A few days ago the led and its driver finally arrived. I believe that the caravan serails were on strike so it took the camels a wee bit longer to cross the Mongolian Highlands. But on the bright side, mounting the pair onto the illuminator turned out much easier and quicker than foreseen. An element from an old 50mm Practica lens, the rear one, I think, perfectly falls into the position of the collimator lens. Two telescopically fitting cardboard tubes and a decent amount of tape made a rather stable fit with the option to vary the distance between the led and the newly installed collimator lens.
Here's the schematic drawing of the set-up:
Heat is still a bit of an issue, what with no heat sink in place. Luckily I had chosen a led driver with three modes. Additional to full power, it can be set to half power, which turns out to be exceedingly useful, and to blinking. For the latter mode I haven't thought of a use yet. The led is mounted on an L-shaped aluminium bracket that gets hot, so is leading the heat out of the house. Up to some 10 minutes of continuous use on half power, I have not yet seen heat related problems.
Does it work? Well, the light Judged from photographic shutter time it's about half the light intensity of the Labophot's backstage light. output is sufficient for comfortable viewing, even at half power. Note that the A-aperture was wide open and no filters were used. Later, when I had brought the led into a better position, the light output increased by about 30%, if the scale of a photoresistor module is anywhere near linear.
This is the first photo made with the led-powered illuminator:
To align the filament (for lack of a better term) with the rest of the Köhlerian system, the collector ought to project a conjugal image of it at the plane of the A-aperture. Even though the illuminator can be dismantled to some extent, I found it impossible to check this directly. I could place a paper disk at the A-iris, but it would be impossible to see it with the lamp mounted. However, I noted that when I hold the illuminator with light on about 10cm above my desk, I can see a projection of iris blades in focus. At first I was misled to think these belonged to the F-(field) aperture, but they move while adjusting the A-aperture. I guess this projection must be the back focal plane (?); if seems to fall in the upper parts of the objective.
Now I bent a piece of metal wire across the led to get some contrast and moved the led forward and backward, unitil I got a focused projection together with the aperture blades of the A-aperture. Am I right to think that this is the optimal z-position of the led? It is not the position with maximum light output, but the difference is less than 10%.
As for the quality, the only thing I can test now is the even-ness of the illumination. I made a photo of the surface of a piece of white acrylic and then plotted the luminace over the diagonal from top-left to bottom-right. Here's the result (blue line), compared with a similar test using transmitted light. I note here that I still have this slight excentricity issue with my camera mount. With transmitted lighting I can center the condenser optimally for the camera; with incident light I'll have to do with how it falls upon me. The near full-stop vignetting at the right of the chart is close to where the field aperture cuts it. The wire is still in front of the led, but I can't discern a shadow of it.
Now, all this is done with a temporary test set-up in order to determine the parameters of the long-term lighthouse. I am breaking my head now on how to make the led center-able in three dimensions while connected to a sufficiently large heat sink and keeping the construction light-weight. Especially for the xy-centering I could do with practical tips. I have ideas, but it's getting complex and that normally is not a good start.
Wim
