Sensor to distinguish between daylight and artificial light in a room?

[maryammagdy]

is there a way to distinguish between daylight and artificial light in a room? I'm thinking my options are a light intensity sensor or a color temperature sensor. The sunlight is supposed to be much stronger than artificial light. But I'm wondering if the case will be the same inside a room. Or does the idea seem silly to distinguish between both?

 

[Pommie]

You may be able to detect the 50/60Hz strobing of artificial lights.

Mike.

 

[dknguyen]

In theory, you could use your RGB colour sensor and program in some tolerances for the spectrum of natural sunlight which is different than most lamps. You can use colour temperature of the sun vs commonly available bulbs to get starting values so you can tune. The sun is 5800K, though being indoors behind windows might affect this. Common bulbs are 3000K (those horrid red orange bulbs for basement and bedrooms), 4100K (most common office white lighting), and 6500K (not so common).
https://en.wikipedia.org/wiki/Color_temperature#Categorizing_different_lighting

But it might not work unless you're actually looking at the light source. The 60Hz detection mentioned earlier might be more effective if you can't.

The intensity sensor won't be any good.

 

[crutschow]

I like Pommie's idea for its simplicity (although the light fluctuation frequency is likely 100-120Hz for most artificial light sources).

 

[JonSea]

I don't think frequency detection will work for incandescent bulbs.

 

[KeepItSimpleStupid]

Dimmable LEDs will have a switching frequency all to itself.

AM 1.5 Global is the solar spectrum, but it's going to have some shift based on time of day. For a given room, you check for some predominance or abstinence of certain wavelength bands.

Brands of lights will also shift the spectrum. Sylvania and GE ELH lamps have a different spectra.
Why do I know. I had an application where it mattered.

 

[Nigel Goodwin]

I don't think frequency detection will work for incandescent bulbs.

Shouldn't be much of a problem in the 21st century

I don't think it would work too well for flaming torches or oil lamps?.

As in most of these threads, we need FAR more information - for a start WHY does he want to do this?.

 

[JonSea]

We still have incandescent bulbs in our house....2 or 3 I think

 

[Nigel Goodwin]

We still have incandescent bulbs in our house....2 or 3 I think

Shame on you!

 

[dknguyen]

Shame on you!

We have a few dozen. All in the ceiling chandelier thingies.

 

[crutschow]

Even incandescent lamps have a slight detectable flicker.
I've used them to detect the strobe pattern on a vinyl record turn-table strobe-wheel to determine if the turn-table rotation speed was correct.

 

[Nigel Goodwin]

Even incandescent lamps have a slight detectable flicker.
I've used them to detect the strobe pattern on a vinyl record turn-table strobe-wheel to determine if the turn-table rotation speed was correct.

Like you say, they have an easily detected 50/60 Hz flicker.

 

[gophert]

I don't think frequency detection will work for incandescent bulbs.

Easily detectable with a phototransistor capacitively coupled to a high gain amplifier.

 

[crutschow]

Like you say, they have an easily detected 50/60 Hz flicker.

Actually 100-120Hz but why quibble?

 

[gophert]

Actually 100-120Hz but why quibble?

Still 60Hz on an incandescent light.

 

[crutschow]

Still 60Hz on an incandescent light.

Never said otherwise.
But the light intensity changes for each peak of the sine wave, giving fluctuations at twice the line frequency.
Another way to look at it is that the power output changes with the square of the voltage, which is a double frequency function.
Sort of acts like a full-wave rectifier.

 

[OBW0549]

I don't think frequency detection will work for incandescent bulbs.

It will work; bulb filaments heat and cool significantly throughout each cycle of the applied AC, giving the light intensity a significant ripple (around 5% for a 100W bulb, if I recall correctly from measurements I made years ago). The ripple occurs at twice the line frequency, since the filament heats up on both positive and negative peaks and cools in between.

Low-wattage bulbs have the most ripple in their light output because their filaments are light and have little thermal mass; therefore they heat and cool the fastest. Bulbs with heavy filaments (high-wattage bults and high-current, low-voltage bulbs) have less ripple.

 

[JonSea]

Would anyone else like to line up and kick me? Geez.

 

[crutschow]

Would anyone else like to line up and kick me? Geez.

Naw, It was fun but I think we're finished.

 

[ronsimpson]

I think a UV light detector might give you information. Maybe a far IR detector. What I want is to know what percent of the light is UV.