Concept: LED Brightness controlled by photo sensor.

Hello...
I've brought up this subject before a long time ago but it didn't really take off... The concept is really quite simple... Take the idea of a standard night-light... You know.. the kind they sell at hardware stores to mount in your bathroom?..They turn on when the main lights go out, and then when the natural light from the morning sun coming up hits the sensor, the night-light turns off...

So.. Have you ever very slowly shined a light at the photo sensor of the night-light in a dark room? If you do it careful enough, you'll notice that the night-light will slowly dim down, and eventually turn off... Ah.. This got me thinking. Watch this video AND this video to understand what I mean.

Many people these days are into variable brightness, multi-level, and so on. So... why not instead of developing a UI to manipulate the levels, just expand on the idea of a photo sensor?

The solution in my mind is simple.. Keep in mind that I am not very knowledgeable in circuitry so I don't know how 'real world' idea would work. But why not attach a photo sensor to the driver of the LED? Have the sensor mounted facing forward, on the edge of the bezel, and perhaps recessed a bit... The idea then being that aiming the flashlight at distant objects will not reflect enough light back to influence the photo sensor and so the LED will be bright, but that as the user shines the light on closer objects, the photo sensor will pick up on the reflected light and dim the LED to a level where the reflected light no longer is strong enough to influence the sensor.

This way, if you shine a light at a book page, the sensor will dim the LED way down to a level that is easy on the eyes, but as soon as you shine the light across the room, the sensor will raise the output level because it isn't detecting enough reflected light to dim the LED. In other words, it would be automatic brightness control. I understand this wouldn't work for many people, but I mean, in this age of innovation why not create it?

Has there been any designs like this? Would it work?

If you read all the way through this, thanks.. I'm very honestly interested in this concept and would like to determine if it is possible in any fashion.
All thoughts and ideas are welcome.

EDIT:5-18-2008
If I were to further describe the system, I would say that there is a tail-cap button, which allows a user to pick the tolerance of light received by the photo sensor, making it either: highly sensitive to light, moderately sensitive to light, or somewhat insensitive to light.
With this type of selection, the flashlight can be told to shift (dim and brighten) only between a specific range of sensitivity. That would enable the user to create a threshold of brightness, which would only vary in output by exactly the amount of lumen in the range between sensitivities.

So one click on the tailcap makes the photo sensor 0-32% sensitive (resulting in bright output that will only slightly respond to signal), two clicks makes it 32-64% sensitive (likely to be dimmer, but is not limited from being bright) and three clicks makes it 64-96% sensitive (likely dim or very dim, but again not limited from being bright) to light. Four clicks will always reset the sensitivity to 100% (infinitely and fluidly dim or bright depending on sensor input). Any one of the three sensitivities can be accessed upon clicking the proper amount for it. One long press turns the flashlight on and off.

We have to remember that the user is not modifying the brightness of the LED. The user is modifying the sensitivity of input from the photo sensor, which in turn affects the brightness.

But that entire explanation also distorts the root of the idea in a way. I'd much rather see a kind of fluid reaction to light levels, a real-time response to the environment. Maybe that would mean no user input at all. Just on, and off.

Sorry if I'm being long winded, but I've had this idea for a while, and I need to air it out.. Somebody, come shoot a hole in it, or perhaps make it come to fruition; either way it just needs to get settled. Lets build one together, or deem it impossible/unreasonable.

EDIT: 5-19-2008
So now, due to more thinking and development, there needs to be some set rules and requirements for this concept.

1.) It would not be the only light you had with you. It's a special purpose light, not general duty. (at least in this conceptual form)

2.) There would be no UI besides On and Off (at first). The photo sensor would be the only thing determining brightness, and its range of function would be between completely off, and the LEDs maximum brightness. (proof of concept)
a.) If there is a UI designed it would be made to adjust the gain and/or threshold max/min settings. (added once proof of concept was achieved)

3.) Add a function to disable the sensory system completely and have a way to immediately go to full brightness. (tactical operation)

4.) The flashlight should only be used in dark environments, where other light sources are not able to influence the input of the photo sensor.
a.) A calibration period may resolve this issue. Upon switching the flashlight on or 'activating' it, there would be a brief moment where the electronics would take a reading from the sensors, average them out (if there is more than one sensor), and then activate the LED at a level configured to coincide with sensor readings.

EDIT: 5-24-2008
Well, it's time for some pictures... It's going to take your bandwidth for a ride (just kidding)..These are pictures of the circuit board found in the night-light that you can see in the videos above.





In the top photo: The little blue cylinder on the bottom left of the circuit has "50 v, 1uF" written on it. Next to that is a three pronged black pot that says "mcr100-6, 031"..Above the two paddles, I see the CDS (cadmium photo resistor), followed by a capacitor banded "Gold, Green, Yellow, Red". Below the capacitor is a black cylinder that says "IN" and it has a silver band on the right side.. Printed on the PCB itself is "07V0, 94v-01, SCR" and in larger digital lettering there is a number "0709" in the center, under the paddles.

The second photo shows the electrical path of those components. From top down the solder points are "CDS, C1 (Red on left, Gold on right), Black Cylinder, AC input and lamp contacts, then three prongs of the black pot, and two from the blue cylinder."

According to the packaging, the lamp is taking 3.7v and an input of 120vac from the wall. Anyone think it's possible to modify the circuit to take an 18650 battery current?

Sorry if I'm not calling things by their proper name. I'll use the right words if I have them.

I think the concept has merit, but it wouldn't be ideal for every situation. I'd like to see something like that made just for kicks.

We have some lighting on machines where I work that works on that principle to some extent. If you could adjust the gain and/or max/min settings, I know I would be interested.

Added to OP.

I think it is a very interesting concept and could work very well.

However the success will depend on the quality / reliability of the photo sensor.

I have a IR sensor yard light that has failed on me. It will fire randomly, maybe 30% of the time when I walk past it. I imagine photo sensor might be susceptible to the same type of sensor failure.

So depending on the quality of the sensor, there could be two outcomes: 1) a very expensive, but functional light with good sensor response, or 2) an affordable light with a sensor that works intermittenly or poorly.

This of course would be a great light for the aging population, and could be a commercial success with the right combination of price and functionality.

Personally as a gadget guy I'm more interested in a light that can be programmed via a PC interface (bluetooth?? USB?? ) for custom UI. That way programming can be done quickly and easily, and everyone gets their own UI that they prefer.

The main gripe I have with the current crop of "intelligent UI" lights is that they are still a bit tricky to program, with a lot of bezel turning, etc...

That would be nice, but different situations call for different lighting needs. For example a LEO incapacitating someone at close range needs lots of light, but at the same distance only needs a little light for reading. With your explanation, one or both will be the wrong intensity.
A better solution might be to run it through a standard driver first(uses the tail button), and then route that through your photo sensor(side buttons). Either can still get very dim, but if you introduce the standard driver, you limit the high end for low light situations, and by adjusting the sensitivity of the sensor, you can boost up the low end. I would prefer a potentiometer on the side, but not everyone likes that.

Everyone's needs and desires are different, so I'll be the first to disagree. I prefer simple and reliable. While many of the lights I have do have electronics for voltage regulation or level selection, the ones I rely on the most are either the simplest or those that are proven to be the most reliable. Why add all of this to a light when the operator can do the same thing if the light has a suitable interface, such as that found on the U2? I do not need a sensor to tell me I need to dim the light before trying to read a map. Plus, no matter how fast the electronics would respond, the sensor would still have to "see" the light to recognize that output needs to be dimmed. To me that means that the ooperator will also "see" the light, even if only for a second.

As a toy for those who like gadgets, I can see the value. I am much more skeptical about real world use in demanding situations.

Thanks, those are some great points..
... There's still so much to think about, including all the problems that will arise.. If i had the knowledge, I would just buy the sensors (which are readily available at radio shack for little solar kits) and find a way to install them in pre-existing driver boards..
The problem i had was that night-lights run on 120vac from the wall outlet.. I don't know how to manipulate it to take a DC source like an 18650 battery. If someone could draw up schematics on how to properly integrate the sensor onto a common driver board, or how to modify the night-light from AC input to DC, then the concept would be able to move forward in my mind.

LED-holic said:

I think it is a very interesting concept and could work very well.

However the success will depend on the quality / reliability of the photo sensor.

I have a IR sensor yard light that has failed on me. It will fire randomly, maybe 30% of the time when I walk past it. I imagine photo sensor might be susceptible to the same type of sensor failure.

So depending on the quality of the sensor, there could be two outcomes: 1) a very expensive, but functional light with good sensor response, or 2) an affordable light with a sensor that works intermittenly or poorly.

This of course would be a great light for the aging population, and could be a commercial success with the right combination of price and functionality.

Personally as a gadget guy I'm more interested in a light that can be programmed via a PC interface (bluetooth?? USB?? ) for custom UI. That way programming can be done quickly and easily, and everyone gets their own UI that they prefer.

The main gripe I have with the current crop of "intelligent UI" lights is that they are still a bit tricky to program, with a lot of bezel turning, etc...

Click to expand...

GreySave said:

...As a toy for those who like gadgets, I can see the value. I am much more skeptical about real world use in demanding situations.

Click to expand...

I agree a light like this would not be suitable for demanding situations. However it would be useful for far more than those who like gadgets. As a convenient / intelligent light for the elderly it would be a great tool.

CostCo sells a package of light sensor 3AAA LED stick-on lights. I use them in my garage for additional ambient lighting and find them useful. So there could be practical uses for a light like this.

EDIT: these are the lights I'm referring to. They turn themselves on when they detect low ambient external light.

It's true... As far as real hard-use applications, this style of 'sensory' seems out of place, as tekno_cowboy stated, it's going to respond incorrectly if subjected to specific types of dynamic life interaction.
Or as you stated, that the user will begin to anticipate the sensor, and therefore perceive a latency between what it senses, and how quickly it can respond. Unlike the U2 though, the user doesn't have to turn a dial and watch the level go down. I would assume the sensor would be faster than a U2 at finding the level appropriate.

Thanks for your input.. This is exactly the kind of discussion i was hoping for.

GreySave said:

Everyone's needs and desires are different, so I'll be the first to disagree. I prefer simple and reliable. While many of the lights I have do have electronics for voltage regulation or level selection, the ones I rely on the most are either the simplest or those that are proven to be the most reliable. Why add all of this to a light when the operator can do the same thing if the light has a suitable interface, such as that found on the U2? I do not need a sensor to tell me I need to dim the light before trying to read a map. Plus, no matter how fast the electronics would respond, the sensor would still have to "see" the light to recognize that output needs to be dimmed. To me that means that the ooperator will also "see" the light, even if only for a second.

As a toy for those who like gadgets, I can see the value. I am much more skeptical about real world use in demanding situations.

Click to expand...

I can't say that I would use one for practical, day-to-day use. I have plenty of lights that do that reliably and efficiently already. It would be nice for some applications, but I can't see myself ever using it in a situation where timing was an important factor.

Tekno_Cowboy said:

A better solution might be to run it through a standard driver first(uses the tail button), and then route that through your photo sensor(side buttons). Either can still get very dim, but if you introduce the standard driver, you limit the high end for low light situations, and by adjusting the sensitivity of the sensor, you can boost up the low end.

Click to expand...

Absolutely.. This would be one method to try in an experiment to get a real-life example of use..
If only I knew how to do that.

Contrary to the nay-sayers, I think this is a great idea! Especially for maintanence folks, mechanics, etc... those non-tactical folks that go between the inside of a car engine then back to the tool chest, or those guys that you see in the hall at work, on a ladder with half their body in the roof playing with the wiring. Anyone that alternates between lit and non-lit environments.

Sorry I can't help with making it work, but I'd REALLY suggest you keep track of this thread and maybe your original, just in case this type of light ever gets made. Then you can prove it was your idea first, and make a scad of royalties (of if you know any good copyright lawyers).

I think it's a good idea for a everyday light.

I'd add a way of making the sensor less sensitive and add another sensor to detect ambient lighting, so if the light in the room is bright, and you still need a light, the light will be brighter so it can be helpful. If that sensor(the ambient light one) senses a low enough level of light, the input will then change to the sensor in front so the light will dim as the reflected light gets brighter.

I'd also add a easy way of over riding everything and go straight to full output.

I'd love to see it used in a real light.

I like the idea of a light that adjusts by itself to changing light levels. There actually are lights that do this now, just not flashlights. My 2009 Toyota Corolla has daytime running lights on light sensors. When it gets dark, the headlights automatically come up to full brightness low beams and all of the guages and readouts behind the steering wheel light up. It would be nice to have my lights do the same thing, but it would be more complicated. Most light sensors act like a switch going on or off only. You need to have one that has multiple setting, not just on and off for your idea to work. I do hope that it can be done and that someone does it.

Enzo Morocioli said:

So.. Have you ever very slowly shined a light at the photo sensor of the night-light in a dark room? If you do it careful enough, you'll notice that the night-light will slowly dim down, and eventually turn off.

Click to expand...

I've never been able to accomplish a smooth, variable transition in extinguishing a night light by aiming another beam at it. They ALWAYS flicker on me and then go out. I trust most sensors aren't sensitive enough for smooth transitions. If they are they're probably fairly expensive.

From the little knowledge i have i can suggest to try this as a proof of concept:

Use a 20KOhm photoresistor ( google it and you will find many ) and wire it instead of the 20KOhm pot on a shark converter board.
In theory it should do the same work as the pot on the shark and increase/decrease the output according to the amount of light it receives. Im not sure on how the sensitivity could be set but maybe some resistors onboard the shark are used to control the sensitivity of the pot and by replacing them they can change the sensivity of the photo resistor too.

Dunno but it might be worth a try.

Hello Enzo --


I've often thought about this very thing ! :thumbsup:


Certainly is an interesting concept, isn't it.



Thank you for putting it all into words. :goodjob:



Not quite so easy to explain, eh ?

:twothumbs
_

This is similar to the idea behind self adjusting electronic flashes for cameras: a sensor measures the amount of light reflected from the scene and cuts off the flash (it is a very fast sensor) when the total light received hits some set level.

This assumes that there is a single or settable total reflection level you want to achieve. If you want to light something brighter, e.g. extra small print on some text, then bringing such a light closer to the page won't help.
I'm probably old fashioned and prefer to let my eye adjust to varying light levels rather than the flashlight...

I've never been able to accomplish a smooth, variable transition in extinguishing a night light by aiming another beam at it. They ALWAYS flicker on me and then go out. I trust most sensors aren't sensitive enough for smooth transitions. If they are they're probably fairly expensive.

Click to expand...

There's no reason a smooth transition couldn't be done. Those night lights were never designed to have a smooth transition, they were designed to provide a sharp on/off switch. It's just that with just the right amount of light, the resistance on the photoresistor is just enough that the transistor logic is forced into a "forbidden region" in between discrete on/off states, which is what leads to unpredictable behavior like flickering etc.

A circuit could be designed however to actually very specifically vary the brightness of the light as a function of the photodiode brightness. IMO a "safer" way of doing it than connecting a photodiode directly to the feedback loop on the LED driver circuit would be to acutally have a digital photodector, which controleld a digital potentiometer on the feedback loop. By implementing a digital system, better stability couild be attained, and the user would have the optino of tweaking the minimum, maximum settings, and the gamma (or rate of change of brightness) for the LED.

IMO though, there really isn't al ot of practical use for this for flashlights or mobile operations, as I believe a human operator with a well designed interface could do much better than any photodiode.

Where this idea (optical feedback) IMO makes sense is with fixed lighitng in rooms where there is a mixture between incident natural light, and artificial light. Sometime in the future I'd like to see rooms with embedded sensors in various places, that adjusted the room lighting so that both illuminance/lux at various surfaces (such as desks) would be constant, AND so that color temperature would be matching. IE, if there is a cool white light coming in fromt the windows (natural daylight) the lighting in the rest of the room should be made to match. At night time, the lighting would then setting at a lower (warmer) color temperature, and slightly lower overall illuminance, all of which would continuously reglated by a series of photosensors distributed throughout the room.