I was walking through my local camping supply store the other day when I came across this water purifier that cost $50 (I know outrageous) Being the DIYer I am I took a closer look at it to find that it was simply Some UV lights. Then it hit me, I can make this way cheaper. So I did.

HINT: start your soldering iron and hot glue gun now so they warm up
Remove these adsRemove these ads by Signing Up

Step 1: What you will need

Here is a list of all the tools and parts you will need.


- some ultraviolet LEDs, I used 4

-A small switch

-One resistor for each LED, Mine were 330 ohm

-Black tape/ Electrical tape

-Some wire

-Small plastic or glass container, This should be watertight and clear

-A battery, Mine was a 7.2 volt RC car battery


-Solder Iron


-Wire cutters/ strippers

-Hot glue gun with hot glue


I had some stuff but I had to buy the LEDs ($2 Each) and the switch ($1 for a pack of 5)


This is from the label on the bag that the UV LEDs came in.

"This LED produces intense ultraviolet light. Exposure to UV radiation can be harmful. Protect your eyes and skin when operating. Never look directly at the LED. Buyer assumes all risks using this LED."

Just so you know and don't try and hold me responsible either.
hai.tay.3512 months ago

put your water bottle in the sun it will kill some bacteria

Wrong uv spectrum
haeshproject9 months ago

There are no UV leds in the right spectrum with enough power for the right price. Wait 10 years and check back.

Please i do not want to be rude or anything, but you could not be more wrong, the wavelength for 254nm +-10% are around and many people have posted sites that sell them on this page one post was 2 years ago, sure if your looking for these UV-C LED's on eBay they are few and far between on eBay and some of the Cree range provide very high powered LED's that are classed as Daylight.

So in short they are out there just few and far between and there are high power LED's out there.

I think the correct thing to say would be right now the LED's for UV-C filters have about the same power consumption as the Fluorescent counterparts to filter water in order for them to work as a UV filter, but the advantage of LED compared to UV fluorescent would be the lifespan and durability of the product.

Taking price into account, at a glance could look similar or more for LED UV-C but add to the fact you would be paying for a device the lasts longer and durable it should be cheaper in the long run, but running cost i do agree it's about the same for UV-C as the efficiency is not there yet in LED's for this spectrum.

As regards to spectrum i mean UVC. In regards to power i mean a 3 watt cree bulb. I buy high power leds and i have a cree UVC led 3 watt bulb and output is not even in the same order of magnitude as my florescent bulbs. Also even high power leds barely can penetrate water more than a few mm with led and maybe a few inchs with off the shelf flourcent. For industrial applications like waste water treatment there are ones that work for 3 foot ditches. An led bulb capeable of sterilizing water is not within the technical capabilities of mankind at this time,. If that was powerful enough to kill germs you would destroy that plastic bottle which is not uv resistant, hence why they use quartz in real sterilizers.

did i say it was to be used for industrial applications: NO

Did i say that the efficiency is not there yet in LED for this Spectrum: yes

did i say that an LED of any wattage works at 3 foot: no

Have i used UV LEDS / UV lamps ect YES.

the fact is i did not indicate any of those values, you just assumed or added them in.

your words "As regards to spectrum i mean UVC." 100nm to 280nm

last time i checked 254nm is with in UVC sure its towards the UVB end of the scale but if you know anything about killing bacteria then you would know that 254nm is the wavelength does the damage.


Germicidal UV is delivered by a mercury-vapor lamp that emits UV at the germicidal wavelength. Mercury vapour emits at 254 nm. Many germicidal UV bulbs use special ballasts
to regulate electrical current flow to the bulbs, similar to those
needed for fluorescent lights. In some cases, UVGI electrodeless lamps
can be energised with microwaves, giving very long stable life and other
advantages[clarification needed]. This is known as 'Microwave UV.'

Lamps are either amalgam or medium pressure lamps. Each type has specific strengths and weaknesses

there are LEDS that have the wavelength of 254nm...............They are few and far between and you will not find them on stated.

I will not post a link to these LED's as others have already done that.

your words "In regards to power i mean a 3 watt cree bulb. I buy high power leds and
i have a cree UVC led 3 watt bulb and output is not even in the same
order of magnitude as my florescent bulbs."

i see where your getting this idea from your thinking about normal LED lights in your home are rated at 5watts and there = 50 watts personally there more like 60 watts with good ones, but as i Stated, UV LEDS are not efficient as there other counter parts, but they do have the advantage of been solid state and durability, last thing bud this is just a note for you in future the term bulb should not be used for LED as LED stands for Light Emitting Diode, the only relation to a bulb they have is they emit light.

so to respond to your question in order to get the same power based on the current technology you be very close to the same 1 per watt as stated.

Sorry if i was not clear enough on that if you like i can get you the exact figures but i am very busy today.

Last LEDS are already in many certified filters.

PS i have access to a microscope if you would like to see the results of LED UV light to bacteria i am more then happy to send a video.

LEDS definitely stand a chance to disrupt the market soon. There are LED disinfection units, and they either dont work well with any water fouling (such as tannins) or they are insanely expensive, but most likely both. For space application they are ideal which is why I was using them. However we resorted to mercury based due to crap performance and cost. I think its very important point you made that even though my LED supplier says it rated at 3 watts that really only 2.8 watt draw and the the UVC bulbs are only a few percent efficient as opposed to orange/red which has a much higher lumens per watt. Usually only 10s of mW of UV from 3 watt input.

Heres a LED purifier source.

nanoBorg889 months ago
Thanks for making this instructable and starting this discussion.
A good method in checking if your producing the right wavelength of light; using the photoelectric effect!
I have access to a gold leaf electroscope and a piece of zinc metal. The near uv spectrum will do nothing to zinc, but anything with a wavelength shorter than about 350 (half way through uva band) will cause electrons to emitted from its surface. On a charged electroscope the leaves will drop as it is discharged.
If you wanted to be surer that you were in the uvc range you will need to use a different metal. Choose one from the list that has an threshold greater than 4.43ev
Looking through the list the more common ones would be aluminium or tin

Btw thought it might be a good idea to remind people that there's very little to none uvc in sunlight because it's absorbed in the atmosphere.

Hope this helps someone!
Blackice5049 months ago

Hi, mate i have spent too much time researching this subject as i want to make a UV Filter to be inline after my Eheim 2226 filter.

Some things i have mentioned in a reply on your page about using UV-C LEDS and i will try not to repeat my self here.

So about your project UV-Filters do work so you are on the right train of thinking how ever with out a pre-filter you device may not work as well as you think because all water has particle matter in it unless been filtered and distilled and even then some can be in the water and that's where UV Filters come in but only after the first process of water filtration has been done as UV is light and can only pass though some objects and even some objects that appear to be transparent may not allow some spectrum to pass through especially in the UV range even water will diffuse UV light and the deeper the body of water the more UV light defuses thus changing the wavelength of UV light making the filter less effective this is why need lots of power to overcome this, so that $50 dollar purifier may not be as outrages as you think, in fact most of the UV Filters i have looked at for my fish are around that price range for 36 watt of UV-C killing power ;) so then add on the cost of a batt and maybe a charger i would think it should cost more if it actually kills bacteria simply because its the same price as the ones that run on mains.

Also i really wound not 4 LED's would do the same job unless you leave it on over time and give it a shake every so often, so please health sake do not drink the water from that device unless you got it from a tap and have tested the time it takes to kill bacteria.

But great job on making your device and getting peoples minds thinking about the subject.

A fun thing you can do with UV with your device is to put old cooking oil and dish washing liquid in and having a really cool glow lamp.

moradofox1 year ago
nrepak, are you sure you get UV-C LED (250-285nm) at $2 each? From your chamber, I see too much visible light. someone's comment is right, it looks UV-A LED(365-395nm), in this wavelength section, the germicidal efficiency is about 1/3,000 of UV-C.
LoganMackey3 years ago
found manufacturer uv diodes with 240nM wavelength

do u get it uv leds newhere in India?
You'd need to buy 1 million of these before they would be cheap enough to be worth it from these people.
TGGrimler2 years ago
Ebay sells something called the P-light, which is supposed to disinfect your toilet at night. $8+/each plus crazy shipping from Australia.

On another note, there are multiple brands of water bottle caps that are meant to convert your water bottle into a camp lantern. They are solar powered and use LED bulbs. If you took them apart you could possibly install the appropriate uv LED bulb. If you replace only a few of the bulbs with UV and have a few with normal white LED and then use colored bottle to filter out the harmful light, then you should be able to have a water bottle, camp lantern, and water purifier in one. Unfortunately that increases cost.
brentonl5 years ago
So, you claim that the water will be 99.99% bacteria free.  Have you tested that hypothesis with bacterial cultures?
Surely the LEDs do not need to actually be IN the water. So long as the container doesn't filter UV light you could surely put the UV LEDs OUTSIDE the container.

Would be interesting to test this with proper cultures on petri dishes. Shame I don't have a lab / money / the scientific experience to do it...
*drum roll*
:( May not have been the case when you posted that link but it's now showing as 'out of stock'
gcc_mtl3 years ago
Your device does not kill bacteria:

1 - UVC is germicidal, thats between 100 and 280nM
2 - most plastics and glass filter out UV light

You need to use Clear poly-styrene plastic or cobalt glass for UVC to pass through, or stick the bulbs into the water without anything infront of them.

elting3 years ago
i made one with a solar panel and baby food jar.i put it into my small platic pond and within 2 days the amount of green crap was cut in half.i wish i would of looked here first befor i went of the trouble of figureing it out myself.well thats the fun in it any way.nice diy
Strycn1n33 years ago
it kills the bacteria sure but some water would still taste like S***
I think heat shrink tubing would be easier to put on. Dab each end of tubing with hot glue if worried about water penetration.
ego733 years ago
If it does indeed nuke 99.99%, how awesome would it be to rig it for solar/rechargeable battery operation! Then rig it, all nice n compact, onto a cap for a 2 gallon plastic container. Cool!
I can't find anything below 350nm! I've been looking at combining a UV-C water treatment (including sand filter) with DIY PV solar panels for a project in Cameroon (remote schools).
I understand UV-C LEDS will cost more but would like to get hold of some to develope the idea with a view that the price will come down eventually. Anyone know where i can get some?
I wouldn't have thought the 400nm LEDS used in this design actually have any effect? Maybe someone could test using a coliform presence test? fairly cheap to buy. I got some for £3.50 ($7?).

varun20214 years ago
could you tell me how much current each of the uv led was consuming?
findoggydog4 years ago
In some, if not all cases the UV light will be neutralised by the small container housing the light, therefore rendering it useless for the purpose outlined.

The danger of that is you will think your water is being purified, when it is actually no different than when you put it in the bottle.

But unless you have a bio-lab at hand there will be no way of knowing.

Personally I'm going to stick with the store-bought purifiers as they get thoroughly tested - I don't want gastro in the middle of no where.
n33r5 years ago
cant u solder em in series connection it would much easier to handle
Pie Ninja n33r4 years ago
You're silly.
Nyxius5 years ago
Where did you get your UV diodes from?
nrepak (author)  Nyxius5 years ago
electric goldmine for the 400s look around for some 240nm ones they will set you back but work a lot better try an auction site to get them cheaper
Nyxius nrepak4 years ago
Thanks man!
vatak5 years ago
UV-c LEDs are the best for this application.
jigoku13 vatak5 years ago
 where can i buy those uv c LED's?? is there any site exist that sell those kind of diodes????
jigoku135 years ago
where did you find your uv diodes with 240nM wavelength? what site can i find one of those diodes?  pls suggest some. . .
mossimo35 years ago
This is an interesting concept, however only UV-C is effective in killing harmful bacteria. But the down-side is UV-C is harmful to your eyes and skin; it can give your skin and eyes, yes EYES too, a nasty burn. So if you manage to find UV-C LEDs, you should enclose them in an opaque container--As should be done with any UV LEDs. Other than that this is a neat idea! Have you tried adding an external power supply or something to have 'unlimited' power for however long you want to purify your water?
manutea5 years ago
Nice job-thank you !
xiuxiubaby5 years ago
hoho it's so cool
phenoptix5 years ago
What wavelength were the UV LEDs? I heard about this as an application for low wavelength UV LEDs, I wouldn't think the UV LEDs I sell that are around 400nm wavelength would be powerful enough (too long a wavelength) to cause the bacteria any harm - which is a shame!
nrepak (author)  phenoptix5 years ago
Ahh good point. I bought two different packages one had 240nM and the other had 450 nm but I mixed them up and am not sure which ones I used. Thanks for reminding me
the 240nm are the real bad boys! how much did they set you back? The 450 would be a blue colour
I remember looking this up once _after_ ordering a bag of 12 cheap UV leds for $5. Worthless for sanitation, pretty cool for other things (I'm building a ring light out of mine now, at least it'll look cool) The low nm uv leds are in the range of $50 if you're lucky. That said you can find gemacidal uv lamps online for under $20, some of them under $10! Sure the driver circuitry will be different, but you can still do this very cheaply.
crc095 years ago
this is incredable, great for camping or survival
Wolf Seril5 years ago
If I'm stranded in the woods near some suspect-looking water and all I have is several UV LEDs, solder, a power source, and some plastic containers, I'll be sure to keep this in mind. Just kidding, nice instructable.
vapoking5 years ago
Isn't the wavelength that fries bugs bad to look at? a book on constructing a mushroom lab advised that if you use uv for sterilization you should set up the room so that the lights go off when the door opens.
nrepak (author)  vapoking5 years ago
any UV light is bad to look directly at and not rie about the shroom lab lol
nrepak (author)  nrepak5 years ago
Woops typo thats sposed to say and not sure about the shroom lab
i dont think thats the right UV light for the job.. like vatak says, UV-c is the best, and i believe this LED mostly puts out UV-a ..... but good idea, dont think i would need to build one of these, i did use an UV LED to make some water glow after i dumped a broken highlighter into it. i just have the waterbottle sitting on my desk and i turn on the UV LED when friends are over or sometimes at night.
dark sponge5 years ago
Judging by the color, this isn't the right frequency. This is about the color af my 390nm UV leds. You need a much higher frequency light to kill bacteria and stuff. And even if this was the right frequency, this wouldn't filter out toxins in the water. At least it makes the water glow a cool color!
Mr. Grumpus5 years ago
Now if we can get some spec's on the output of the LED's in the Store purifier to compare with the homegrown one, we'll be in bidness.

Feelin some "Blade Runner" kickin in....

..."Enhance"...String of UV LED's in PVC or Clear pipe of similar diameter with regulated water flow to allow for 30-60 sec of exposure to UV light....

..."Enhance"...Utilize Solar Panels or Wind Generators for power supply...

..."Enhance"...Add rain fed cistern, some sort of pre-filter to clear out the chunkies, use gravity to power most of the water flow, and some not-so-subtle electronic wizardry to turn the system intake water valve "OFF" when power from batteries to UV LED's drops below a certain limit, leaving enough lag time to purify that which is left in the system.

Viola...or, small cello...You now have a higher tech, more pure, more nature-friendly* system of water purification/reclamation.