Information and discussion about some LEDs

[Stefano]

I have no affiliation with any sellers or companies mentioned in this topic.

 

I recently did some searches on LEDs, since technology improves as time passes, and there could be new LEDs that were unavailable before.

 

About 365 nm LEDs, I posted a topic some years ago (time flies!), and I haven't done new searches on this. Maybe some LEDs with improved efficiency or radiant power are available.

 

If you search for LEDs online, especially if you search for a wide variety of wavelengths and power LEDs, you will notice that LEDs with wavelengths ranging from 365 nm to about 1050 nm are easily available with input powers between 1 W and 100 W (sometimes even more). They all look the same, with 1 W LEDs commonly available as a single chip (sometimes on hexagonal star PCBs), 10 W ones being made of 9 chips, usually connected in a 3S3P configuration, and 100 W ones being made of 100 1 W chips connected in 10S10P. The fact they all look the same makes me wonder if there's some kind of standard manufacturers are following. LEDs in this wavelength range are the easiest to find, they are efficient, and they are available in a lot of wavelengths. Between about 365 nm and 425 nm (UV to violet), you can find them in 5 nm increments (365 nm, 370 nm, 375 nm, ect.). The hardest ones to find are in the cyan (~480 nm) and yellow-green (~560 nm) regions, but if you search long enough you will find them. I have them in 10 W power, I still need to mount them on a heatsink (remember, it's important for power LEDs to mount them on a heatsink, they will overheat otherwise).

 

I have recently discovered this company, which seems to make LEDs with a very dense selection of wavelengths (every 10 nm in the UV and visible range): https://www.bing.com/ck/a?!&&p=fa0e995133e1d8fbJmltdHM9MTY5OTA1NjAwMCZpZ3VpZD0yZjFkYzQ3My0zZDI0LTZlMzUtMjJkNi1kNmEyM2M1MzZmOGUmaW5zaWQ9NTIyMQ&ptn=3&hsh=3&fclid=2f1dc473-3d24-6e35-22d6-d6a23c536f8e&psq=Lumixtar&u=a1aHR0cHM6Ly93d3cubHVtaXh0YXIuY29tL3Byb2R1Y3QtY2VudGVy&ntb=1

 

I don't know how one can buy those LEDs. They would be useful if one wants to make a righ resolution LED monochromator.

 

LEDs outside the 365-1050 nm portion of the spectrum are available, but noticeably harder to find. They are typically more expensive, avaliable with input powers of at most a few watts, and they are much less efficient. You can find them on eBay, but you should look for specialised companies, such as Thorlabs, for a wider selection.

 

In particular, the next LED wavelength you will easily find below 365 nm is 340 nm. Thorlabs sells one (link), and you can find them on eBay (link). The LEDs look similar, but have different specifications. Both have changed around the same time to a newer model, and they looked identical previously. The eBay one should be model CUN4GF1B, produced by Seoul Viosys. What I find very weird is that there's no mention I could find on this LED on Seoul Viosys' site. I don't know if Thorlabs' LED is produced by Seoul Viosys too, but there's no mention about this company I could find. Can someone explain this?

 

Thorlabs' LED claims a typical power output of 69.2 mW, and a power input of 3940 mW, with an efficiency of 1.75%. From the datasheet of CUN4GF1B we have a typical input power of 1025 mW (4.1 V and 0.25 A), and a typical power output of 40 mW, for an efficiency of 3.9%. Also, CUN4GF1B's datasheet specifies a minimum, typical and maximum current rating. What does that mean?

 

There's another notable 340 nm LED I found, this one: https://shop.boselec.com/products/340nm-uva-led-smd-packages-medium-power

It has an input power of 1575 mW (4.5 V, 0.35 A), and an output power of 160 mW, which means a very high efficiency of 10.16%.

[lonesome_dave]

At the other end of the spectrum I ordered another set of 940nm LED mini-torches to use with my various IR viewers and expected them to be totally dark like the previous set with the same specs (3W) and the same seller (China). When I fired them up I assumed they must be 850nm as the red glow was obvious. But my spectrometer showed approximately 940nm like the first set. However, my broadband power meter showed that the new set was about 6 times more powerful. (The logarithmic output curve of the LED torches was overlapping the logarithmic sensitivity curve of human eyesight somewhere in between).

 

Not really what I wanted because it was supposed to be a demonstration of IR viewing with a totally dark light source. They are almost as bright as the much more expensive 940nm torches I have (which also have a visible red glow due to their high power). I guess I shouldn't complain about a little 1-AA mini-torch for $12 that has such high output. Can't be much heat-sinking in that little thing and they do get noticeably warm rather quickly so I don't expect them to last very long.

 

I can still do the demo with the few lower-powered ones I ordered about 8 months ago. Even with dark-adapted eyes they have no visible output in the dark. You need an IR viewer to make sure they are turned off. I'll try some nearly dead batteries in the new ones to see if they can stay dark but still work with the IR viewers.

[colinbm]

@Stefano thanks for the latest news on LEDs.

@lonesome_dave perhaps it is time to use a filter in front of the LEDs.

[Andy Perrin]

Yeah, I was going to say, just filter the LEDs and you're all set, lonesome_dave.

[lonesome_dave]

Agreed. If I had a good bandpass filter at 940nm that would probably do the trick but I don't have one close enough to work. I tried filtering with the cheap color dye longpass filters with shallow slope (950nm & 1000nm) but the red glow is still visible from several feet away in the dark. I have a Thorlabs 1000nm hard-edge longpass that blocks the red glow but it also blocks most of the output. However, I don't want a solution that costs 15x what the mini-torch costs because I want it to be repeatable at low cost. This was to be an accessory to the IR viewers I'm donating to schools.

 

I think I have found a supplier of the original torches in the UK for a good price. Depleted batteries in the brighter ones sorta work but not the sort of solution I'm comfortable with.

[dabateman]

You could contact Tangsinuo and ask for a specific size.

He does sell narrow band 940nm filters.

https://www.ebay.com/itm/275733102708?mkcid=16&mkevt=1&mkrid=711-127632-2357-0&ssspo=aW1-_wTCSp-&sssrc=4429486&ssuid=RxqKLNPYQr-&var=&widget_ver=artemis&media=COPY

 

https://www.ebay.com/itm/275733102696?mkcid=16&mkevt=1&mkrid=711-127632-2357-0&ssspo=aW1-_wTCSp-&sssrc=4429486&ssuid=RxqKLNPYQr-&var=&widget_ver=artemis&media=COPY

[lonesome_dave]

Thanks dabateman, those are rather inexpensive at $10 for the 25mm round. Peak transmission above 90% too.

[Lou Jost]

A large US retailer for UV LEDs (including UV-C) is Boston Electronics.

https://www.boselec.com/products/

They ship extremely quickly and sell any amounts, including single LEDs, and they also sell the matched drivers for them.

 

[Stefano]

One of the 340 nm LEDs I linked above is from that retailer. They don't offer 340 nm LEDs in high power, but only up to medium power (still powerful at 160 mW of output power). Their high power LEDs seem to be made of 4 chips connected together.

 

I hope they will offer more powerful options in the future. Still, a 10% efficiency at 340 nm is a promising sign of technological improvement. It's impressive to think that blue LEDs began to be commercially available in the 1990s, and were still somewhat expensive in the 2000s. Today some of them exceed 80% efficiency: https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://www.led-professional.com/products/leds_led_modules/industry2019s-brightest-most-efficient-royal-blue-led-by-cree&ved=2ahUKEwjIuNuK4a-CAxX-RPEDHQTTAR4QFnoECBYQAQ&usg=AOvVaw20qzAPe9PSZB2i9B5rWYYA

[ulf]

It is indeed a promising trend of improved efficiency.

It is nice to see improvement for the LEDs of 340nm too.

 

The drive for research and improvement normally very often commercial.

 

The blue LEDs have been driven forward to improved efficiency by the vast global need for good illumination. They are mainly used for making white LEDs.

As LEDs are sensitive to heat it is ideal if as much as possible of the input power is converted to light beside the obvious need for high efficiency.

I think the illumination-market is by far the biggest market for LEDs 

 

 

LEDs in the 365-405nm range is mainly used in the industry for curing lacquers, adhesives and bonding agents.

 

I have no good idea of any big volume market with a need for shorter wavelength LEDs that can drive an evolution very fast for those LEDs, making them quite efficient and cheap.

AFAIK they are mainly used in measurement and analysis equipment that are expensive and with a limited market.

 

I would love to hear more about other possible usage areas of different LED types.

I might be totally ignorant about some areas that really need many efficient LEDs with 340nm or shorter.

[Lou Jost]

16 minutes ago, ulf said:

It is indeed a promising trend of improved efficiency.

It is nice to see improvement for the LEDs of 340nm too.

 

The drive for research and improvement normally very often commercial.

 

The blue LEDs have been driven forward to improved efficiency by the vast global need for good illumination. They are mainly used for making white LEDs.

As LEDs are sensitive to heat it is ideal if as much as possible of the input power is converted to light beside the obvious need for high efficiency.

I think the illumination-market is by far the biggest market for LEDs 

 

 

LEDs in the 365-405nm range is mainly used in the industry for curing lacquers, adhesives and bonding agents.

 

I have no good idea of any big volume market with a need for shorter wavelength LEDs that can drive an evolution very fast for those LEDs, making them quite efficient and cheap.

AFAIK they are mainly used in measurement and analysis equipment that are expensive and with a limited market.

 

I would love to hear more about other possible usage areas of different LED types.

I might be totally ignorant about some areas that really need many efficient LEDs with 340nm or shorter.

 

 

 

 

Boston Electronics LEDs go down to 265nm. They aren't very efficient though. They are very well documented on their website, with efficiencies, power versus angle graphs, linewidth plots, etc.

[Stefano]

As for the really shortest wavelengths, here's a 235 nm LED: https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://silannauv.com/products/sf1-235nm-flat-lens/&ved=2ahUKEwiemMeFuLCCAxXpg_0HHfWAAmkQFnoECB0QAQ&usg=AOvVaw1d0lyzcIs4ZyAk9VDUiOd9

 

The shortest wavelength LED so far made emits at 210 nm: https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://www.led-professional.com/technology/light-generation/aluminum-nitride-light-emitting-diodes-with-the-shortest-wavelength-towards-dioxin-pcb-decomposition-technology&ved=2ahUKEwjhsPGYurCCAxXvgv0HHU6tA0EQFnoECCsQAQ&usg=AOvVaw1qqMCAYka3OYZXPGbu5Vmo

 

To my knowledge they aren't commercially available yet.

[lukaszgryglicki]

I would love to get some efficient 305-310nm LEDs.

[Lou Jost]

1 minute ago, lukaszgryglicki said:

I would love to get some efficient 305-310nm LEDs.

I have some good 310nm LEDs on the way from Boston Electronics

[Andy Perrin]

7 hours ago, ulf said:

I would love to hear more about other possible usage areas of different LED types.

I might be totally ignorant about some areas that really need many efficient LEDs with 340nm or shorter.

Maybe semiconductor manufacturing? I think the most recent generations may use electron beams or X-rays, but I bet a lot of chips could still be made using UV-C...whether they would consider using LEDs for the purpose, I have no clue.

[Lou Jost]

6 minutes ago, Andy Perrin said:

Maybe semiconductor manufacturing? I think the most recent generations may use electron beams or X-rays, but I bet a lot of chips could still be made using UV-C...whether they would consider using LEDs for the purpose, I have no clue.

As far as I know, no large-scale photolithography has been done with LEDs. What's needed in photolithography is high power all concentrated in a single wavelength. Photolithography lenses are not corrected for CA (except possibly for a fraction of a nanometer around the designed wavelength). This freedom from CA correction lets them be better optimized for other things like flatness of field and low distortion. LEDs are not good choices because they are much less monochromatic than atomic emission lines.

[Stefano]

[Quoting Lou Jost]

 

Did you get the high power one? It has an efficiency of about 2.38% (0.4 W output at 16.8 W input). 400 mW of output is a lot.

[Lou Jost]

Yes, I got the high-power plug-and-play package.. Jonathan has shown that mercury lights are a lot brighter, but there may be some advantages to a nice compact LED w/lens if one is doing microscopy

[Stefano]

I would be interested to know how this LED performs.

[Lou Jost]

I'll let everyone know. But it sill take months to get here to Ecuador. I expect a friend will travel to Ecuador with it in January.

[JMC]

23 hours ago, Lou Jost said:

Yes, I got the high-power plug-and-play package.. Jonathan has shown that mercury lights are a lot brighter, but there may be some advantages to a nice compact LED w/lens if one is doing microscopy

Good luck with it Lou. If it is bright enough, then it certainly offers the advantage of size (and reduction in unwanted wavelengths, which will reduce the extent of filtering required). Also, cost. Are you going to combine it with a fused silica condenser lens to try and get a nice parallel beam and harness as much out the output as possible for your microscope?

[Lou Jost]

1 hour ago, JMC said:

Good luck with it Lou. If it is bright enough, then it certainly offers the advantage of size (and reduction in unwanted wavelengths, which will reduce the extent of filtering required). Also, cost. Are you going to combine it with a fused silica condenser lens to try and get a nice parallel beam and harness as much out the output as possible for your microscope?

Jonathan, it comes with a lens built in, and I will have to see what that lens does exactly. Ideally I would have liked it without any lens, so that it could be put physically in the back focal plane of an inverted microscope objective used as a condenser, eliminating the need for a separate home-made condenser.  But I do have a few random silica and calcium fluoride lenses to play around with.

 

I read your blog post about how you made a condenser that worked really well. That was impressive. Would you be willing to share the plans? This would be really useful.

[JMC]

2 hours ago, Lou Jost said:

Jonathan, it comes with a lens built in, and I will have to see what that lens does exactly. Ideally I would have liked it without any lens, so that it could be put physically in the back focal plane of an inverted microscope objective used as a condenser, eliminating the need for a separate home-made condenser.  But I do have a few random silica and calcium fluoride lenses to play around with.

 

I read your blog post about how you made a condenser that worked really well. That was impressive. Would you be willing to share the plans? This would be really useful.

Ah right, hope the inbuilt lens works out.

 

As for sharing plans, that's easier said than done. I don't really have 'plans'. I tend to just position things by trial and error and then find ways of fixing them in place at the position which works the best. I know that sounds very unscientific for a person who calls himself a scientist, but I am very much a trial and error person.

 

Do you mean my home made UV condenser I shared here - https://jmcscientificconsulting.com/uv-microscopy-designing-a-uv-condenser/ ? If so drop me a message and I can try and get some photos of what I did sorted out.