Hi dear people,
I try to design a simple microscope that uses a LED with a collimation lens for fluorescence. The light of the LED gets coupled into the beam path with a dichroic mirror and is focussed into the back focal plane of the microscope objective (infinity corrected). One restriction of course is that I want the diameter of the image of the LED not to exceed the diameter of the optics at the back focal plane, or else I will loose light. At this point I have two questions:
- How can I find this maximum diameter? I can measure the optics or look at the mechanical desing, but this is only a rough estimation. Would it actually be the same as the diameter of the entrance pupil?
- Would it cause any trouble, if the diameter of the image that I project into the back focal plane would be very small? My guess is that this would be fine, as every single point in the back focal plane of the objective will be imaged into the whole aperture on the object side of the objective. Does this mean I could theoretically focus the LED into an infinitely small point at the back focal plane and still get a homogeneously illuminated object (also considering that I do not exceed the image sided numerical aperture of the objective)?
Keep in mind that I do not use a Koehler illlumination here, as I want a very simple setup.
Thanks a lot for any hints!
This is a problem I'm having at my lab that's really got me (and my PI) befuddled. Forgive me if I'm overlooking something obvious, as I'm simply an EE undergrad with no formal optics training/classes (yet).
Essentially, for the project we're working on, we need to confidently characterize the axial magnification of our system. From a few different sources, it seems like common knowledge that the axial magnification of an optical microscope is the lateral magnification squared divided by the refractive index of the immersion medium (assuming the detector/image of the sample is immersed in a medium of n = 1).
However, I'm having a VERY hard time getting experimental results that match this. Here are the results I'm most confident with: https://imgur.com/a/hnncNbp. I've done a few other experiments, but this is the most comprehensive one.
In this experiment, I measured the axial magnification for 5 different objectives (purple dots, a few different trials), all of which are immersion objectives (n = 1.518) using the same sample. The red line is M^2 / n. As you can see, the measured magnifications are consistently greater than the red line would predict. The scale doesn't show it, but the 10x and 20x see a comparable amount of error to the red line (relatively) as the 40, 60, and 100x.
More details about this experiment:
-
Loaded in the microscope is a test sample with fluorescent test structures that have easily measurable axial features at various scales (important, as I will be doing measurements for many different magnifications)
-
Attached to the microscope is a 4f re-imaging system (this is part of the research being done) - however, I've performed a few measurements and the re-imaging system does not appear to effect the magnification of the system (as is desired). What I mean by this is that the magnification of an image on the detector is very close (0.1% error) to the marked magnification of the active objective on the microscope.
-
To measure the axial magnification, I have the detector camera mounted on a translation stage. After locating the desired structure in the test sample, I move the detector using the translation stage in small increments, taking an image at each increment. This produces a 3D image of the image projected out of the optical system.
-
I load the image sequence into ImageJ, take the orthogonal projection, and measure the size of the axial structure in pixels (which is equal to the number of stage movements it took t
So I am in a need of a microscope with which I could quickly check my fluorescence samples by eye (eye peace or camera), so that I wouldn’t need to boot up the confocal. I am aware of a paper that suggests how to build one by 3D printing most of the parts. I was wondering if someone tried DIY’ing something like that or found a way to modify a cheap amazon microscope for fluorescence analysis? Cheers!
I'm building a fluorescence microscope for materials science/electronics purposes. I need to put this together cheaply. My background is in vacuum materials science, so I'm new to optical optics (rather than X-ray and ion optics).
My plan is to buy a commercial CMOS microscope eyepiece camera for ~200 dollars and a 365 nm UV LED source for ~200. I have a spare stereo microscope and that's what I am using as the base of the scope. It is for reflection illumination only. I have a ton of questions though.
1.) Do I need a UV pass filter for the 365 nm source? I saw videos of 365 nm UV and it looked like it was almost faint white light, and I'm worried fluorescence will be swamped by the white light.
2.) Is 365 nm better as a fluorescence source than 385 nm? 365 nm does have higher photon energy, but due to the lower efficiency the spectral brightness is weaker than 385 nm for the same power draw.
3.) How intense do I need the UV source to be anyways?
4.) Would I need a low light CMOS camera specialized for fluorescence?
5.) Do I need a UV block filter for the CMOS camera?
6.) How do I achieve spectroscopic capabilities for <1000 dollars?
Thanks!
The GFP was C-terminal tagged to the protein.
Locating the protein in the peroxisome after the cell fractionation was done by a Western Blot and not by microscopy.
