample helps

Edit: Getting the height of the illumination right for each sample is crucial to getting the effect to work, I guess - I’m thinking of adding a thumbscrew for the illumination to control the vertical position better. Here’s some focus stacked moldy bread.

[ebob]

Sorry for the noddy-level question (I’m still in the foothills of a vertical learning-curve here), but where can I go to get hold of these stl files?

I see a reference to the “Customisations and alternatives” page, which I assume is this one:
Assembly Instructions - Customisations and alternatives

But nothing there relates to dark field, or stl files with names starting “condenser_”

Thanks!

[WilliamW]

You need to enter a parallel universe, via a Gitlab Merge Request. In post 5 on this thread there is a link to Merge Request 332. (If the acronym PR used in that post is confusing, it from a similar platform Github, where they call Merge Requests Pull Requests)

If you go to that Merge Request you will see a description of the intention to have parts for a polarised microscope. Below the description is a section about a ‘pipeline’, and with that is a button labelled view app. The view app button takes you to a complete version of the instructions that is associated with the merge request, you will see that the web address is different from the instructions that you get to from the OpenFlexure website. In those parallel universe instructions the customisations and alternatives tab has a link to a version of all_stls.zip which should have the parts mentioned in this thread.

[ebob]

You need to enter a parallel universe

Thank you, that makes a lot more sense now. Onwards into the multiverse!