UV Phone Sanitizer Shows The Power Of Modern DIY

Editor’s Update: According to the schematic for this project, SST-10-UV-A130-F405-00 (PDF) LEDs are used which produce 405nm UV-A light. The manufacturer, Luminus, does not recommend that part for disinfection or sterilization. Luminus sells UV-C LEDs for that purpose, generating 275-285nm. After publication the part number used was changed to and American Opto L933-UV265-2-20 which is a UV-C LED producing 265-278nm.

The global COVID-19 pandemic has had a serious impact on the hacking and making scene, though it hasn’t been all bad. Sure, shipping on average is taking a lot longer than we’d like when ordering parts, but otherwise being stuck at home has given many people far more time to work on their projects than they would have had otherwise. In some cases, it’s also been a reminder of just how far we’ve come in terms of what the dedicated individual is capable of producing within the confines of their own home.

As a perfect example, take a look at this UV sanitizer box built by [Md Raz]. Looking for a way to quickly and easily kill germs on smartphones and other small devices, he used the considerable capabilities afforded to the modern hacker to produce a professional-looking device in far less time than it would have if he had to outsource things like PCB manufacturing or injection molding.

Inside the 3D printed enclosure is an array of SMD UV-C LEDs that, according to the manufacturer’s specs, will destroy viruses and bacteria in 5 minutes. To make sure the LEDs are given enough time to do their job, [Md] is using an ATtiny85 to control the countdown and a seven segment display to let the user know how much longer they have to wait. All the electronics are held on PCBs produced with a BotFactory SV2 desktop PCB printer, but for those of us with somewhat more limited budgets, a mill or even a modified laser engraver could be used to produce similar boards.

With everything going on, there’s understandably been increased demand for germicidal lights. But unfortunately, some unscrupulous manufacturers are trying to take advantage of the situation. Being able to select the LEDs for this device based on their specifications is arguably just as important as how quickly it was produced. Though we’d still advise a position of “trust, but verify” when it comes to UV-C.

93 thoughts on “UV Phone Sanitizer Shows The Power Of Modern DIY

        1. Thanks for the feedback. Currently, we use the following LEDs which according to the manufacturer emit around 272nm (you can see the spectral power distribution on page 5 of the datasheet https://www.digikey.com/product-detail/en/american-opto-plus-led/L933-UV265-2-20/2460-L933-UV265-2-20TR-ND/12180111 ). Of course, you should always test and verify that your components are behaving as expected, especially for an application like this.

    1. Note the traces are such that both diodes are an equal distance from the power supply, insuring that they turn on at the exact same time. If you don’t do that, they could be out of phase, and they might cancel each other out!

      1. Nah. They are omnidirectional lights so they are going to create an interference pattern of constructive and destructive phases no matter what you do. Timing isn’t really important here.

        Like this: https://www.simply.science/images/content/physics/waves_optics/interference_diffraction/Concept_map/interference1.gif

        Changing the phase of one of the light sources will change the interference pattern, but it will still result in that characteristic field of high and low intensity.

          1. forty-2 – no, recombination of electron hole pairs in AlGaN LEDs is stochastic in nature, the (average) radiative lifetime at room temperature should be in the order of 1 ns (1E-9s) and the emission itself happens on an even faster timescale (~fs). A 300 nm photon has a frequency of ~1000 THz (period duration 1E-15 s). So neglecting the inherent randomness in photon energy, there is no chance getting two individual photon sources producing spontaneous emission to produce an interference pattern, even with the most precise means of electrical external control.

            RW ver 0.0.1 – LEDs produce narrow-band emission which has everything to do with the densities of states in the semiconductor over energy and nothing to do with a natural line width which is substantially smaller than that. At higher current densities, the peak wavelength shifts due to band filling, and it also shifts with the lattice temperature. There is no characteristic “wavelength”, not even in laser diodes – these can be tuned using feedback, temperature and current.

            qwert – as mentioned above it is practically impossible to get two light sources to produce interference patterns. Even optical heterodyne detection requires spatial coherence and requires photons to have almost the same wavelength (see https://www.rp-photonics.com/optical_heterodyne_detection.html). The picture you linked shows the wave picture of two coherent sources, for example corresponding to a double slit hit by a plane wave from below. This case does not apply here, as I’ve described above. Photons coming from a single source can interfere with themselves and create interference patterns, but for that to happen, they must pass through a beam splitter or one or more slits.

            Alan H – Continuing with the above, on the target or detector, the optical path lengths for each path must be almost identical. Look at optical coherence tomography for example, where interference disappears within micrometers of optical path length mismatch (see https://en.wikipedia.org/wiki/Optical_coherence_tomography).

            YGDES unfortunately nailed it – nothing is right about the setup. By visible-light LED technology standards, UV-C emitters are effectively broken, with around 2 to 10 mW of optical output power at 1 to 2 W input. They require proper heatsinking, but then again they look to be behind a slab of float glass that absorbs everything below 300 nm. As others have pointed out, these LEDs might not even emit around 265-280 nm, so all is lost anyway.

            Y’all have a nice day.

      2. It’s especially commendable for the track that is common to the LEDs and the power pin. Lesser mortals would haven given in to the temptation to just run a straight track through all three of them. And then the electrons at the far LED would be under significantly less electron pressure than at the near LED; I just tested that with my garden hose water analog. Md Raz did the right thing and made a cut and routed a dedicated track to the middle of both LEDs to equalize their pressure. Hey, if it works for water, it works for electrons as well, right?

    2. Perfeshnuls done did it too, seen it in several laptops where they have a PCB for a couple of indicator LEDs or PCB for power button and one indicator…. worst thing is, some junior designer prolly got a bonus by reducing part count from 2 PCBs to one…

    3. idk whatever. Corona hacks are usually pointless gestures, but so was the quarantine after all. All of that was for nothing, since the state never decided to provide adequate support for people who aren’t independently wealthy to weather the storm. So it’s effectively over way too soon and the spike is gonna happen just the same. Sanitizing your phone and making touch-free braille isn’t going to save anybody. It’s all a performative politicized thing, because that’s what we do with every aspect of culture now. It’s just part of the spectacle.

      I think people should just stop pretending already. It doesn’t help no matter what your opinions or positions on things are.

      1. The idea that quarantining an infectious disease doesn’t save lives is false and dangerous.

        China, Korea, Singapore did better than any western nation, because they’d been through this drill before. France, Italy, and Spain underestimated the disease in the beginning, but reacted quickly thereafter and got things under control. Germany quarantined rather early, partly because of the Italian experience, and with a complete shutdown in the South where I live. We walked away with 8,000 dead in a country of 80M.

        The US, with around 2M cases and 120k deaths on 320M population, demonstrates how a poorly coordinated response can play out in a rich, large country. It would be a lot worse if the US healthcare system weren’t relatively strong. The UK, the same but in a small country. Sweden, which took the “herd immunity” strategy, adds twice as many cases per day as Germany, with 1/8th the population, and shows no sign of slowing down.

        Look at the daily incidence curves internationally. The sharpness of the decline correlates very well with how strict the various contries quarantine and testing regimes were/are. The US and UK are recovering slowly because they are following weak/mixed quarantining strategies, and the cost in lives will be real. And to make matters worse, it will delay a return to “normalcy” (whatever that will be in the future).


        Brazil, Russia, and India look like they’re set to fall off a cliff as of this writing. I hope they do not listen to internet trolls.

        This hack is about UV LEDs. One should use a mercury tube instead.

    1. Agreed. The same thing would apply to the actual LEDs: real UV LEDs that produce the kind of killer UV we need are eye-wateringly expensive, because the window is quartz. Thats why all the current functional sterilizers are using tubes. I wonder how much of the new battery powered “UV steriliser” products out there actually have an affect.

      1. We just used a random piece of glass we had lying around to keep our stuff off the PCBs but you make a good point. Instead of using a transparent (to, hopefully, UV-C) material, you could use standoffs to keep your device off of the PCBs. And you should test your transparent material to ensure that it does transmit UV-C.

    1. Ya, I was thinking the same, that the UV would damage the plastic.
      I’m not entirely sure a phone chassis would respond well to frequent alcohol wipes though either.
      And what effect does all that UV have on the display?

    1. most modern phones are waterproof, just wash it in the sink with hand soap and dry with a towel. If it’s good enough for your hands, it’s good enough for your phone.

      1. Don’t forget that IP ratings are tested on a brand new device. A device that’s several years old, has had a few decent drops, worn out rubber around the buttons etc, will not stand up to the same as a new device. And water jets under pressure (from a tap) is a different thing than leaving it submerged. Most phones also claim proofing against regular tap water only, not salt or soapy water. Some of Samsung’s warnings: https://www.samsung.com/sg/support/mobile-devices/is-samsung-galaxy-s9-and-s9-plus-waterproof/

        “- Do not expose the device to water moving with force, such as water running from a tap, ocean waves or waterfalls”
        “- If the device is exposed to any liquid other than freshwater (salt water, ionised water etc), gently rinse it with still fresh water immediately and dry it thoroughly with a clean soft cloth. Otherwise, salt may accumulate and block primary earpiece and external speaker when it dries”
        “- If the device is exposed to any liquid other than freshwater (salt water, ionised water etc), gently rinse it with still fresh water immediately and dry it thoroughly with a clean soft cloth. Otherwise, salt may accumulate and block primary earpiece and external speaker when it dries”

        At the very least I would not press the buttons while washing it and some time afterwards.

        Personally I just use an alcohol wipe to wipe it down. Smooth surfaces are easy to clean and the oleophobic coatings make it even harder for virus-containing droplets to stick to it.

        1. Sorry the last quote was doubled up, I was trying to paste this one:

          “- If the device is dropped or receives an impact, the water and dust-resistant features may be damaged. Physical damage to your phone, such as chips and cracks can significantly reduce its ability to resist dust and water”

          1. If your phone has cracks that render it non-waterproof then no amount of cleaning will remove the stuff that has accumulated within the cracks and it will be a permanent home for microbes. Ditch the phone and buy a new cheap one. Seriously is your life worth a phone???

          2. @X:

            Coronavirus isn’t a microbe. But yeah I wouldn’t use a phone with broken glass as I’d be too worried I’d cut myself.

            But even a good drop without visible damage can compromise the waterproofing of the phone. The display is only glued on with a thin line of glue around the edges in most cases, and it doesn’t take too much to damage this. In some teardowns they use a cardboard playing card to cut through it.

            The waterproofing of phones is meant to protect against accidents, it’s not really meant for daily submersion. Similar to “10M waterproof” watches which you can’t actually shower with.

        2. I used waterproof cameras quite a lot when the kids were small – normal got about 2 years out of them before they would die in the pool… I think the same would be true for phone – no way would I be putting a 2 year old phone fully in water..

  1. Real UV-C LEDs look like tiny EPROM chips: ceramic case with quartz window, and chip inside. No plastics, as these would perish. Also glow isn’t very visible, you can smell it though, smells like burnt skin :>

  2. I hear there is alot of fake uvc leds about. Claiming they kill all bacteria but don’t because of them being fake. May I ask how do you know that the leds are the real deal. I thought they still cost alot of money.

    1. Some fake UVC leds emit a purple light, and those are easily spotted as fake. Others emit an icy blue light, similar to the light emitted by UVC bulbs, but those may only be “light blue” LEDs.

      The “lens” on a real UVC led will be made of quartz, so an LED with a “lens” that isn’t thin and flat is probably fake.

  3. Interesting thought, if you slice a piece of chicken thin enough, so the UVC can penetrate through the entire slice, then you could kill all the salmonella bacteria in the slice and have safe to eat chicken sushi (if the heat in the box didn’t cook the chicken that is.)

  4. There is not enough information provided to determine if the design is viable, and to get enough information you must surrender your Email address – which is NOT gonna happen in my case.

  5. 0 dollar solution, save enough string off the potato bags until you can dangle the phone off a first floor window to the ground, then sneak into a gas station and dangle it in the E85 tank whenever you need it cleaned…. ;-)

      1. I literally just pulled a full rom dump off one of those!

        A year or two ago I wrote a quick arduino program that continuously read the eprom and compared it with the previous read, and displayed only the erased bits. Ran that while exposing it to a cheap UV flashlight. After 5 minutes only about 4 bytes had changed…

  6. Tinfoil hat time: What if…. we shouldn’t be supporting the LED industry because LEDs caused covid-19 to become a problem…

    So there’s an association of higher covid virulence in persons who are short on vitamin D, how do we produce D? Sunlight or UV exposure. Also while UVC is highly antiviral, UVA is still weakly so. Now, what might cause us as a population to be shorter on vitamin D that in previous years, reduced UV exposure? Now what if I told you that fluorescent strip lamps and curly bulbs put out small but measurable amounts of UVA continuously? I wonder what could have happened to that auxiliary source of UV exposure and thus vitamin D supplementation, oh that’s right, we swapped them all out for LEDs that don’t put out any UV at all. Also, while weakly antiviral, if you had lights on at home for hours and hours, that might well reduce virus viability time to half when continually dosed with weak UVA, vs getting off completely scott free with LED light.

    So obviously big LED is behind all this, they don’t want you to know!!!1111 ;-)

    C’mon, it stands up better than the 5G crap :-D

    1. But the 5G thing is perfectly reasonable! 5G causes the virus – even if there is no 5g in the country at the time! It’s a know fact! It’s the reason we have a new virus – 4G caused swine flue etc, 3G caused …

      By the way, I have a great deal on a bridge I’d like to sell you..

    1. Reading the spec sheet, it says each one has up to 100mW. That does’nt seem like much power. So I would assume that if it does disinfect, the opbject would have to be inside the unit for an extended period of time.

      1. Let’s be generous and say that 5 mins is enough to sterilise a circle the size of a quarter right next to the LED.

        Goes back to what I was saying about UV sterilisation for air handling a couple of months ago, if it does it in less than half an hour or handles more than a cubic foot at once, you can tell if it actually has enough power to work because it will blow a regular 15A breaker as soon as you plug it in.

      2. The Wikipedia article on germicidal irradiation says a 90% kill of bacteria and virii can be achieved with 8 mW-s per sq cm.

        So, that 100 mW source could disinfect (not sterilize) 12 square cm per second, or for the metric-averse: a square foot in less than a minute and a half.

        1. Which is spurious, 8mW inactivates the big clumsy stuff, giardia, e-coli, like shooting bunnies in a barrel with a shotgun, but adenovirus for example will take 120mW square centimetre. Then the LEDs are rated at 100mW by power consumption not UV output and are less efficient than discharge lamps, so if you’re getting 10mW out of them in UV you’re doing good.

          1. Also I do not see any notes about careful selection of the transparent stage material the phone is sitting on, which if any of the usual suspects perspex, lexan, glass etc, is probably 100% opaque to UVC so the bottom of the phone gets nothing at all.

          2. Perhaps if you used the units correctly or read the datasheet you might have some credibility.

            Dose is measured by energy per unit area (watt-seconds per square centimeter), not total power (watts) and certainly not a power-area product (“mW square centimetre”).

            The device datasheets quote actual output power (flux) in milliwatts (in addition to input current & voltage).

            The Luminous Devices ones that appear to be used here are rated 875-1015 mW output flux, at an input power of 1.85 watts. They will, of course, promptly melt themselves off those boards with the heat sinking [Md] is (not) providing, so he must clearly be driving much more gently.

            Incidentally, the shortest wavelength available in devices in these packages is 365 nm — short enough for photocatalytic breakdown, but not disinfection.

          3. Thanks for the comments and great discussion. Indeed, our original design used the incorrect UV-C LED. Since our initial post on Instructables, we have updated the bill of materials and the design (but not the pictures :D) to use a more appropriate UV-C LED, which RW ver 0.0.3 has correctly linked to.

        2. https://www.instructables.com/id/UV-Sanitizer/

          That’s not the same thing?

          Well think what you like but I wouldn’t lick anything put in there until it toasted 83 minutes each side, with the glass decontaminated in between, because the glass just blocks the bottom two. Actual irradiance at 1cm is quoted as 0.6 mW per cm^2 because it’s spreading it 120 degrees with that lens. If we take a wild hope that no part of the phone is more than 5cm away and divide that by the square of the distance, the furthest parts are getting 0.024 mW per cm^2 dividing that into the 120 mW-s sq cm needed to be sure of killing virii not just “hey look at me I’m the broad side of a barn” e-coli etc, then you need 5000 seconds for four nines deactivation. But also, don’t lick the edges that you hold all the time, because they’re in shadow, not to mention what’s hiding in the seams and around any buttons.

  7. It’s a nice build and it certainly looks good from a design perspective. But i am having a hard time wether it will reach it’s design objective of killing the bugs. While LEDs are the better choice over traditional lamps in almost every application, this is an area where i would still choose to use fluorescent lighting. For instance Philips has a range of UVC lighting which is used in applications like killing bacteria and fungicide in aquariums and they also provide the bigger lights for application in hospitals https://www.lighting.philips.com/main/products/uv-disinfection

    A simple solution could consist of al Philips TUV PL-S 11W placed inside a corner of a microwave, and then place your smartphone on a stand in the middle of the rotating table. For the best results, don’t disenage the microwave generator ;)

  8. Tried opening the schematic in Eagle. Got the following message that makes me think this is all a joke:

    Loading /UV Sanitizer Source Files/Electronics/UV Sanitizer ATTINYT85.sch …


    line 1, column 1: Start tag expected.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.