DIY light-activated switch
I am considering a project that involves setting up a 4x5 to take an (unattended) early morning photograph once the light reaches a certain level. This would be dawn, but pre-sunrise. I would appreciate any feedback/advice on the approach I am considering.
I have some experience building devices that use microcontrollers to monitor temperature, and so want to do the same with photoresistors. My thought was to create a diffuser (would half a white ping pong ball work?) that has a photoresistor placed under it to read the brighness. I would then do some tests on location -- wait for the desired light level, use my hand-held light meter to see what the exposure would be, and at the same time record the resistance value from my DIY setup. The idea is that the microprocessor would watch for the desired resistance, and when it reaches the threshold value take the photograph. Would a lens to focus the diffuser shell onto the photoresistor make it more accurate? Would it be necessary?
On a related front, any thoughts on devices to press the shutter release cable?
As a child I built such a switching device out of a Philips electronics teaching kit. Don't remember the circuit, but sure the kit did not contain microprocessors basck then...
Once there were solenoids with push-pin and male shutter threading. At least my gives quite a bang, so use a long extension cable release inbetween (in case you could even get the latter...)
What about activating a cable release by an electrical motor and an excenter-wheel?
Last edited by AgX; 10-31-2011 at 12:47 PM. Click to view previous post history.
I doubt the need of a lens behind the diffuser, I don't believe any incident meter uses them.
If you're going out in the woods at sunrise(ish) wouldn't it be just as easy to adjust the sensitivity of the device while you're out there?
After considering several options I'd suggest an alarm clock that activates a human who's programmed to execute the following code:
if light_level = desired
wait 60 seconds; retry
re: adjusting while I'm out there ... that's the thing -- I won't be there. I need to be able to set up the camera to take the photograph in situations where I can't be on-hand. (But perhaps NDRS has a programmable human he can donate to the cause ...)
re: lens ... the more I think about it, the more I think a condensor is needed. The diffuser is going to be at least a half-inch across, maybe twice that. The photoresistor is roughly .12" in diameter. Without a lens, most of the light passing through the diffuser won't hit the photoresistor. Yes?
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
For those of us old enough:
Originally Posted by ndrs
10 RELEASE = 156
20 INPUT LIGHT
30 IF LIGHT => RELEASE THEN 50
40 GOTO 20
50 PRESS SHUTTER
"People who say things won't work are a dime a dozen. People who figure out how to make things work are worth a fortune" - Dave Rat.
Most incident light meter has the dome diameter approximately 3/4 inch and yet the diameter of the photocell is 1/2 inch or less and there is no lens.
Originally Posted by couldabin
I've no personal experience with them, but the servos used in radio controlled models might offer an actuator that could run the shutter release.
I suspect you won't need a lens unless you're trying to trip at extremely low light levels and need to concentrate the received light -- on the other end, getting too carried away with that could potentially fry the sensor in bright sun!
I would have probably used a phototransistor instead, they usually have somewhat better sensitivity. An analogue comparator is the only device you need to implement the rest of the logic.
There are a number of IC comparators available, in the worst case most 16 bit PIC processors have at least one comparator on the chip.
But in your pre-dawn low light situation even that wouldn't be enough I guess. The problem would be that dark current would be very similar in the magnitude to your light signal. You will be detecting noise rather than the required amount of light. I know what I am talking about, because I tried building a low-light light meter last year. I would recommend using TSL230R Light to Frequency Converter from The Texas Advanced Optical Systems - it is the only thing that worked for me in the end for that amount of light.
PS: PM me if you need more details - my light meter worked in the end
Last edited by vyshemirsky; 10-31-2011 at 07:11 PM. Click to view previous post history.
Reason: added PS
You don't need a lens behind the diffuser, in fact you really don't want one at all because that will introduce an angular-dependence to the meter's response. You don't need "all" the light from the diffuser, just a constant fraction thereof. Ping-pong ball is perfect unless you can source a white acrylic dome, which will be a bit more transmissive.
Certainly you can make a light-triggered circuit, but photoresistors are notoriously wonky and variable in their output (temperature dependent among other things) so it won't necessarily trigger at exactly the light level you want. It'd probably be OK for a B&W shot if you aim to overexpose a little, but I wouldn't trust it at all for colour. Some experimentation would be required!
A (much) more accurate option would be to use a phototransistor, i.e. the same devices present in a real light meter. IMHO you need to admit that what you're building is a light meter and since you have a microcontroller involved, actually build a proper light meter that will compute exposure. That would allow you to (assuming B mode on the shutter) accurately set the exposure time, or if the shutter time is fixed, the light meter can wait until the measured level is exactly the required level. Beware though that phototransistors are very IR-sensitive, so you either want a hot-mirror to cut off radiation below 700nm or you need to accept that you will have an exposure error that is dependent on the time of day.
The other option is to just build a timer with your micro. If you know what time of day will have the lighting you want, a timer will suffice and is much, much easier to build.