P3D Viewers and white LEDs

[Tom Hubin <thubin@xxxxxxxxx>]

Tom Deering and I traded Emails recently about using white LEDs in a
viewer. Since I last wrote to Tom I have had occasion to use white LEDs
in a design for microscope illumination. I have also had a chance to
look at some homebrew viewer designs on the internet. 

I see no reason why a pair of white LEDs would not be sufficient for a
viewer. That would require a 9vdc battery, 2 white LEDs, a 90 ohm
resistor, and a switch wired in series. The current drain from the
battery would be 20ma. Power consumed by the LEDs and resistor would be
180mw.

The greatest waste of available light in viewers is due to using a wide
angle diffuser instead of condenser optics. Using a diffuser just
spreads light all over the entire inside of the viewer. Some of the
light reaches the eyepiece so you can see the slide. But most of the
light does not pass through the eyepiece. Very inefficient.

Let the LED or bulb light travel sufficent distance to fill the slide
area evenly. An LED with a 20 degree viewing angle will fill a 42mm
diameter circle after traveling a distance of 119mm. Using an LED with a
larger viewing angle will allow you to shorten this distance. A mirror
can also be used to squeeze this distance into a smaller package.

Then place a lens near the slide. About the same place you would
normally have a diffuser. The lens is to image the light source to the
center of the eyepiece. Since there is no heat to deal with you can use
a Fresnel lens. If the eyepiece is a 50mm lens then you will need a 35mm
lens near the slide to image the LED onto the eyepiece. You may need a
pair of 70mm Fresnel lenses combined to get the 35mm focal length over a
42mm diameter.

If you place your eye at the center of the eyepiece then all of the
light enters your eye and the slide should be very bright. If your eye
is a few milimeters off center then it is possible that none of the
light will enter your eye and the slide will appear very dark. This is a
problem with stereo because of the variety of interocular distances.

One solution is to diffuse the light just enough to fill the eyepiece.
So we put the diffuser back into the system in the usual place. But we
use a very weak diffuser. One that will diffuse the light only 20 to 30
degrees. This would fill a 50mm eyepiece lens with a 20mm diameter.

Most diffusers scatter light over a much larger angle. It is important
to use a small angle diffuser. An ideal diffuser would be a rear
projection lenticular screen. That would spread the light horizontally
but not vertically. Thus accomodating a variety of interocular distances
while giving up very little light.

I think that a very fine sandpaper or fine sandblasting on one side of a
clear plastic might be sufficient. To test the diffuser, point a laser
pointer at it. Then look at the light exiting the diffuser. Is it now a
20 to 30 degree cone? If so, this is great. If the cone angle is too
small then maybe both sides need to be sanded. If the cone angle is too
great then a finer sandpaper or grit might work.

Other materials can be tried as diffusers. Tissue paper. Plastic bags.
Any suggestions? Arrays of microlenses? 

Anyway, just some thoughts on improving the illumination efficiency of a
viewer.

Tom Hubin
thubin@xxxxxxxxx