Some experiments to see how a selection of series capacitors affects the power dissipation and output of a 50W near-ultraviolet LED.
The small motor-run capacitors look very promising for reducing the heat dissipation and extending the LED lifespan of these LED modules. Be aware that these capacitors do not have discharge resistors and can deliver a zing, so a couple of 470K resistors in series across the capacitor is a good idea to drop the voltage fairly quickly.
With standard LED floodlights the capacitors could either be mounted internally, or in an external box. This could be an interesting way to prolong the life of external security lights.
All tests were done with a 240V supply. For 120V use the capacitor values will have to be higher.
I Have no way to test the wavelength accuracy.
The gloves I wore in the video are standard half-dipped work gloves. They are not officially rated for live electrical work, but do provide a secondary layer of protection against shock at mains voltage. If working in a wet or outdoor environment take extra safety precautions.
If you enjoy these videos you can help support the channel with a dollar for coffee, cookies and random gadgets for disassembly at:- https://www.bigclive.com/coffee.htm
This also keeps the channel independent of YouTube's algorithm quirks, allowing it to be a bit more dangerous and naughty.
#ElectronicsCreators

I have a purple LED Let's experiment shall we? and do lots of experiments with it. So this came from AliExpress it is branded Yxo Yuen now 220 volt 50 wat. It's probably going to be available in lower voles as well. High Power UV purple LED 395 NM for basically for printer curing this 3D resin curing and I get the feel out.

This is sold for getting existing Uh 50 W LED fixtures and basically taking the original white LED out and putting in the ultraviolet one instead. and the ultraviolet one will I say ultraviolet. it's deep violet at 5 395 NM H it um is probably just the beer chips that nor be used to excite warm white phosphor, but it's a nice color, very nice color. That's one of the reasons I got it.

Now if we take a closer look at the chip and I'll Zoom down this if we take a closer look at that, we can see the incoming power here. We've got a metal oxide veror. we've got a brid re fire. We've got the LEDs Wir as a large series string of parallel pairs that basic down here.

go up again down here, go up and just zigzag all the way until they get to the end. But unlike other uh, high power LEDs that have the simple rectifier and then they've got loads of linear current regulator chips stacked in parallel. This one has a mosfet a 4 N65, a 4K and it's driven by the very mysterious L1015 or 358 K10 and this chip is doing the linear current regulation with a high power mosfet. So you set the current sensing presumably with a couple of resistors under here.

uh and it will then regulate the power through the LEDs I wonder why they're doing that? Um, I Was going to dig this out but I thought I'll just rack it if for do that. let's do the experiments first. If we take a closer look at the chip though, we can see that it uses uh, the flip chips flip chips are LEDs that don't have the wire Bonds on them going to the adjacent pads. They literally have tiny little bare copper pads with the a blob of solder in each and then they place the chips onto the sort of the soda paste and it's the bare LED chip itself.

No case that just the bare chip literally flows on like a component. They're tiny I wonder how they get it so accurate. Or maybe that is why they've got these large areas to reduce accuracy because the chip will kind of Lan itself as it melts. Anyway, what I'm going to do is I am going to bring in the LED and I'm also going to bring in a notepad because I have some capacitors to put in seies with this.

Let's Zoom back out again a bit cuz I'm going to be bringing the anti here is the anti and we'll hook it up to this. and uh, since I'm going to be picking this up and pointing at things I'm going to put a pair of gloves on as an electrical safety measure. Rare, but true. So first of all, I'm going to start by sticking these wires in here.

They're just push in, should push in, they are pushing in and if you want to release them, you, uh, use a gently rounded item and you push down in these while pulling the wire gently and it should just pop out. So if I pop this in here, we're going to be swamped with Ultrav. well near Ultra out light very shortly. I'm going to stuff those in there.
but I am going to put the gloves on just generic plastic dipped gloves because uh, when you're working with electricity, it helps to just have an extra layer of protection Personally I don't feel I need to at this bench during this experiment. but I'm going to do it because I feel it is important to demonstrate and also I would normally recommend that the heat sinks these are mounted on actually have um, a ground on them because these uh circuits are a tiny thin shim of fiberglass. The tip of the finger is missing just for the touch pad. Boop boop boop.

just to get you know stuff done. Uh, other things worth noting: this is a solid state relay heat sink. It's just one that I had to hand and the holes on it for the solid state really are exactly identical. for the diagonal spacing of the LED that's very handy right to what it's about to get.

very piple and I shall plug this in. it's very purple I'm going to Shield myself. Now look how purple it is. it's Mega purple and it says everything's lit up in the fac I've got a hat here just basically to uh to test it with and it's just basically it's it's making everything glow I don't want to look directly into that too much.

Um, I can also feel the heat from it. Uh, 89 Power Factor: 44.9 wat. So let's say 45 wats and we'll fill in the first value 45 watts and then we'll put LEDs in the series and see what we get that does not look bright looking at it. I'm going to point it over there.

Oh, that is making everything in the room glow. Yes, that is very bright, right? Tell you what. I'm ping off now and then we'll put the first capacitor in line and this is going to be a 100 Nar capacitor. I zoomed right out here.

I shall zoom out a little bit more so technically speaking. this is live and that's neutral so tell you what. I'll put them in the correct slot just for the picky people actually. I'll put this one in here and I'll put a capacitor in series.

Let's dig out the capacitors. I've marked them all up a very small 100 nanard capacitor. I'm going to put this one um into a Chinese clone Wigo style connector. Straighten that wire up, shove it in there, click it down.

put it into the 100 Nano capacitor. What do you reckon the power is going to be from this with 100 nanopar in line with it? Are we ready? Let's power up. It's very dim I mean it's not bright at all. It says 6 watt5 power factor? Uh, 0.6 wat.

that's really very low. Still useful. Hold on I'm just going to do a scientific test to going to turn the light off I'm going to take the exposure off I'm going to get the baseball cap and I'm going to point this at it. You know it's not bad.

You know, for for about half a watt, that's really not bad. Is it okay? Tell you what. I'm going to pause a moment Charlie while I bring the light back so you're about to get dazzled. The light One moment please.
the light is back. Uh, let's unplug this. Let's discharge that capacitor. So I didn't get a little tingle off.

itth. What have I got to discharge that with? Hold on. Just give give us a second. do I have a there's a screwdriver.

There's a uh security screwdriver. Not much of a spark off. That good. Having said that, uh, it's intill in circuit with this so that LA to like a discharge resistor and I've disconnected the wrong thing anyway.

So let's put that wire in there and we'll stick the 220 N F in. I Would expect that to be round about one watt perhaps. So here is a 220 n F capacitor. We'll pop that in here.

This is kind of just really experimenting to see what sort of power levels you could get. You know if you wanted to just tame it down a bit because under running these LEDs significantly like say, running at 10 wats instead of the 50 wats, it's going to make it last for ages, but still put out a useful amount of energy. 220 nart, it's visibly brighter already. It's 1 to 7.

Do we say 1.3 Let's say 1.3 watt per factor5 Again, okay, because there is the capacitor limiting it and again I don't think I even need to. Um, you can see the flicker there because it is. Basically, it's not smooth. but I can't see the flicker, but it's put out a decent amount.

it's actually doing. Yeah, that's actually not bad one. What's surprisingly good for that deep violet, a good ultraviolet effect, but not Mega Let's unplug it. next.

capacitor is going to be I Think this will be self- discharging. Actually, you know what? Let's just, uh, shun it. Anyway, it's self- discharging. That's fine.

Um, but I'll I'll play Sa and discharge them anyway. let's put the 330. Nano This is a huge 330 Nano for capacitor. This is where if your board just watching me do this, you could just basically jump to the end and see the results.

It's your call 330 n looking a lot brighter. It's now let's say 1.9 5 I'd say that's close to 1.9 watt. Okay, 1.9 Watts it is. Could just say 2 Watts I'm going to say 1.9 Wats and that is going to be very visibly bright.

Yeah I can see it glowing even in the studio lighting I can see where you can see the Shimmer off it just at the side. That's very good. Unplug unleash, capacitor shunt capacitor anyway and we're on to 470 nard now, which is an unusually small capacitor. just depends on the manufacturer and construction.

The Varian size quite a lot. uh 47 N Farad brighter again looking pretty good in the making this pop this hat, it's very visible. uh 2.7 Watts Okay, and then actually I'm just going to unplug this like this right? just to show you that it does hold a charge if you don't and it just made a mockery of Me by not making a l crack noise. Oh okay, I just plug unplugged it at the zero crossing point.
Then excellent to try dabbing that across. Let's go to uh, let's plug unplug this this time and let's go to one microfi this time. which I would expect to be about 5 wats then based on what we've done so far. So let's plug it in at that.

I might even parallel two up. Very. that's actually making the hot light off camera. That's pretty good.

Um, it's 6 wats 5. Well, 5.9 let's call it 5.9 Watts So actually it doesn't take to much capacitance. I'd Kind of like to get a 2.2 Megard capacitor. Wonder if I've got one? H I Shall go and look for that one moment please.

I Did not find the capacitors I was looking for but I did find some motor uh run capacitors. Let's cut these open. let's rip them over and pop them out. These are from CPC in the UK and this one is rated 4 microfi which would be about 4.

microfi should be about 24 wat shouldn't it? Theoretically, at this rate, could this be an interesting way to tame them down? Just use motor capacitors and this one is 2 Mard so it's going to tame it down to about 12 wats. Theoretically, let's find out: I wonder if these have a discharge resistor I don't think they've got a discharge resistor and this is uh Shady cuz these will deliver quite a bang. Okie Doie I shall test that. So I shall get a bit of wire here.

This will do and I shall get this Wi and hook it around one terminal like this. just temporary connection and the other wire shall be hooked around the terminal like this. Again, a temperary connection and let's see what happens if I stuff this into this terminal here and plug it in. What are we going to guess? So this is 2 Mard Motor Capacitor What? you Reon is it going to be 12 wats? Let's find out the power is 11.5 wat 11.5 wats that is actually starting to get a very useful level of ultra violet while also extend the life and you could fit this capacitor in the case.

Now to test this, if this has a discharge capacitor resistor, should I say I'm going to short it out but I'm going to uh, basically just disconnect it from here. First, do Not do what I am doing and that should have left a charge on if there is a hold on. there was a bit of a pop there, not much. I shall test it I I'll make a note Down Below in the description.

If there is H a discharge resistor in these, I'll disconnect them and just see the voltage goes down at a modest rate. So let's get the next connect C on. So this is the four micard and that is theoretically going to run the LED somewhere around kind of half its rating. That' be quite useful and these motor capacitors are very common and you don't have to worry about them going short circuit cuz the worst case happening.

Uh, if they went short circuit is the LED would just go up to full brightness. So I should pop this in here. Place your bats. what's it going to be with for Mard very bright it's going to be that's really making everything light now and it's 23 Watts So that's effectively rendered it.
4 Microfarad I'm shiing my aate here equal. uh, let's say 20. it's 23.7 let's say 24 Wats that's interesting. Uh, that's going to half the power dissipation from this Led.

Oh I didn't check the power factor. Hold on. let me just check that fiveish 58 Yeah, that's what I'd expect. So all worth doing those experiments.

They were interesting. I should just discharge capacitor. It should have been discharged because of the Uh. it was in circuit with that.

But there we have it. Uh, the Ultrav Viet LED dubing it with capacitors. Here are the end results. Very easy to tame it down and in a suitable fixture.

Uh, that would just be a great way of getting a long lasting LED that still put out fairly decent power, but the LED chips would last a lot longer. so that was A. That was a good experiment that was well worth doing. Kind of want to build that into a case now with the Uh with an actual capacitor just to see how that uh, how that works.

But there we have it. Uh dub being an UltraViolet or a white LED just for prolonging LED life in a very simple way.

13 thoughts on “High power ultraviolet led experiments”
  1. Avataaar/Circle Created with python_avatars @-r-495 says:

    3M sells good safety glasses that will protect you from that light.
    It may leave you with yellowish spots all over if you get too high of an exposure.

    I‘m interested in these LEDs but am looking for ones under 300nm.

  2. Avataaar/Circle Created with python_avatars @unmanaged says:

    Clive could you not make a dimmer this way and just switch each in?

  3. Avataaar/Circle Created with python_avatars @capitalinventor4823 says:

    All of the resin for 3D printing that I've seen, that is for the hobbyist, states that it cures using 405 nm UV light so that's what I thought the printers and curing stations were using.

  4. Avataaar/Circle Created with python_avatars @WouterWeggelaar says:

    I am using these LEDs for developing cyanotype photography, and I bought these about a year ago. Bought 4, using one so far to expose cyanotype coated papers.
    Now with these capacitor hacks, I might have to start messing with the LED (which I planned for any way hence purchasing a few!)

    I did also ground the heat sink to be extra sure

  5. Avataaar/Circle Created with python_avatars @toveryonder1115 says:

    i wonder if your camera is 60hz and that is why everything has flicker with your 50hz electricity.

  6. Avataaar/Circle Created with python_avatars @Glanmire3 says:

    As I'm a professional at a big industrial OEM LED company I almost had a hearth attack when he turn the unit on without any personal protection (the gloves is almost nothing, compared to any other uncovered body parts). It doesn't matter if you turn the emitter away as the indirect emission (bounced from walls and other object back) can make the same injury.
    If it's not even close to 50W AND a really UV-A emitter (so somewhere 395 – 360nm emission range) then he is blind already or have serious IRREVERSIBLE eye damage to the retina. (just Google it)
    One more advice. Never touch any surface what is participating in Uv relation by bare hand, like the encapsulation of the LED and cover glass because even a very small amount of grease from the hand causing discoloration (turning into black).

  7. Avataaar/Circle Created with python_avatars @mevk1 says:

    This knob is wondering if adding an inductor would be practical way to increase power factor. If so maybe you could follow up with video by experimenting with some inductors you have laying around?

  8. Avataaar/Circle Created with python_avatars @executive says:

    the power isn't of interest to me at all. What matters is spectrum. You should do an experiment to measure the actual wavelength. Producing real UV LEDs is very expensive, and in my experience, this low cost aliexpress/ebay garbage is usually just a really deep purple. It'll make things glow, but it's no good for curing resins (no better than ambient light). Fake shanzai crap. The listings usually misrepresent the peak emission wavelength by a large margin (they just copy datasheets from legit products). It's very easy and fun to measure. All you need is a narrow slit, a piece of CD-R for a diffraction grating, a ruler, and some math.

  9. Avataaar/Circle Created with python_avatars @TomTRobot says:

    Apparently UV curing is basically a matter of proportion of t*W/cm2. Reducing power to increase longevity would either increase time and/or require changing distance, both of which may be acceptable trade-offs depending on the application. Given that there are are exponential decreases in longevity with high junction temperature you could get a lot of bang-for-the-buck tuning the power usage of these devices rather than investing in cooling solutions.

  10. Avataaar/Circle Created with python_avatars @grotekleum says:

    Wah, I've gone blind now.

  11. Avataaar/Circle Created with python_avatars @charliesoffer says:

    That's one heck of a SAD lamp, Clive! Mandatory sun screen and shades must be worn though… 😊

  12. Avataaar/Circle Created with python_avatars @4lecsg says:

    For this to work, the LED board needs to be somewhat decent quality. I tried to limit a very cheap led projector like this, and it ended up powering only a quarter of the LEDs on the board, and it still failed after a year while also lowering the light output.

  13. Avataaar/Circle Created with python_avatars @analoghardwaretops3976 says:

    All capacitor series droppers are passing full
    " inrush current " at the instant of
    " switch on " ( cap fully discharged state )… and is worst , if it happens at peak of the ac voltage waveform…
    Only later does it do the
    " average " current limiting…
    Therefore an additional NTC series component will reduce the liabilities…

Leave a Reply

Your email address will not be published. Required fields are marked *

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