How do light meters work?

I'm wanting to know how light meters (spot meters in particular) work. Not how to use them, but the actual electronic bits! I'm thinking about the kind of details that would enable me to build one from scratch using off-the-shelf components. Obviously I'd get a much more accurate and more reliable meter if I just bought one, but like learning about lens design I'm interested in it even if I never actually go out and try it. Things like what kind of light-sensitive parts do they use, how does that get converted to an exposure value, and how do they limit the measurement angle to one degree without an extremely long focal length lens? This may be too complicated to answer in a forum post so if someone could direct me to appropriate resources that would be great!

Some info here:

http://bigbro.biophys.cornell.edu/~...on/Light_Meter/Build_Your_Own_Light_Meter.pdf




Steve.

Steve Smith said:

Some info here:

http://bigbro.biophys.cornell.edu/~...on/Light_Meter/Build_Your_Own_Light_Meter.pdf

Click to expand...

If anyone chooses to follow those instructions, I have about 27 metres of 1.5"/38mm plastic pipe spare after finishing my darkroom plumbing, that anyone is welcome to collect . A bit narrower diameter than they suggest, but enough to make about 100 light meters...

(Damn the combination of minimum order quantities and my massive impatience .)


Regarding the spots, I would think you wouldn't need a ridiculously-long-focal-length-lens. I'd presume you take a relatively normal length lens (for the viewfinder,) and use a half silvered mirror to split the image to eyepiece and sensor. Just put the sensor on the focal plane and make the sensor small enough to only pick up the central spot (rather than integrating the whole of the captured light) and mask everything else off. I'm just guessing though.

I made one way back when but not a spot meter. I just remember it was a simple electronics project. Made a pretty nice enlarging meter though.

I have an interesting enlarging meter that came from an estate lot. It was a module that was supposed to have a thermal sensor, and be a digital thermomenter. The creative chemical engineer that made it realised that a variable to light resistor ( this electrical engineer forgets the proper term - it is not a CdS cell in this case) would also drive the digital display. It is not too useful for absolute readings, but is very handy to use to duplicate the base exposure time after you change the head height.

back when I was starting to use IR film I heard that CdS light meters don't see IR

so I went to Radio Shack, got a CdS cell and an IR sensor ( a phototransistor IIRC ) and tested them both with a volt/ohm meter

turned on the stove...CdS cell didn't see the glow from the hot stove....but the IR sensor sure did

Keep in mind that CdS works with one type of circuit, Selenium works with another type of circuit, silicon photodiode yet another, and CCD and CMOS in yet others.

A light-detecting circuit should be simple to build with off-the-shelf parts indeed. The difficulty lies in interpreting the output and calibrating it such that it is meaningful. Depending on the linearity characteristics of your sensor, this could be hard. I know, because I've been trying. CdS LDRs are green sensitive and cheap but slow and have memory effects (no good for flash meters). Most photodiodes have peak sensitivity in IR and are theoretically linear, but have very low output and a finite dark current requiring the ability to zero-point calibrate. Phototransistors should be pretty decent except most of them are also tuned for IR and may not be fast enough for flash metering, and may consume more energy. Most silicon solar cells have a radically unlinear response to light and may not be sensitive enough at the dark end. Selenium cells would seemingly have pretty good properties, but good luck finding them. Most of these issues nowadays are conquered rather easily in software processing. And if you simply want something to compare values to a known (such as a comparitive enlarging meter) then absolute readings don't really matter. As for your question about spot meter optics, I'm curious of that as well. It seems the only thing to do would be use a long focal length lens or pinhole arrangement and a small, sensitive point source sensor.