Choice of power UV leds for a "multicolor" UV flashlight set

I have always wanted to have a set of small UV lights, say, 365, 385, 395, and 405nm. The Jet-U which Clemence has used as a host for his mods looks good to me.

What I need to pick are the quality LEDs to fit the xpg mcpcb. They dont need to be really powerful as the host provides just like 200ma, but there should be as little non-uv light as possible, and within $10 per led.

I do have that black zwb2 glass transparent for the 365nm, and may try zwb3 for longer wavelength but it would still filter out most of the 395/405 light

Would appreciate you suggestions

I don’t think this will really work out.

Many UV LEDs emit a whole spectrum of UV 395-365 and anything below about 400 ish shouldn’t be visible and therefore you’d just have a collection of barely visible/invisible lights which you were told are different.

You can get a set cheap on Aliexpress TMWT ultraviolet led 365nm 380nm 395nm fluorescence detector cheap uv penlight flashlight. Get a few and let us know how they perform. I don’t expect much out of these but who knows.

As far as the filter, you just need LEDs with low visible light emission. I have not seen much data on this except comments by djozz in his testing. So far the Nichia 365nm appear to have the least visible light emission. I expect inexpensive leds at longer wavelengths to have increasing amounts of visible light, but just my prejudice. I haven’t been able to find other filters tailored for wavelengths other than 365nm.

I also had found these pens and they have been in my cart for a couple of days. But I dont expect too much either. I think they are direct drive 5mm LEDs and I had a flashlight with a combination of 365/390nm 5mm leds. I think it was called Streamlight nightcom. Nothing to write home about. But these pens are so inexpensive that I think I’ll buy them.

Re filters. While the zwb2 is a dedicated 356mm narrow band filter, the zwb3 has a much broader range. But on the other hand, djozz found that those higher wl leds did not have that much of non-uv light, so no need for the filters. I may just settle on a clear lens that is transparent to uv, eventually

What do you want to look at with the longer wavelength emitters? For mineral fluorescence, the longer wavelengths give much less fluorescence, or at least I haven’t seen any minerals that fluoresce better at longer wavelengths. I did find this reference on 395nm fluorescence with scorpions. I don’t think that is a concern of yours, but I may be wrong. I am always trying to keep up on fluorescence topics.

Here is an example. I have a more detailed PDF document, but can’t attach it here

Djozz tested some UV emitters, you might find useful.

I use a Seoul Viosys 365nm with ZWB2 filter in my DQG Hobi pocket carry.


You almost always get some stray light from these emitters so appropriate filters are a good idea.

Thanks, CRX for the photos. I really wanted to make use of one of my DQG Spies but just ruined the LE trying to reflow an optisolis LED. So would keep the other one as it is.

Now, I found MecArmy PS14 Dual Color Temperature EDC AA/14500 Flashlight, which is quite far from being a budget one, but it has an ideal UI for my purpose, if I want to have, say, 365nm and 385nm in one flashlight. On the other hand, I have no idea if it is moddable at all. I haven’t found ane mods on this forum or elsewhere.

I have ordered these UV penlights, in all three UV flavors, as well as some zwb2 and zwb3 filters. I have a 365nm light with zwb2 filter so I know what to expect from it, but I am really keen to see how the zwb2 works for different wavelengths.

There is also a KDLITKER P6-TRI CREE XP-E2 TRIPLE COLOR P60 LED DROP-IN , where the original color LEDs could be swapped for UV LEDs. I am not sure at the moment, which of good UV LEDs have the XP-E footprint, also this is going to be another step up in size, even if it is just a 3p body.

You only need a ZWB2 filter if the LED produced white waste light which is interfering with UV output

Nichia LEDs are known to produce almost no visible light, just a bit of >380nm on the edge of spectrum is visible, a ZWB2 filter won’t help much here while reducing UV output quite a bit

if you want really invisible UV light you need a really good low bandwidth UV laser mirror that simply lets the unwanted light through building an angle light

if you look on filters they always got a wide range, so to pass at 365nm 80% they also pass like 10% of light you can see

a really expensive filter would blast budget but they are there

https://www.edmundoptics.de/document/download/359803

For a 365nm LED you are certainly right, zwb2 filter covers just the right band. What I want to experiment with are LEDs in 380-405nm, and here the broader band of the zwb3 may help, or not.

One particular case where it could is if I decide to stick to 5mm uv LEDs.

My opinion is that the longer the wavelength, the less fluorescence you will see. And there will be more problems with visible light emission requiring an expensive tailored filter as Lexel has shown. The longer wavelengths seem to be better for fluorescence of organic molecules, like chlorophyll, so if you are looking at that it could be interesting. I want short wavelengths for minerals.

The problem of white-ish stray light seems typical for 365nm leds in my limited experience. The 385nm leds from LG don’t have that, they only emit a narrow band around 385nm. Also going further up to 400nm you have no extra emission bands. So a filter is only needed for 365nm.

If I eventually decide to put a 365nm and a 380/385nm leds in a single host like that from mecarmy, so that they are powered by the same driver, what would be your suggestion regarding the choice of particular leds so that

  1. They have similar output at a given driver level
  2. They have the same footprint.

And btw, what kind of power do I need in a mule for:

  1. Note/document detection from say 20cm
  2. Dirt/urine detection from 2m?

Thank you!

I guess that 500 mW output gives good illumination and is reasonable, that is well under 1A drive current for most UV leds.

I have no specific led recommendation apart from that the 3535 footprint is convenient (which rules out Nichia), UV leds from all kinds of manufacturers perform well. Perhaps LG for 365nm and 385nm, and if you can find it 400nm. But the SST-10 UV leds are good too.

I stand corrected. Interesting to hear that Jos.

SST-10 has a fairly high forward voltage as well as the Nichia NVSU233B
they are at about 3.8-4.0V at max current from specs

https://www.nichia.co.jp/de/product/uvled.html#NVSU233B

Nichia NCSU276A and NCSU276C have a very uncommon footprint hard to get well performing MCPCBs for it, MOSLED may be best option or virence MCPCB with proper insulation

Found this, on aliexpress. They claim to have all wavelengths on all board sizes. But the leds look different from what I see in the djozz’s threads, dont they?

LG 3535 UV LEDs

I just found an amazing reference paper dealing with food quality testing with a professional 365nm UV system. It is in russian, unfortunately, but I am sure similar ones are available in every major language since it is a reference manual for a standard food safety camera.

It says for example, that natural butter glows yellow and any hint of blue glow means that margarine had been added. Different plant oils glow different colors. And none of them should dlow deep red, while my current Spanish oil does. Meat glows differently depending on how fresh it is.

So there seems to be A LOT more than note/docs testing

That’s interesting. Now I really want to shine some UV at my food and see what happens! :sunglasses:

Peas are bright red :party: