Go Down

Topic: UV Light Sensor (Read 19286 times) previous topic - next topic


I am new to arduino and I want to use an arduino board to measure the amount of uv light, specifically the 254 nm range.  I was wondering what would be the best option to do this.  I am thinking of using a simple LDR and using a source LED of 254 nm wavelength.  I will then be able to measure the amount of light present and directly relate to absorbance along that wavelength if I pass a gas through this light path.  I know an LDR'r range of wavelengths does not really include that of the ultraviolet spectrum, but I wondered if some sort of amplifier for the LDR's analog output would help me at all?  I am new to this arduino and fairly new to electronics in general so forgive me if any of the aforementioned is wrong, but any help or suggestions would be greatly appreciated.


Feb 05, 2012, 08:55 am Last Edit: Feb 05, 2012, 09:08 am by focalist Reason: 1
One possible solution is to use the UV LED as your sensor!  LED's actually are quite good as light detectors, and I would think each wavelength of LED emitter would hold true.  I have a bag of 100 of them waiting for a project, pennies on ebay.. Or search out the exact component wavelength and then order the proper one from an electronics supply house.  I've not tried it, but the concept and circuit are simple enough to just breadboard with a couple of off the rack LED's in the right UV range, one emitter the other a frequency matched detector, in terms of wavelength, and give it a go.  

There has been several projects which use LED's as a light sensor, I expect UV would be no different. Photons at the junction, frequency-tuned doping as is needed to EMIT that light as normally used.. It would seem it should work both ways, just like visible range diodes..

Look in the playground if I remember right.. Or just do a search looking for LED as a sensor.  Really cute, cheap, might be just what you need...  in fact, there's a current project in the exhibit section of the forum that uses Infrared LED as the sensor for a poor mans thermal camera.  way cool, and it supports the idea of LED as a frequency selective sensor...


You might have a problem with just picking a random device as sensor - some plastics absorb hard UV heavily (just fluoresce) and wouldn't be very transparent at the relevant wavelengths - so be prepared to try again if one device performs poorly.  Many plastics will degrade if exposed to bright UV for extended periods of time.
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]


UV LED's are made with UV-transparent and tolerant resins, I would suspect- kinda self-defeating otherwise!


Feb 05, 2012, 04:15 pm Last Edit: Feb 05, 2012, 06:36 pm by jhjhjh0 Reason: 1
We used a UV filtered photodiode in a school project years ago.

This is where we got it from: http://www.eoc-inc.com/UV_detectors_silicon_carbide_photodiodes.htm

You can probably find a much cheaper (but less accurate) one on mouser, digikey, ...

Remember sunscreen!


Thanks for all of the responses.  I think a uv specific photodiode and an led that outputs 254 nm wavelength will work, atleast for early on testing, but are either of these available for fairly cheap and if so any recommendations as to where to purchase?


Using UV to measure gases sounds like a cool application.

We used a UV photodiode that responded to ~200-400nm. It was used to take a measurement to display a good estimate of the current UV index. As of 2006-2007, a photodiode that responded to only that window of the spectrum was about $100 USD. That's not to say they have not gone down in price or there are not more options now.

I will keep looking too and keep you posted because I would like to make a new and improved UVI meter.

A couple things I remember that might help during your design phase is that florescent light gives off hardly any UV. Whereas incandescent and, obviously, sunlight do. If you avoid other sources of UV, you might be able to make due with a photodiode with a larger sensitivity range.

Another important factor to keep in mind is that the smaller the window of the spectrum the photodiode responds to, the less current it will produce. On the brightest days, we could only produce measurements in the low nano-amp range. It was difficult (for us, at least) to amplify that.

Please keep us posted on your progress.


Well, for the cost, you aren't going to beat the LED.

I really think you should give this a read:


And you can get a pair of em from pretty much any electronics shop for a buck or two at most, even if they gouge you on the price.  Seems that an LED is just a specifically doped photodiode in the first place-- with sensing windows that are of the emitter frequency or shorter wavelengths.  If you can't tell, I'm interested in the application, and if it can be done with LED's, because I have literally have a bag of 100 5mm UV LED's I got off ebay a couple of months ago about a foot and a half from my hand.  I paid a couple of dollars for the lot.  Now I'll assume these are the cheapest on the planet- but better ones are only slightly more expensive..

Anyways.. interested to see where this goes, no matter what kind of sensor you end up making work!


Have you found a 254nm LED?

According to http://www.maxmax.com/aUVLedLights.htm
"There are no commercially available UV LED's that emit below 365nm, and certainly none at 254nm.  "

Will it still work if you go 365nm or higher?


I actually haven't had much trouble finding 254 nm LED's, they are just nowhere near the 12 cent price that typical LED's are.  All of the ones that I have found have been 200-300 dollars.  I also was also able to find a few LDR's that would measure the specific wavelengths I am after, which are a few in the UV range (254 nm) at about the same price as the LED's.  I am unsure if I want to purchase these or not as I am unsure if this would work.  I want to pass a gas containing different metals in the UV spectrum through multiple different wavelength lights with its specific sensor to quantify the absorbance.  I am not sure if an LED and LDR will provide me with enough sensitivity for me to even generate any range of absorbance concentration.  I am still looking into this more to determine if this is a viable route.  I am also considering using the specific LED as a light sensor as well, but I am also unsure if this will work or not.


Feb 07, 2012, 04:46 am Last Edit: Feb 07, 2012, 05:14 am by focalist Reason: 1
Just thinking on ways this might work.

As long as the sensor works in the frequency range, you won't have to tune the sensor, only the emitter.  That might make it a little easier.

How about this:


This was a home-made phosphor screen I made using Rustoleum glow in the dark paint, quite a while ago.  Still in the basement, and may actually come into play with the UV LED's I got.  It's something I made while messing around UV and black lights for some photo stuff.. but this demo is using a simple single-LED UV "flashlight".  The important point being the VISIBLE reaction to INVISIBLE frequencies of Ultraviolet, not the continued glowing effect.  There's a phosphor in the paint.  Phosphors fluoresce in the VISIBLE range in the presence of UV light.. it's cheap to get a visible-range light detector.  The more I think about this, the more I think using a phosphor to indirectly measure might just work.

Anyway, it seems that detection and quantification of UV light might be expensive DIRECTLY.. but what about if you use a material which will fluoresce in the presence of the UV, then measure the fluorscence?  Absorbance blocks the photons, the indirect measurement might be viable if you can "calibrate" against the material's reactivity to UV.  Even the paint I bought has that info available.  The can of paint (think it was like four ounces or something) cost about $10.  I'm sure you can buy a "calibrated" phosphor (or just calibrate it yourself) that's better than Home Depot glow in the dark paint for relatively reasonable money.  Is there a current real-world device that does this, that you can skim over to see what IT uses?  Kind of like a reverse scintillation tube, except for UV.

Then, all you need is sources for the UV wavelengths, and a single detector.. in the visible range.  By the way, maybe there's UV fluorescent lamps which would be a lot cheaper than than the LED emitter.  MacGuyver UV gas chromatography.  Yeah, I love Arduino.... hehehe

Anyway, Just a thought.  I'm sure there's shortfalls, but I've heard FAR crazier things.. of course most of those are echoes.


There is a sensor called a Uv-tron.  I don't recall mfgr at the moment but Google will.


Using the UV Tron as a detector for 254nm wavelength light sounds promising and I am thinking of pursuing this option.  However, now I am stuck trying to find an LED light source of 254 nm wavelength for somewhat cheap.  I am thinking I may have to use a monochromator to filter out the wavelengths I am interested in from a white LED or something along those lines, but I am unsure how much these cost or if its just more efficient for me to invest in one of these higher priced deep UV LEDs.


I highly doubt that you are going to see any energy at that wavelength from an LED of any sort.
One way to get there might be an electric spark. Nitrogen transitions somewhere in the UV spectrum IIRC.

Go Up