Changing out LEDs in solar garden lights - why does white work and green not?

[tooki]

It can be difficult to find green InGaN diffused LEDs. I've seen some on eBay, but I'm suspicious about whether they're really the old yellow-green type. The Kingbright WP7083ZGD/G is the only one I know of which is widely available, but it isn't cheap.
https://www.digikey.com.au/product-detail/en/kingbright/WP7083ZGD-G/754-1890-ND/3084179

It's certainly easy enough to find out: look at the wavelength. A green LED under about 540nm is a modern "true" green with InGaN or GaN chemistry, if it's 550 or above it's the old yellow-green. Both kinds are readily available on eBay and Ali as well as DK, etc. The vendors are usually much more likely to list the wavelength than the chemistry, so that's what I go by.

[Zero999]

It can be difficult to find green InGaN diffused LEDs. I've seen some on eBay, but I'm suspicious about whether they're really the old yellow-green type. The Kingbright WP7083ZGD/G is the only one I know of which is widely available, but it isn't cheap.
https://www.digikey.com.au/product-detail/en/kingbright/WP7083ZGD-G/754-1890-ND/3084179

It's certainly easy enough to find out: look at the wavelength. A green LED under about 540nm is a modern "true" green with InGaN or GaN chemistry, if it's 550 or above it's the old yellow-green. Both kinds are readily available on eBay and Ali as well as DK, etc. The vendors are usually much more likely to list the wavelength than the chemistry, so that's what I go by.

I agree and for vendors such as Digi-Key, I do trust specifications such as the wavelength, although I do check the data sheet as well, but I'm often hesitant to rely on eBay, especially from vendors who also sell unrelated items.

[Jwillis]

Slap a reflector collimator on top  and that should spread out the light better. Depends on how much room you have to work with.

[tooki]

It can be difficult to find green InGaN diffused LEDs. I've seen some on eBay, but I'm suspicious about whether they're really the old yellow-green type. The Kingbright WP7083ZGD/G is the only one I know of which is widely available, but it isn't cheap.
https://www.digikey.com.au/product-detail/en/kingbright/WP7083ZGD-G/754-1890-ND/3084179

It's certainly easy enough to find out: look at the wavelength. A green LED under about 540nm is a modern "true" green with InGaN or GaN chemistry, if it's 550 or above it's the old yellow-green. Both kinds are readily available on eBay and Ali as well as DK, etc. The vendors are usually much more likely to list the wavelength than the chemistry, so that's what I go by.

I agree and for vendors such as Digi-Key, I do trust specifications such as the wavelength, although I do check the data sheet as well, but I'm often hesitant to rely on eBay, especially from vendors who also sell unrelated items.

So buy from the electronics vendors, or from specialized vendors that just do LEDs.

[stryker]

Thanks for everything guys.  I've gone through the parts bin and found I have some of these 10mm green waterclear LEDs that would appear to do the job at least based on V_F, however in reality none I've tested show anything higher than V_F 2.5-3.0V, nowhere near the advertised 3.2-3.4V.

What test current did you use? The 3.2V to 3.4V is normally specified at a relatively high current such as 20mA. If you were using the diode test on a multimeter, then you should expect a much lower voltage.

I'm using one of those ATmega328 based component testers that come as a kit, similar to this one on banggood to get a guide on the V_F value on the LEDs.

[stryker]

Slap a reflector collimator on top  and that should spread out the light better. Depends on how much room you have to work with.

This looks like a workable idea however all I've found focus the beam rather than spread it more diffusely...do you have a source & product in mind that might suit here?  Thanks

[Zero999]

It can be difficult to find green InGaN diffused LEDs. I've seen some on eBay, but I'm suspicious about whether they're really the old yellow-green type. The Kingbright WP7083ZGD/G is the only one I know of which is widely available, but it isn't cheap.
https://www.digikey.com.au/product-detail/en/kingbright/WP7083ZGD-G/754-1890-ND/3084179

It's certainly easy enough to find out: look at the wavelength. A green LED under about 540nm is a modern "true" green with InGaN or GaN chemistry, if it's 550 or above it's the old yellow-green. Both kinds are readily available on eBay and Ali as well as DK, etc. The vendors are usually much more likely to list the wavelength than the chemistry, so that's what I go by.

I agree and for vendors such as Digi-Key, I do trust specifications such as the wavelength, although I do check the data sheet as well, but I'm often hesitant to rely on eBay, especially from vendors who also sell unrelated items.

So buy from the electronics vendors, or from specialized vendors that just do LEDs.

I'm more cynical about eBay for buying diffused InGaN green LEDs, than the water clear type, simply because they're not many options from the usual trusted supplies such as Digi-Key, Mouser, TME, RS Components etc. I wouldn't buy anything, without  asking the seller first. I know it's easy to return things, but I'd rather not have the hassle.

tooki,
What sort of purple are you after? 410nm  to 420nm gives a nice violet colour, but is quite close to UV and is quite harsh to look at.

The only manufacturer of phosphor converted purple LEDs, I'm aware of is OptoSupply, but I've never tried them. Look at part numbers: OSCD4L5111A (Lavender) and OSK64L5111A (Sakura). They also make a range of other colours.
http://www.optosupply.com/uppic/201683604722.pdf
http://www.optosupply.com/uppic/201682360345.pdf

They are available from Rapid Electronics, TME and some eBay sellers who appear to be buying from the aforementioned and adding a profit margin.

PC purple is really what I'm after. I found a string of battery-powered christmas lights that have them, and I wish I could find those separately. As I said, even if they do exist on eBay/Ali, they're impossible to find, since the word "purple" has been hijacked by all the UV LED listings.

As for OptoSupply, I think I did come across those once, I just don't know of any vendor that carries them that can ship to Switzerland with customs pre-cleared (DDP incoterms). :/

Can you now buy from TME, Mouser or Digi-Key? More expensive, than eBay and you need to put in a reasonable size order to get free shipping, but at least you know what you're getting.

Kingbright offers special colours pink, purple, special blues. I wish these would be more prevalent, aren't we getting bored with LED colours.

They only do the pink ones in the common 5mm T-1 ³/₄ package. It would be good to get the other colours in 5mm and some diffused options would also be nice.

Huh? Look at the link. The pink ones also come in T1 and various SMD packages, as do many other colors.

Yes, that's what I meant. I misworded it.

I, too, am baffled at why diffused LEDs are so rare now.

Diffused InGaN green is difficult to find, but I can't fine diffused phosphor converted yellow anywhere. The only yellow LEDs, available in a diffused package are the cappy old dim ones.

I don't know why. Perhaps it's because they can make them seem brighter with a narrow beam?

It's incredible to me that the blue LED went from being the elusive color, hard to make and not very bright, to the one that not only became the brightest chemistry possible, but also became the foundation for all the phosphor-converted LEDs that we now rely on. (I have heard mention of phosphor-converted-UV white LEDs, but I don't think I've encountered one in the wild, and strongly suspect they're extinct now.)

Yes, photometrically speaking, i.e. in terms of raw power out vs power in, deep blue is the most efficient LED colour, followed by red. Pure InGaN green is more efficient, than blue and red, in terms of lumens/Watt, but that's only because the human eye is much more sensitive to green and relatively insensitive to blue. Photometrically, InGaN green is much less efficient than red and blue.

I believe larger bandgap LEDs will generally be more efficient, assuming everything else is equal. I think it's due to Ohm's law. To produce red, a lower bandgap, thus voltage drop diode is required, hence relatively high Ohmic losses, compared to a larger bandgap diode, with a higher forward voltage. In theory more efficient infrared LEDs could be made if it was possible for a higher bandgap semiconductor could be made to emit two photons, per excitation.

The quest is still on for more efficient green and yellow LEDs. InGaN is good enough for blue-green to pure green, but deep InGaN blue/violet and a phosphor is more efficient for greenish-yellow through to pure yellow and amber region.

[Jwillis]

Slap a reflector collimator on top  and that should spread out the light better. Depends on how much room you have to work with.

This looks like a workable idea however all I've found focus the beam rather than spread it more diffusely...do you have a source & product in mind that might suit here?  Thanks

 

If this is the kind of thing that may work ,you want  120^o. 180^o exist but seem rare.
You've probably already converted your LEDs . But why wouldn't a green cylindrical filter placed on the inside work and just use the white lights. You can get clear plastic sheets of what ever colour you like. Cut to length and roll into a cylinder and slip it inside.   

[Zero999]

Thanks for everything guys.  I've gone through the parts bin and found I have some of these 10mm green waterclear LEDs that would appear to do the job at least based on V_F, however in reality none I've tested show anything higher than V_F 2.5-3.0V, nowhere near the advertised 3.2-3.4V.

What test current did you use? The 3.2V to 3.4V is normally specified at a relatively high current such as 20mA. If you were using the diode test on a multimeter, then you should expect a much lower voltage.

I'm using one of those ATmega328 based component testers that come as a kit, similar to this one on banggood to get a guide on the V_F value on the LEDs.

Does it not display the test current on the screen? It's probably quite low.

I Googled the M12864 and found this document, containing a schematic. The test pins are connected to some I/O ports, via 470k and 680R resistors, presumably to limit the current to different levels. Going by the forward voltages you've given, the supply voltage of the MCU is 5V and the 680R resistors, the diode test current will be under 4mA.
http://www.n5dux.com/ham/files/pdf/M12864%20Transistor%20Tester%20Instructions.pdf

[stryker]

Does it not display the test current on the screen? It's probably quite low.

It doesn't.  All it confirms is the battery and supply voltages at startup

the supply voltage of the MCU is 5V and the 680R resistors, the diode test current will be under 4mA.
http://www.n5dux.com/ham/files/pdf/M12864%20Transistor%20Tester%20Instructions.pdf

Thanks for the link.  IIRC there was a similar doc for the one I built.  It's been a very handy little swiss army knife of the workbench to have around.

[tooki]

It can be difficult to find green InGaN diffused LEDs. I've seen some on eBay, but I'm suspicious about whether they're really the old yellow-green type. The Kingbright WP7083ZGD/G is the only one I know of which is widely available, but it isn't cheap.
https://www.digikey.com.au/product-detail/en/kingbright/WP7083ZGD-G/754-1890-ND/3084179

It's certainly easy enough to find out: look at the wavelength. A green LED under about 540nm is a modern "true" green with InGaN or GaN chemistry, if it's 550 or above it's the old yellow-green. Both kinds are readily available on eBay and Ali as well as DK, etc. The vendors are usually much more likely to list the wavelength than the chemistry, so that's what I go by.

I agree and for vendors such as Digi-Key, I do trust specifications such as the wavelength, although I do check the data sheet as well, but I'm often hesitant to rely on eBay, especially from vendors who also sell unrelated items.

So buy from the electronics vendors, or from specialized vendors that just do LEDs.

I'm more cynical about eBay for buying diffused InGaN green LEDs, than the water clear type, simply because they're not many options from the usual trusted supplies such as Digi-Key, Mouser, TME, RS Components etc. I wouldn't buy anything, without  asking the seller first. I know it's easy to return things, but I'd rather not have the hassle.

Availability on DK, Mouser, etc. is no predictor of availability on eBay/Ali. There are tons of components the big guys carry that you never, ever see on eBay/Ali, but then there are also tons of components that are easy to find on eBay/Ali but that none of the big guys carry at all.

If you see an LED being sold just as “green” without a wavelength or voltage spec, then you may or may not get a modern LED. But if they say it’s low-500s nm wavelength, then it is. I’m a bit puzzled as to why you have such extraordinary skepticism about buying LEDs on eBay/Ali: if anything, they’re one of the items that makes the most sense to get there, since the most critical specs (color, voltage, package type) are easily verified without special skill or equipment. It’s not like, say, op-amps or transistors, where fakes may work, but don’t meet specs in ways that require expertise and equipment to verify.

For what it’s worth, I’ve had no trouble whatsoever with eBay/Ali LEDs, with the understanding that I may not going to get carefully binned LEDs whose color matches perfectly.

[Zero999]

It can be difficult to find green InGaN diffused LEDs. I've seen some on eBay, but I'm suspicious about whether they're really the old yellow-green type. The Kingbright WP7083ZGD/G is the only one I know of which is widely available, but it isn't cheap.
https://www.digikey.com.au/product-detail/en/kingbright/WP7083ZGD-G/754-1890-ND/3084179

It's certainly easy enough to find out: look at the wavelength. A green LED under about 540nm is a modern "true" green with InGaN or GaN chemistry, if it's 550 or above it's the old yellow-green. Both kinds are readily available on eBay and Ali as well as DK, etc. The vendors are usually much more likely to list the wavelength than the chemistry, so that's what I go by.

I agree and for vendors such as Digi-Key, I do trust specifications such as the wavelength, although I do check the data sheet as well, but I'm often hesitant to rely on eBay, especially from vendors who also sell unrelated items.

So buy from the electronics vendors, or from specialized vendors that just do LEDs.

I'm more cynical about eBay for buying diffused InGaN green LEDs, than the water clear type, simply because they're not many options from the usual trusted supplies such as Digi-Key, Mouser, TME, RS Components etc. I wouldn't buy anything, without  asking the seller first. I know it's easy to return things, but I'd rather not have the hassle.

Availability on DK, Mouser, etc. is no predictor of availability on eBay/Ali. There are tons of components the big guys carry that you never, ever see on eBay/Ali, but then there are also tons of components that are easy to find on eBay/Ali but that none of the big guys carry at all.

If you see an LED being sold just as “green” without a wavelength or voltage spec, then you may or may not get a modern LED. But if they say it’s low-500s nm wavelength, then it is. I’m a bit puzzled as to why you have such extraordinary skepticism about buying LEDs on eBay/Ali: if anything, they’re one of the items that makes the most sense to get there, since the most critical specs (color, voltage, package type) are easily verified without special skill or equipment. It’s not like, say, op-amps or transistors, where fakes may work, but don’t meet specs in ways that require expertise and equipment to verify.

For what it’s worth, I’ve had no trouble whatsoever with eBay/Ali LEDs, with the understanding that I may not going to get carefully binned LEDs whose color matches perfectly.

You have a point about availability of items from eBay/Ali and the usual distributors.

I'm not that sceptical about buying LEDs from eBay/Ali. As a general rule, I believe most sellers are honest and are not trying to rip people off. I would not hesitate in buying ordinary red, yellow, blue water clear/diffused or transparent green LEDs from eBay/Ali. The issue is, when buying from a distributor, they give the part number and data sheet, so I can be more sure about what I'm getting. Most Bay/Ali sellers don't have the part number or data sheet, so it could be anything. Quite often the specification listed will also be inconsistent, such as specifying the same intensity for all LEDs in the store, irrespective of their colour. You're right, any idiot can tell a red LED from a green one, but a layperson wouldn't necessarily know the relationship between colour and wavelength.
https://www.ebay.co.uk/itm/Diffused-LEDs-1-8mm-3mm-5mm-Red-Blue-White-Green-Yellow-Orange-1st-CLASS-POST/121116641268?_trkparms=aid%3D1110002%26algo%3DSPLICE.SOI%26ao%3D1%26asc%3D20200116171858%26meid%3Df494b5f861284305985eed5123449fff%26pid%3D100008%26rk%3D3%26rkt%3D12%26sd%3D291252890071%26itm%3D121116641268%26pmt%3D1%26noa%3D0%26pg%3D2047675%26algv%3DPromotedSellersOtherItemsV2%26brand%3DUnbranded&_trksid=p2047675.c100008.m2219

Regarding searching for purple, rather than UV LEDs on eBay. Try adding -UV to the search string. I find I get more phosphor converted purple LEDs that way, although it doesn't eliminate violet or UV sold as purple, because the seller doesn't know and just puts purple in the listing, as that's the colour they see when they're powered.

Does it not display the test current on the screen? It's probably quite low.

It doesn't.  All it confirms is the battery and supply voltages at startup

the supply voltage of the MCU is 5V and the 680R resistors, the diode test current will be under 4mA.
http://www.n5dux.com/ham/files/pdf/M12864%20Transistor%20Tester%20Instructions.pdf

Thanks for the link.  IIRC there was a similar doc for the one I built.  It's been a very handy little swiss army knife of the workbench to have around.

I have one of the more expensive Peak component testers I bought awhile ago, which does show the test current.
https://docs.rs-online.com/36ac/0900766b803abf3e.pdf

If you have an oscilloscope you can measure it, by connecting a low value resistor, say 10R in series with the LED and pushing the test button and looking at the waveform. The current can be calculated using Ohm's law.

[tooki]

You have a point about availability of items from eBay/Ali and the usual distributors.

I'm not that sceptical about buying LEDs from eBay/Ali. As a general rule, I believe most sellers are honest and are not trying to rip people off. I would not hesitate in buying ordinary red, yellow, blue water clear/diffused or transparent green LEDs from eBay/Ali. The issue is, when buying from a distributor, they give the part number and data sheet, so I can be more sure about what I'm getting. Most Bay/Ali sellers don't have the part number or data sheet, so it could be anything. Quite often the specification listed will also be inconsistent, such as specifying the same intensity for all LEDs in the store, irrespective of their colour. You're right, any idiot can tell a red LED from a green one, but a layperson wouldn't necessarily know the relationship between colour and wavelength.
https://www.ebay.co.uk/itm/Diffused-LEDs-1-8mm-3mm-5mm-Red-Blue-White-Green-Yellow-Orange-1st-CLASS-POST/121116641268?_trkparms=aid%3D1110002%26algo%3DSPLICE.SOI%26ao%3D1%26asc%3D20200116171858%26meid%3Df494b5f861284305985eed5123449fff%26pid%3D100008%26rk%3D3%26rkt%3D12%26sd%3D291252890071%26itm%3D121116641268%26pmt%3D1%26noa%3D0%26pg%3D2047675%26algv%3DPromotedSellersOtherItemsV2%26brand%3DUnbranded&_trksid=p2047675.c100008.m2219

As I’ve said at least once before: if you’re this skeptical (which you are, despite your claims to the contrary, since your explanations are just trying to justify that skepticism), then buy from the eBay/Ali vendors that specialize in LEDs. You’re not usually going to get the kind of datasheets you get from name brands that literally charge 10-20 times as much, but you will get all the key specs. (And on Ali especially, some are the actual web storefronts for manufacturers, and their websites sometimes have actual datasheets.) I don’t know why you keep trying to convince me that there are eBay vendors out there that have crappy specs; I never claimed otherwise. But that doesn’t mean that good vendors don’t exist!!!!

Regarding searching for purple, rather than UV LEDs on eBay. Try adding -UV to the search string. I find I get more phosphor converted purple LEDs that way, although it doesn't eliminate violet or UV sold as purple, because the seller doesn't know and just puts purple in the listing, as that's the colour they see when they're powered.

Yeah that’s kinda web basics, innit? Believe me, I’ve done extremely careful search term pruning, but never found PC purple LEDs being sold from China, other than built into Christmas lights and the like. Invariably, they were actually UV. The only discrete PC purple ones that came up were from western vendors selling them at outrageous prices. It’s been a while since I searched for them, so maybe it’s changed now, but I have extremely good google-fu (and eBay/Ali), and I could not find any.

[Zero999]

You have a point about availability of items from eBay/Ali and the usual distributors.

I'm not that sceptical about buying LEDs from eBay/Ali. As a general rule, I believe most sellers are honest and are not trying to rip people off. I would not hesitate in buying ordinary red, yellow, blue water clear/diffused or transparent green LEDs from eBay/Ali. The issue is, when buying from a distributor, they give the part number and data sheet, so I can be more sure about what I'm getting. Most Bay/Ali sellers don't have the part number or data sheet, so it could be anything. Quite often the specification listed will also be inconsistent, such as specifying the same intensity for all LEDs in the store, irrespective of their colour. You're right, any idiot can tell a red LED from a green one, but a layperson wouldn't necessarily know the relationship between colour and wavelength.
https://www.ebay.co.uk/itm/Diffused-LEDs-1-8mm-3mm-5mm-Red-Blue-White-Green-Yellow-Orange-1st-CLASS-POST/121116641268?_trkparms=aid%3D1110002%26algo%3DSPLICE.SOI%26ao%3D1%26asc%3D20200116171858%26meid%3Df494b5f861284305985eed5123449fff%26pid%3D100008%26rk%3D3%26rkt%3D12%26sd%3D291252890071%26itm%3D121116641268%26pmt%3D1%26noa%3D0%26pg%3D2047675%26algv%3DPromotedSellersOtherItemsV2%26brand%3DUnbranded&_trksid=p2047675.c100008.m2219

As I’ve said at least once before: if you’re this skeptical (which you are, despite your claims to the contrary, since your explanations are just trying to justify that skepticism), then buy from the eBay/Ali vendors that specialize in LEDs. You’re not usually going to get the kind of datasheets you get from name brands that literally charge 10-20 times as much, but you will get all the key specs. (And on Ali especially, some are the actual web storefronts for manufacturers, and their websites sometimes have actual datasheets.) I don’t know why you keep trying to convince me that there are eBay vendors out there that have crappy specs; I never claimed otherwise. But that doesn’t mean that good vendors don’t exist!!!! 

Of course I'm sceptical of anyone selling something very cheaply. I would be silly not to be. If I just want some generic, LEDs, not fussy about the colour or chemistry, then eBay/Ali seem ideal, but as I'm after something more specific, it's more tricky.

I thought the store I linked to above looked decent, as they specialise in LEDs, until  I noticed they specify the same brightness for all LED colours. Do you know of any sellers of InGaN diffuses green LEDs, who are good and will not sell me the crappy yellow-green ones?

Regarding searching for purple, rather than UV LEDs on eBay. Try adding -UV to the search string. I find I get more phosphor converted purple LEDs that way, although it doesn't eliminate violet or UV sold as purple, because the seller doesn't know and just puts purple in the listing, as that's the colour they see when they're powered.

Yeah that’s kinda web basics, innit? Believe me, I’ve done extremely careful search term pruning, but never found PC purple LEDs being sold from China, other than built into Christmas lights and the like. Invariably, they were actually UV. The only discrete PC purple ones that came up were from western vendors selling them at outrageous prices. It’s been a while since I searched for them, so maybe it’s changed now, but I have extremely good google-fu (and eBay/Ali), and I could not find any.

Firstly few LEDs sold as UV from eBay/Ali are actually UV. They're normally deep violet, around the 400nm to 420nm mark. This is not an issue in most applications, as it will still make most florescent materials glow brightly and the human eye isn't that sensitise to such wavelengths anyway. It's also much safer than real UV.

Secondly, how do you know the Christmas lights aren't actually deep violet, rather than phosphor converted?

Finally, have you actually tried messaging your favourite LED seller, asking them if it's something they can source?

[tooki]

Of course I'm sceptical of anyone selling something very cheaply. I would be silly not to be. If I just want some generic, LEDs, not fussy about the colour or chemistry, then eBay/Ali seem ideal, but as I'm after something more specific, it's more tricky.

But InGaN green LEDs aren't anything remotely exotic these days. If anything, they've already significantly displaced traditional green LEDs. In every LED assortment I've ordered from eBay, the green have been InGaN. (But of course, they're also usually either SMD, or THT with stupid clear housing.)

I thought the store I linked to above looked decent, as they specialise in LEDs, until  I noticed they specify the same brightness for all LED colours. Do you know of any sellers of InGaN diffuses green LEDs, who are good and will not sell me the crappy yellow-green ones?

That's a general component vendor who sells a few LEDs. Not an LED specialist by a long shot.

With that said, if you look at their other LEDs, they are listing different brightnesses per color. It is (distantly) possible that the diffused ones, being intended as indicators, were actually selected to have the same brightness.


The best LED vendor I've found so far is https://chelead.aliexpress.com/

Everything I've gotten from them has matched the specs perfectly, packaging was good, and the selection is spectacular (nobody else carries even distantly as many white LED displays, for example, nor the 1-2mm thin LEDs). Anyhow, this store generally refers to the traditional yellow-green ones as "green", "light green", or "green 570nm", and the InGaN ones as "pure green". They have diffused InGaN green LEDs in diffused white (milky), diffused green, with and without flanges, in various sizes…

Even so, they suck at listing the chemistry (they list everything as being AlGaInP, which is plainly incorrect). But the colors and voltages have all been absolutely correct in my experience. (In general, if a vendor is savvy enough to distinguish between yellow-green and other shades of green, and lists the wavelengths, you're not going to get surprised.)

Another vendor that looks promising, but which I've never ordered from yet, is the one with the hands-down most bizarre name of any electronics supplier I've ever seen: https://molesmell.aliexpress.com

(I wonder how much that name costs them in sales. And if they even realize what it means in English.  )

Firstly few LEDs sold as UV from eBay/Ali are actually UV. They're normally deep violet, around the 400nm to 420nm mark. This is not an issue in most applications, as it will still make most florescent materials glow brightly and the human eye isn't that sensitise to such wavelengths anyway. It's also much safer than real UV.

Well, they're usually listed as 395-400nm. Better vendors have quite granular wavelengths, from 400nm all the way down to even the mid-200nm, though the price increases dramatically the farther you go.

What counts as UV and what as violet seems to be a matter of opinion, as one sees various different cutoffs (heck, just within wikipedia, there's massive disagreement between pages). Without question, 395nm is right up at the longest wavelengths of what can be considered UV. But it's certainly not visible violet!


Interestingly, if you search for e.g. 430nm LEDs, which are definitely in the true violet range, you find plenty — but only the high-wattage kind for illumination and glue curing. But never small indicator-style ones.

Secondly, how do you know the Christmas lights aren't actually deep violet, rather than phosphor converted?

Because I own some, and they're nothing like the UV or pure violet LEDs. It's not the feeble glow of UV LEDs, it's a rich, bright purple. (And one that is easily split by a prism into its constituent blue and red.)

The other clue is in pictures: the UV LED epoxy tends to glow a pale whitish blue, while pictures of the PC purple tend to show the epoxy as being more purple to pink. (Just like in real life.)

Finally, have you actually tried messaging your favourite LED seller, asking them if it's something they can source?

Nah, since I'm just a hobbyist. I wouldn't be ordering enough volume to warrant them custom ordering them. (If they had them at hand, they'd already list them.)

[I wanted a rude username]

Another vendor that looks promising, but which I've never ordered from yet, is the one with the hands-down most bizarre name of any electronics supplier I've ever seen: https://molesmell.aliexpress.com

(I wonder how much that name costs them in sales. And if they even realize what it means in English.  )

Bet you an Ittinger it's actually good for sales. Most sellers have forgettable names, but if ever see Molesmell in search results even once, you'll remember them. 

(These sellers are too expensive for me ... usually go with Songxin Light Store, but that's mostly by virtue of them often having the best price, although they have been reliable so far.)

[Chris56000]

Hi!

At the risk of causing howls of derision, is it possible to make "brown" and "grey" light sources to complete the full set of Electronic Engineering colours?!

Chris Williams

[tooki]

Hi!

At the risk of causing howls of derision, is it possible to make "brown" and "grey" light sources to complete the full set of Electronic Engineering colours?!

Well, brown is just dark orange, and gray is dark white.


Technology Connections just did a video on this, in fact:

[stryker]

Hi again brains trust,

I found this Instructable by Becky Stern that illustrated hot glue as one of the ideas for an LED diffuser.

I'm pleased to say that combining that, with a square of aluminium foil tucked in around the base, shiny side up (Aldi Goliath brand, nothing but the best) and the 10mm green clear LEDs with the 3V V_F I linked earlier, and the result is pretty much what I was looking for.


The hot snot expands the beam of light to slightly past the width of the base, and the shiny foil bounces that back to fill the cylinder nicely.  Without the foil, only the top half is visibly lit (with much of the light absorbed by the black plastic base that doubles as the top of the plastic garden spike).  With the variability of the LEDs in that parts bin even from just one order I'll choose those with the highest measured V_F for these.

Thanks to everyone for the helpful discussion and suggestions.  I've learned a heap in this thread.

One down, just 29 to go...cheers!
Geoff

[Zero999]

Of course I'm sceptical of anyone selling something very cheaply. I would be silly not to be. If I just want some generic, LEDs, not fussy about the colour or chemistry, then eBay/Ali seem ideal, but as I'm after something more specific, it's more tricky.

But InGaN green LEDs aren't anything remotely exotic these days. If anything, they've already significantly displaced traditional green LEDs. In every LED assortment I've ordered from eBay, the green have been InGaN. (But of course, they're also usually either SMD, or THT with stupid clear housing.)

Yes, it's the InGaN green diffused, I'm sceptical about. I'll just have to order some and give them ago.

I thought the store I linked to above looked decent, as they specialise in LEDs, until  I noticed they specify the same brightness for all LED colours. Do you know of any sellers of InGaN diffuses green LEDs, who are good and will not sell me the crappy yellow-green ones?

That's a general component vendor who sells a few LEDs. Not an LED specialist by a long shot.

With that said, if you look at their other LEDs, they are listing different brightnesses per color. It is (distantly) possible that the diffused ones, being intended as indicators, were actually selected to have the same brightness.

The best LED vendor I've found so far is https://chelead.aliexpress.com/

Everything I've gotten from them has matched the specs perfectly, packaging was good, and the selection is spectacular (nobody else carries even distantly as many white LED displays, for example, nor the 1-2mm thin LEDs). Anyhow, this store generally refers to the traditional yellow-green ones as "green", "light green", or "green 570nm", and the InGaN ones as "pure green". They have diffused InGaN green LEDs in diffused white (milky), diffused green, with and without flanges, in various sizes…

Even so, they suck at listing the chemistry (they list everything as being AlGaInP, which is plainly incorrect). But the colors and voltages have all been absolutely correct in my experience. (In general, if a vendor is savvy enough to distinguish between yellow-green and other shades of green, and lists the wavelengths, you're not going to get surprised.)

Another vendor that looks promising, but which I've never ordered from yet, is the one with the hands-down most bizarre name of any electronics supplier I've ever seen: https://molesmell.aliexpress.com

(I wonder how much that name costs them in sales. And if they even realize what it means in English.  )

Thanks. Do you know of any ebay sellers? I don't have an Alie account, but will obviously get one, if needs be.

Firstly few LEDs sold as UV from eBay/Ali are actually UV. They're normally deep violet, around the 400nm to 420nm mark. This is not an issue in most applications, as it will still make most florescent materials glow brightly and the human eye isn't that sensitise to such wavelengths anyway. It's also much safer than real UV.

Well, they're usually listed as 395-400nm. Better vendors have quite granular wavelengths, from 400nm all the way down to even the mid-200nm, though the price increases dramatically the farther you go.

What counts as UV and what as violet seems to be a matter of opinion, as one sees various different cutoffs (heck, just within wikipedia, there's massive disagreement between pages). Without question, 395nm is right up at the longest wavelengths of what can be considered UV. But it's certainly not visible violet!

Interestingly, if you search for e.g. 430nm LEDs, which are definitely in the true violet range, you find plenty — but only the high-wattage kind for illumination and glue curing. But never small indicator-style ones.

Secondly, how do you know the Christmas lights aren't actually deep violet, rather than phosphor converted?

Because I own some, and they're nothing like the UV or pure violet LEDs. It's not the feeble glow of UV LEDs, it's a rich, bright purple. (And one that is easily split by a prism into its constituent blue and red.)

Yes, UV is not quite that well defined, since the sensitivity depends on the viewer. It probably also depends on the application. There are other areas of the EM spectrum which are divided depending on the origin of the radiation or application. One example is gamma, vs X-rays, of the same energy, the former being from atomic decay and the latter from the rapid deceleration of an electron.

The other clue is in pictures: the UV LED epoxy tends to glow a pale whitish blue, while pictures of the PC purple tend to show the epoxy as being more purple to pink. (Just like in real life.)

A true UV LED should light a dim pinkish/white colour, rather than purple, when vied with the naked eye, obviously cameras may differ and high quality UV LEDs use an epoxy which doesn't fluoresce.

Finally, have you actually tried messaging your favourite LED seller, asking them if it's something they can source?

Nah, since I'm just a hobbyist. I wouldn't be ordering enough volume to warrant them custom ordering them. (If they had them at hand, they'd already list them.)

If it's just for hobbyist purposes, then how about buying some cheap Christmas lights and dismantling them?

Hi again brains trust,

I found this Instructable by Becky Stern that illustrated hot glue as one of the ideas for an LED diffuser.

I'm pleased to say that combining that, with a square of aluminium foil tucked in around the base, shiny side up (Aldi Goliath brand, nothing but the best) and the 10mm green clear LEDs with the 3V V_F I linked earlier, and the result is pretty much what I was looking for.


The hot snot expands the beam of light to slightly past the width of the base, and the shiny foil bounces that back to fill the cylinder nicely.  Without the foil, only the top half is visibly lit (with much of the light absorbed by the black plastic base that doubles as the top of the plastic garden spike).  With the variability of the LEDs in that parts bin even from just one order I'll choose those with the highest measured V_F for these.

Thanks to everyone for the helpful discussion and suggestions.  I've learned a heap in this thread.

One down, just 29 to go...cheers!
Geoff

That looks good. I'm not sure if the forward voltage really matters. InGaN has superior efficiency to the older yellow-green LEDs. A higher forward voltage drop might be slightly brighter, as it will have a higher power dissipation or it might just emit a slightly more bluish green. I suggest you experiment.