Building an Abbe Diffraction setup

With our microscopy course coming up, I’ve had to once again rebuild my optical bench microscope for demonstrating that the back focal plane of the objective shows the Fourier transform of the sample, and that modifying information at the back focal plane changes the resulting image of the sample. I thought I’d provide some documentation of the microscope, both in case someone is interested in reproducing it, and as an example of how to build a simple microscope with off-the-shelf lenses. This is the same setup I used in my Abbe Diffraction lecture on iBiology.

An optical bench microscope made from simple lenses with readily accessible back focal plane.

A photo of the completed scope; click for larger version.

The microscope, as shown above, consists of a white LED set up in a standard Kohler illumination configuration with field and aperture diaphragms. The sample is imaged by an objective lens, and there is a 50-50 beamsplitter in the detection path; half of the light passes through a tube lens to form an image of the sample on the object plane camera (on the right). In the other light path, a projection lens reimages the back focal plane of the objective on to a second camera. All of the parts are from Thorlabs except for the cameras, which are from The Imaging Source (at the time I originally built this, Thorlabs didn’t sell the cheap cameras they do now; if I were building it from scratch, I would probably use Thorlabs cameras).

All lenses are achromatic doublets to minimize aberration, although some of the illumination lenses could probably be replaced by singlet lenses without a problem. The objective lens has a 30 mm focal length and the tube lens has a 150 mm focal length, giving a magnification of 5x. The LED collimating lens has a 20 mm focal length and the field lens has a 150 mm focal length, giving a 7.5x magnification of the LED on the aperture diaphragm. This doesn’t fill the back aperture of the condenser lens, but it comes close.

To align this system, I started by aligning the tube lens so that it focused an image from infinity on to the camera. The tube lens is placed with the flatter side facing the camera, and I then pointed the breadboard out a window and moved the tube lens to get an image of a distant object (>100 feet away) in focus on the camera. Next, I assembled the LED assembly and focused the collimating lens to give a collimated beam. Thorlabs has good instructions for doing this; you can also get close by focusing an image of the LED on a distant wall. Then, I placed the beamsplitter, objective lens, and sample holder (none of these distances are critical) and adjusted the sample to objective distance until an image of the sample was in focus.

Once this is done, you can align the transmitted light optics; if the LED collimating lens is focused at infinity, then placing the condenser lens will produce an image of the LED one focal length away from the condenser. This can then be focused on the sample. Finally, you can place the field lens and the two irises; the aperture iris should be placed at the back focal plane of the condenser lens so that it controls the condenser NA; the field lens is placed to produce an image of the LED on the aperture iris; and the field iris is placed so that it is in focus at the sample.

Now you can focus the back focal plane camera. Place a sample in the sample holder that produces a strong diffraction pattern; a stage grid works well. Focus the objective so that the grid is in focus, close down the aperture iris all the way, and close the field iris so that only the grid is illuminated. You should see a clear diffraction pattern at the back focal plane of the objective (I have a second sample holder placed here to allow manipulating the diffraction pattern). Then place the projection lens in the reflected path of the beamsplitter; you should see an image of the diffraction pattern some distance from that lens. Place the back focal plane camera to collect this image and adjust its position until the image of the diffraction pattern is clear.

If you want to build one yourself, here is a partial parts list (I believe it has all the parts except for the cameras, green filter, screws, and sample holders, which were 3D printed). However, if I were to build this again from scratch I would probably build it on an optical rail to minimize the number of degrees of freedom when aligning the optics. It shouldn’t be too hard to replace the post mounts with appropriate rail carriers to simplify alignment. The total cost for the parts listed is $2640.

PartDescriptionNumberPart NumberPriceTotal Price
Objective lens30 mm Achromat1AC254-030-A77.0077.00
Tube lens150 mm Achromat1AC254-150-A71.3071.30
Field lens150 mm Achromat1AC254-150-A71.3071.30
Condenser lens50 mm Achromat1AC254-050-A71.3071.30
Projection lens50 mm Achromat1AC254-050-A71.3071.30
Irises2ID2554.06108.12
Lens mounts6LMR115.2391.38
Mounts for 50mm lenses0.5" tube2SM1L0512.5925.18
Objective focusing adapter1SM1V0529.6029.60
White LED1MCWHL5187.51187.51
LED driver1LEDD1B284284
Power supply1TPS00125.0025.00
SM1 Coupler1SM1T219.5019.50
Collimator mount1" tube1SM1L1014.2514.25
Collimating lens20 mm asphere1ACL252017.2217.22
1/2" posts12TR25.1962.28
Post holders3"11PH3E24.00264.00
Post holders2" for cameras and mirror3PH2E23.5070.50
Mounting forks13CF1759.90128.70
Flip mount for filter1TRF9078.0078.00
Mirror MountFor Beamsplitter1KM200T89.8089.80
Beamsplitter mirror1EBS268.0068.00
Breadboard1MB2436700.00700.00
Breadboard handles1BBH115.0015.00