[Updated] Surveying interest for 5mm high-power high-CRI LED

Honestly, this looks somehow interesting to me but, given the rated power/current/output versus unit price, I'm not that sure. Of course this has to do with the fact I'd buy the leds to later come up with some way to use them.

The special pricing is due only if at least 1500 pieces are reached, isn't it?

Yes, that’s my current plan right now.

To be clear, these prices are kind of worst-case analysis. I have no idea of what I will be running into, or if these LED needs to be “overengineered” so much that you need to slap a chip meant for 3w LEDs on a 5mm thru-hole body just to get 30lm at 70ma.

Interested in 10, more so if the price goes down.

In for 30

Hi, just to temper your expectations a bit.

I have inspected the Nicha GS more throroughly and I can say that achieving Nichia GS level of efficiency is quite challenging. Adding 95 CRI into the mix will lower the output even further. Therefore, I’d say I expect the final output to be around 18-24 lm in the best case, may be a bit less.

That’s very disappointing.

The number may seems disappointing, but going 95 CRI may cost up to 30-40 lm/w. That’s equivalent to 7-9 lms being lost compared to with 70 CRI.

That means to get 18-24lm on 95 CRI, you have to achieve 25-33lm on 70 CRI. On the other hand, maintaining 30lm at 95 CRI means you need to reduce the thermal density of the blue LED dies underneath the phosphor to 1/4th of the Nichis GS, while running at the same 70ma current.

I hope you aware of the implication of this. At that point, it may become too expensive.

With decades-old technology. :weary:

Are none of the improvements in LED technology over the past two decades applicable to this form factor? Are the mid and high power chips so different that they can’t be used for this style emitters? For instance, some CREE high power emitters can get over 300 lumens/watt when used at low power usage.

I really don’t want to be a downer on your project here. It is exciting that some manufacturer is willing to do some experimentation. And maybe we can get a little more experimentation done in the future. I’m only disappointed that we have to wait around for small improvements instead of being able to use much more advanced technology from other form factors, and apply it to this form factor.

It’s the same as Li-Ion batteries. For decades, only the 18650 size got serious development, and the other sizes lag behind. Now, some manufacturers are working on developing better 18350, 14500, etc. but they still haven’t reached the power density of 18650 cells. Not even close. We get excited when some sizes get a little better performance or a little better capacity. But, in the end, it’s still a disappointment knowing where it could/should be.

That’s exactly what we’re doing. Exactly what Nichia did with the GS. They threw in what amounts to 1w LED (or even bigger) inside and ran at 70ma. You have to account for limited thermal dissipation compared to the ideal scenario you derived the numbers from.

On the other hand, BOM is also another limiting factor. Sure, you might be able to throw an XP-E die in it and get great output, but at what cost?

The problem is not efficiency but maximum output. As rngwn already pointed out, dies from modern midpower leds can be used in 5mm leds which then have modern efficiency at low <20 lumen output, but the 5mm led form factor does not allow a very good heatpath so as soon as you go to a decent current, the output maxes out because of overheating.

For flashlight use, such a led could be worth the price. The advantage of 5mm leds is that when built into a flashlight it is mechanically virtually indestructable which is an interesting feature for some and worth paying 10 times the price of a normal 5mm led if the output doubles or triples.

If such LED still have enough mainstream appeals, that may be something to consider as a last resort. The final price may need to move up somewhat from the projected worst-case to reflect the increase in BOM.

Marketing High-CRI stuff always annoy me a bit. As many would say, It’s easier to make better marketting buzz selling 40lm (70CRI) than 30lm 95 CRI.

I do not really pay attention to whatever gamed or inflated figures. If I look for high quality light, a lot more measurements matter.

Marketing can be cleverly reworked to make people understand. Many years ago this happened in the computer CPU arena during the MHz versus actual performance race. Now single core performance is many times higher (and CPUs vastly more capable) despite the actual clockspeed hasn't really raised much (if at all), and certainly the MHz numbers aren't used or valid as marketing strategy.

For record, the projected “18-24 lm” @70ma is of course a nominal spec, where you can expect a decent lifespan out of it. This is important in a mainstream application where the LED is expected to be constantly turned on for an extended period.

You might get away with running these LEDs at 100ma or even 150ma and get extra lumen out of these without blowing up outright. But you might end up with less than 1000 hours of useful lifetime. For flashlights, this may be acceptable to some.

Not-so-good news:

The company has said that altering the power rating of such LEDs is beyond what they are comfortable to do.

They might technically able to pull that off, but according to them, they simply cannot arbitrarily alter the construction parameters (i.e. the cup size, the chip size, the amount of phosphors). The main issues being that the amount of phosphors will need to be tightly controlled. Altering the amount of phosphors will create the quality control problems. For example, if they overfill the phosphors, the epoxy will crack during the curing process. If they underfill it, the CRI and tint will suffer. Seeing their reason, I see no point pushing them to do that for me.

Hey, at least they don’t go ahead and try knowing that it’s not practical, then deliver defective products to me.

My only remaining saving throw is perhaps Nichia itself. I don’t know if it’s even financially feasible (or even possible to begin with) to ask Nichia to basically customize a hi-CRI verison of the 570GS itself for me. I don’t know what will be the MOQ for those. All in all, I got to have more fund than what I have now before hoping to do such thing (in case they require an astronomical amount of MOQ and/or unit cost).

For now I considered this project shelved for the near future.

I’d in for 10 pcs first.
Thank you.

Is this LED suitable to put it into Nitecore Tube instead original bluish led?

100% yes I'd say. By the way, someone already did it with a Yuji led.

Hello, I managed to find such LED after all. It won’t be the “be-all, end-all” for the 5mm leds, but it’s a nice compromise between the power and tint of the Yuji and Nichia GS.

Check out the RWN05-32K-C-EX in this thread:

The price is at $10 for 20pcs (incl. fees).

Thank you for the reference!

I didn’t know about the current limiting resistor. Thus far, all my Tubes with nrgwn’s LEDs work, but they are probably not used as much as to show visible degradadtion by now. I’ll take one and try this 33 Ohms mod.

Guess the new LEDs can be used in unmodified Tubes on a regular basis. Ordered them and looking forward to a nice-tinted Tube that can be used on high without hestitation.