Jump to content
UltravioletPhotography

What portable lightsource would you recommend for UVIVF?


Recommended Posts

So, you might have seen my previous post where I asked about this somewhat as well as I also asked about filtering options.

I have talked with a few members but I still find myself indecisive.

I would prefer to have a strong 365nm LED torch with a ZWB2 or equivalent filter mounted on it's front as to purify the output.

I would like as much power as possible since I want to be able to stop down or light paint larger objects.

Thus far I have considered the Nemo, and the Alonefire H42 45W.

https://www.aliexpress.com/item/1005002133384717.html?spm=a2g0o.cart.0.0.25733c00Uec9BH&mp=1

https://www.ebay.com/itm/402989627275

The Nemo is only 15W but it's output is supposed to be decent all things considered. It's also significantly cheaper but I'm not sure that it's output is enough. There's another problem though that the Alonefire could have some issues as having a genuine 45W worth of LEDs in your torch is difficult to pull off well, you need good components and batteries, and I'm not sure it's worth it for the price, it could only be slightly more powerful than the Nemo.

 

Is there someone who perhaps has both and can compare? Or does anyone have any other suggestions? Really, I just want a lot of fluorescence and the option to take the given source outside at night. Some of you might wonder why I don't just go and order one of the options and see for myself, but my funds are limited and I put a lot of emphasis on spending wisely. I have already discussed this with a few members but I thought I would make a public post to see if anyone knows more.

 

Thank you for reading!

Link to comment

I think the wattage numbers can't be trusted at all and you shouldn't use that to compare any lights sadly.

Just look at the batteries to get an idea. It can't push more than the little 18650 can give it.

 

I would see if you can still get a $20ish Nemo. Those are excellent.

Alternatively the 365nm convoys are good too. If they have come down in cost, that can be an option.

 

If your technically inclined, you can buy cheap LG 3535 385nm leds for a dollar or so and solider a 7135 driver and make your own light for $10.

Link to comment

Thank you.

The Nemo I linked is in fact $20-ish. Problem is that I really don't know how strong give or take 7W of output is.

The LEDs you linked seem very interesting and cheap. I'm just not sure if filtering them with a ZWB2 would be ideal. Plus as I said I would like my source to be portable. And I'm not very technically inclined indeed. I want to hook up a few LEDs but it's come to the point where I might be seeking help from an electrician of a hobbyist robot constructor.

Link to comment

Thank you.

The Nemo I linked is in fact $20-ish. Problem is that I really don't know how strong give or take 7W of output is.

The LEDs you linked seem very interesting and cheap. I'm just not sure if filtering them with a ZWB2 would be ideal. Plus as I said I would like my source to be portable. And I'm not very technically inclined indeed. I want to hook up a few LEDs but it's come to the point where I might be seeking help from an electrician of a hobbyist robot constructor.

 

Ok in that case you're not saving any money and would pay more than if you just bought the Nemo.

 

I would go for a Nemo now over the others at that price point.

 

I only became aware of the LG 3535 chips when I broke my original convoy. I bought a 5 pack of LG 3535 385nm leds on 16mm boards for $5. I then just switched one out for the broken Led board from my convoy and then got a strong 385nm light. I now have 4 spares. I can't remember the board diameter for the Nemo off the top of my head, but it too is common. If it breaks you can do the same in the future. The Nemo also uses 7135 drivers.

Link to comment

 

 

Ok in that case you're not saving any money and would pay more than if you just bought the Nemo.

 

I would go for a Nemo now over the others at that price point.

 

I only became aware of the LG 3535 chips when I broke my original convoy. I bought a 5 pack of LG 3535 385nm leds on 16mm boards for $5. I then just switched one out for the broken Led board from my convoy and then got a strong 385nm light. I now have 4 spares. I can't remember the board diameter for the Nemo off the top of my head, but it too is common. If it breaks you can do the same in the future. The Nemo also uses 7135 drivers.

 

I looked into the Convoy S2+ nichia UV 365nm and the price they're asking for it given that it is not filtered and only 3W is quite ridiculous.

At the price point of $20 ish I guess the Nemo is my best option, but what if I wanted to spend up to 60$? Is there something significantly better in that price range? Power is key.

Thanks.

Link to comment

Fandyus, There may be more powerful torches, I don't know, there are so many that are advertised as being more powerful, but the two torches that we have tested and used the most on here in recent years are the Convoy S2+ 365nm UV Nichia LED,

and there are several of those. Also the torch that Andy named the Nemo, and it comes with 365nm U glass. It is slightly more powerful than the Convoy.

So I can recommend either of those. You can look on Gearbest for those also.

The s2+ seem a little more expensive now, $30?

Link to comment

Fandyus, There may be more powerful torches, I don't know, there are so many that are advertised as being more powerful, but the two torches that we have tested and used the most on here in recent years are the Convoy S2+ 365nm UV Nichia LED,

and there are several of those. Also the torch that Andy named the Nemo, and it comes with 365nm U glass. It is slightly more powerful than the Convoy.

So I can recommend either of those. You can look on Gearbest for those also. 'Nemo' I have seen on eBay.

The s2+ seem a little more expensive now, $30?

The Nemo's optical power is actually 3-5 times higher than a Convoy S2+, but it has a bit wider light- beam.

The convoy's beam is rather narrow and in shorter distances the peak power density can be higher.

 

The U glass in the Nemo is rather ideal as it's peak transmission is close to the 365nm LED's peak.

Both Nemo and Convoy have LEDs and internal structures that have a rather low amount of fluorescence.

That is an unknown factor for other torches and LED arrays.

 

The U-glass used here in the Nemo is NOT suitable to filter out VIS from LEDs with longer wavelengths like 385nm and 395nm as the transmission of the u-glass is low at their peaks.

Most of the UV-light would then be absorbed and heating the filter glass instead of exiting as useable light.

A 395nm LED's light will lose 90% or more of its intensity, when filtering it.

 

If you want to clean up light from those types of LEDs (385nm and 395nm) A SEU2- or SEU3-filter is the best alternative, but that is rather expensive to filter a torch.

The SEU-filter will also absorb energy in it's non dichroic parts and if the LED-array is powerful enough there is a risk of destroying the filter due to thermal stress as it is not made of one glass type, but is a glued stack.

Link to comment

I have never used the Alonefire h42 light. But you got me curious. What I have found from some searches is it uses 3 of the 4LED 365nm group. A couple reviews on YouTube say its a broader smoother light than the typical Convoy UV light, that isn't as center concentrated and doesn't reach as far.

 

So if you want to flood a broad area with UV light, than this maybe a great option. But if you want to excited a small area with intense light and paint around that area, than it might not be the best.

 

One crazy video the guy was illuminating his neighborhood from his balcony with it. So it does look broad to me. The Convoy does have a very tight hot center. Which is surprisingly close to the Nemo.

 

The Convoy is only 1 365nm led, the Nemo is a single group of 4 (which isn't the best at very close distance, you can see the separation line) the Alonefire seems to be three groups of four for total of 12 365nm leds, for broad angle spreading light.

 

So think which way you want to go. Spotlight (Nemo) or floodlight (Alonefire).

Link to comment

dabateman, I'm glad to have showed you something that sparks curiosity. Have fun if you do decide to explore that someday, and be sure to tell us :)

 

After what you all are saying, I'll probably get the Nemo as having a spotlight seems to be more important for flowers and small objects, which is my aim. I just hope the seller on eBay will sell it to me without the charger as I don't need that. I have ordered a better one.

Link to comment

Here are Ulf's scans of the Nemo LED and of the U glass used in the Nemo, and Nichia's graphs of the Convoy LED.

Personally, I don't think the Nemo is as much stronger than the S2+ as David expresses. Perhaps it is on paper, but my experience with them doesn't seem to be that much stronger.

Also, it is my opinion that the best U glass filter to use is U-340 which cuts belong 400nm instead of any U-360 or UG1 type that slightly transmit above 400nm.

 

post-87-0-23990600-1627268689.jpg

 

post-87-0-04171900-1627268778.jpg

 

post-87-0-46767100-1627268893.jpg

Link to comment
Thanks, Cadmium. I actually just saw the post of yours that you originally posted this in. But thanks for having the effort to show me here as well.
Link to comment

I've been looking at this one by UvBeast which has 540mW output : https://www.ebay.com/itm/uvBeast-Black-Light-UV-Flashlight-V3-365nm-FILTERED-Ultraviolet-HIGH-POWER-a/402093546119

It is out of the price range Fandyus has stated, but UvBeast also offers a smaller one (450mW) which looks nice: https://www.ebay.com/itm/402302778887?_trkparms=aid%3D1110006%26algo%3DHOMESPLICE.SIM%26ao%3D1%26asc%3D20200818143230%26meid%3Dc83e445fa813436ea4d8afac7cbde67f%26pid%3D101224%26rk%3D1%26rkt%3D5%26sd%3D402093546119%26itm%3D402302778887%26pmt%3D0%26noa%3D1%26pg%3D2047675%26algv%3DDefaultOrganic&_trksid=p2047675.c101224.m-1

I like this smaller one because it has a built in USB port for charging. Cost is $39.00 (US).

 

Both are 365 nm. So many of the torches offered on Ebay are 395 nm, so be sure to carefully note the wavelength. It isn't always clear.

Both are supposedly filtered. No info though on the filter transmission.

 

I haven't used either of these. Simply pointing out something which I've been looking at which seems worth checking out. :lol:

 

I would definitely go for the 45W output or better for shorter exposure times.

Link to comment
enricosavazzi

I have never used the Alonefire h42 light. But you got me curious. What I have found from some searches is it uses 3 of the 4LED 365nm group. A couple reviews on YouTube say its a broader smoother light than the typical Convoy UV light, that isn't as center concentrated and doesn't reach as far.

[...]

the Alonefire seems to be three groups of four for total of 12 365nm leds, for broad angle spreading light.

I have an Alonefire H42 UV torch but have not been using it much so far. It is obviously more powerful than a Convoy, but I made no quantitative comparison. I do not agree with the mentioned YouTube reviewer that the Alonefire is much more diffuse than a Convoy. In fact, at about 1 m distance, it clearly shows a rather narrow doughnut-shaped highlight, darker both at its center and outside the doughnut. It might do with a little diffusion. I literally felt the heat of the radiation on my face when the torch accidentally switched on for a couple of seconds while I was handling it (I was wearing polycarbonate goggles, so no exposure of the eyes). Make a point of keeping your fingers away from the power button when handling this torch, because the button is sensitive to touch.

 

The Alonefire does have a few quirks that one should be aware of. For one thing, it accepts four 18650 batteries (supplied), but in addition to opening the rear screwed-on hatch one needs to undo four Phillips screws and remove a PC board to insert the batteries, then reassemble the torch in the reverse order. In other words, forget about switching batteries in the field to continue working (although you may connect an external power bank via the supplied USB cable).

 

Correction: There is a simple way to access the battery compartment, which initially I had missed (my torch came without instructions): unscrewing the body from the head roughly in the middle of the torch exposes the batteries, allowing a quick swap.

 

Another thing to mention is that the torch has a type C USB socket, but requires an old-fashioned 5V DC USB power supply or power bank. It does not work with USB 3 PD power supplies, unless they supply 5V by default.

 

Corrected: The torch power button lights up green when the torch is on, red when the torch is connected to a USB charger. It takes many hours of charging for the LED to turn green at the end of charging. So the green light is used to show both that the batteries are fully charged, and that the torch is currently on and spewing out UV. This can be confusing in a UV torch where you may not be immediately aware of whether the LEDs are emitting radiation.

 

Each of the LED packages contains 4 chips, so 12 chips in total. The built-in UV-pass filter is most likely one of the Chinese 365 nm ionic glass filters, probably not coated. The individual chips are visible as violet dots through the glass when the torch is on, but the amount of leaked VIS is not much.

 

Finally, the four batteries are connected in parallel, not in series. One bad cell may cause the other three to discharge through the defective one, which cannot be a good thing.

Link to comment

The heat things is something I would expect from such a torch (based on personal experience). Heating up your skin with UV is probably not the best thing. SWIR would work much better.

 

Those 4x chips are usually rated for 10 W of input power, 15 W might be an upper limit. If they are 50% efficient (I doubt it) they would put out about 5 W of UV each, for a total of 15 W minus the losses of the UV-pass filter. A Convoy S2+ outputs ~1 W at most with the new LED, even less with the U-340 filter, so 15 W is still a lot.

 

The "doughnut" beam can happen with parabolic reflectors if the light source is not in the focus point. Old flashlights with incandescent bulbs had the same problem, but if you could focus the beam you could get rid of the doughnut shape.

Link to comment

The heat things is something I would expect from such a torch (based on personal experience). Heating up your skin with UV is probably not the best thing. SWIR would work much better.

 

Those 4x chips are usually rated for 10 W of input power, 15 W might be an upper limit. If they are 50% efficient (I doubt it) they would put out about 5 W of UV each, for a total of 15 W minus the losses of the UV-pass filter. A Convoy S2+ outputs ~1 W at most with the new LED, even less with the U-340 filter, so 15 W is still a lot.

 

The "doughnut" beam can happen with parabolic reflectors if the light source is not in the focus point. Old flashlights with incandescent bulbs had the same problem, but if you could focus the beam you could get rid of the doughnut shape.

So without all optimistic assumptions it is likely 3 x10W input or less ,with an efficiency of around 30%, as higher selection tiers of LEDs are more expensive.

Then the total optical output before the filter would be 9W or less.

The filter would lose another 25%, leaving less than 7W optical output.

Link to comment
Well, that went down quickly. One possible solution could be to replace the LED chips with more efficient ones with the same size, voltage and most importantly maximum forward current. They often have similar values, but it depends whether they are wired in a 4P, 2S2P or 4S configuration.
Link to comment

Well, that went down quickly.

Life is a bitch, isn't it?

 

One possible solution could be to replace the LED chips with more efficient ones with the same size, voltage and most importantly maximum forward current. They often have similar values, but it depends whether they are wired in a 4P, 2S2P or 4S configuration.

I cannot follow your reasoning here except for finding identically configured LEDs in exactly the same case.

The optimal configuration is always serial coupling for a multi chip array, as then the chips cannot affect each other.

The gain of a more efficiency, by a different selection tier is seldom very big.

 

Even if you have seen some datasheet with an efficiency closer to 50% that kind of LED is extremely rare, if it at all exist in the real world.

 

Getting 20% more optical power, ( 30% => 50% ), is not very significant in the real usage situation either.

It would barely be detectable as a meaningfully shorter exposure time.

The same gain could be obtained by moving the torch just a little bit closer to the motif.

 

If you had unlimited funding, you could always improve the filter too.

A dichroic filter with less losses at the LEDs peak wavelength would help a bit.

 

Your reasoning about these improvements are linear, but the problems are not.

 

Have you heard about the tale of the payment for the design of the chess game?

https://purposefocus...th-b1f7bd70aaca

Getting more meaningful light from a light source is a similar problem, as a doubling gains just one stop of light.

 

To make it even worse we have to factor in the Inverse Square Law:

https://en.wikipedia...cal%20quantity.

 

Also look at this:

https://en.wikipedia...ishing_returns.

 

You can always hope for a quantum leap in UV-LED technology. Sometimes those happen.

A doubling of efficiency from 30% to 60% would also mean that the losses are cut by a half.

Then you could double the number of LEDs with the same amount of cooling.

Then two times of input power would give you four times the optical power.

Unfortunately such quantum leaps do not happen very often in the same technical area.

 

For LEDs there were one already:

https://www.nobelpri...csprize2014.pdf

 

Life is a bitch, isn't it?

Link to comment

So what did we finally end up calculating.

The Convoy is really only 1W, with 3x 7135 drivers.

The Nemo max is 7.5W, but realistically 4W with 5x 7135 drivers.

Both same battery.

The Alonefire is different with 4 batteries. But does that make a difference if parallel. And 3 times the LED of the Nemo. We don't know the Alonefire drivers. But maybe funny if the same. Maximum number of drivers I have seen on a board is 8. So 8x 350mA would only be 2.8A, not far from the 1.75A of the Nemo.

 

My guess is the Alonefire isn't 3x that of the Nemo, but only slightly brighter or the same. The Nemo is about the same as a convoy in the center, it just has a wider center.

Link to comment

Convoy and Nemo, same battery? No, they have different batteries. One battery each.

Convoy = 18650

Nemo = 26650 (larger battery)

 

I don't have any experience with the Alonefire, so I can't speak on that.

Link to comment

So what did we finally end up calculating.

The Convoy is really only 1W, with 3x 7135 drivers.

The Nemo max is 7.5W, but realistically 4W with 5x 7135 drivers.

Both same battery.

The Alonefire is different with 4 batteries. But does that make a difference if parallel. And 3 times the LED of the Nemo. We don't know the Alonefire drivers. But maybe funny if the same. Maximum number of drivers I have seen on a board is 8. So 8x 350mA would only be 2.8A, not far from the 1.75A of the Nemo.

 

My guess is the Alonefire isn't 3x that of the Nemo, but only slightly brighter or the same. The Nemo is about the same as a convoy in the center, it just has a wider center.

I think you meant the same number of batteries when you said "Both same battery"

 

If you compare the Nemo's bigger 26650 battery with the four 18650 in parallel in the Alonefire, there are one advantage:

The internal resistance in a good battery is lower for the bigger one, but by putting four smaller in parallel the total serial resistance might get even lower.

There are also four times as many contact points between torch and batteries, also lowering that contact resistance.

 

If the Alonefire consumes more current like David describes above, the total design situation could be a bit better in this respect for the Alonefire.

Link to comment

Life is a bitch, isn't it?

 

I cannot follow your reasoning here except for finding identically configured LEDs in exactly the same case.

The optimal configuration is always serial coupling for a multi chip array, as then the chips cannot affect each other.

The gain of a more efficiency, by a different selection tier is seldom very big.

Yes, this is what I meant, finding compatibile LEDs with better efficiency that can be soldered on the same board and be driven by the same circuit. The optimal configuration is always all-serial as you say, also because the current is the same in each die, but I have seen 2x2 chips in all three possible configurations, even all-parallel. My 10 W LED has a 2S2P configuration.

 

Even if you have seen some datasheet with an efficiency closer to 50% that kind of LED is extremely rare, if it at all exist in the real world.

 

Getting 20% more optical power, ( 30% => 50% ), is not very significant in the real usage situation either.

It would barely be detectable as a meaningfully shorter exposure time.

The same gain could be obtained by moving the torch just a little bit closer to the motif.

A 30% -> 50% improvement may not seem a lot, but it means the following:

 

You get from 70% to 50% of wasted energy, and so at the same input power you get ~1.66 times more light and at the same time only ~71% of the heat, with the same runtime;

 

keeping the output power the same, you cut the input power by 40% and you produce less than half the heat as before (about 57% less). That would be a nice improvement in my opinion, but you would have to modify the drivers to do so, so this is all theoretical.

 

If you had unlimited funding, you could always improve the filter too.

A dichroic filter with less losses at the LEDs peak wavelength would help a bit.

Yes, that would also improve things. A 365 nm bandpass filter, with a rectangular transmission 10 nm wide and 95% of transmission in the passed band would be great, the only problem being that this kind of filters are quite expensive at that size, especially if custom-made.

 

Too many quotes in a single post (wasn't 5 the limit? It seems it's three). See next post.

Link to comment

Your reasoning about these improvements are linear, but the problems are not.

 

Have you heard about the tale of the payment for the design of the chess game?

https://purposefocus...th-b1f7bd70aaca

Getting more meaningful light from a light source is a similar problem, as a doubling gains just one stop of light.

 

To make it even worse we have to factor in the Inverse Square Law:

https://en.wikipedia...cal%20quantity.

 

Also look at this:

https://en.wikipedia...ishing_returns.

Oh, you remind me of the past, I think it was about 8 years ago (time truly flies!), I was in middle school and in our math textboox there was this very example about exponential growth. I wasn't very impressed at the time because I already knew that exponential growth is fast and starts slow, but somehow I remembered this example.

 

Yes, bringing the torch closer increases the irradiance significantly, at the expense of a narrower beam.

 

You can always hope for a quantum leap in UV-LED technology. Sometimes those happen.

A doubling of efficiency from 30% to 60% would also mean that the losses are cut by a half.

Then you could double the number of LEDs with the same amount of cooling.

Then two times of input power would give you four times the optical power.

Unfortunately such quantum leaps do not happen very often in the same technical area.

 

For LEDs there were one already:

https://www.nobelpri...csprize2014.pdf

Such a leap would be so great with deep UV LEDs, I think it will happen one day, but probably not very soon.

 

Life is a bitch, isn't it?

Yes, sometimes it really is.
Link to comment

What I tried to point out is that what you suggest is gaining rather little in a practical situation, compared with costs and efforts to do the modification, if it is at all possible to find slightly better components.

With the Convoy S2+, Nemo and similar torches we are already on a rather steep slope of the exponential improvement curve.

 

Different kinds of problems begin to emerge with increasing power levels and modifying a torch a bit by changing the used LED will not improve the practical usage very much.

Link to comment

Yes, in simple words, it's not worth it. For the cost of the modification one can buy an even bigger torch with more power, even if the efficiency is lower.

 

 

 

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...