[1900k edition is in!] WTS: 1900k-5800k 5mm LED 95+ CRI

Works with Eneloops, although with greatly reduced runtime.

Or perhaps add some tiny buck/boost converter in it with ni-mh.

It is one PCB including LEDs. A PCB you stare at for some seconds and copper traces will come loose. Well designed in concept and firmware, but really s* electronics. Had to rewire this and that :person_facepalming:

Oh yeah the infamous FR1 boards. I love it when the copper trace blew open and spew all the hot flux and solder over my face every time the iron touches it. Luckily I’m wearing glasses so it didn’t get into my eyes.

Haha, well from a design engineer perspective, that sounds like it was designed just about perfect for the “cheap as possible, basically disposable” intent of this light! Obviously we at BLF want different, but we aren’t the intended market for this light. :SHRUG:

For the sake of identifying my LEDs I now add the SKU numbers (check on the OP).

Also Introducing RWN05-32K-C-EX. A love child hybrid between Yuji's 3200k and Nichia GS.

You get some of the High CRI and tint from Yuji's, and you get some of the power handling from Nichia GS (though not quite as good).

Here's Djozz's test compared with my standard 3400k R9080 and Nichia GS 4000k Rnn.

This one is much more expensive than the other 5mm LEDs. The price will be $9.5/20pcs (or $10.0 incl. handling fees).

The other SKU intruduced is RWN05-45K-B, the 2nd batch from the manufacturer whom I asked for 4000k emitter with 4500k CCT. The first batch ended up 5000k. This batch is closer to 4000k I asked, but ended up at 4500k instead, as tested by Djozz. The price will be $8.0 incl. handling fees.

CRI test pics are coming later.

Pics for RWN05-32K-C-EX (images taken by Djozz)

( ^ tint measurements were in the hotspot with led at 20mA)

Wow, those look nice! Are there any special requirements for running this 5mm LED at the higher end of its permissible amperage range?

To run (any LEDs on the menu) at high current, it’s preferable to:

  • Drive the LED in constant-current mode: As the LED heats up, it will take more current in a feedback loop until it either
    a) finds a thermal equilibrium or
    b) enters the thermal runaway and die.
    Constant current driver will ensure that the current will stay at a managable level, or at least keep the deterioration rate at the constant level (as opposed to rapidly)
  • You may need to enhance the heat dissipation on the cathode lead: The only way thru-hole LEDs can realistically dissipate heat is by thermal conduction through the cathode lead, where it connects to the base that the die is attached to. Based on my educated guesses, you could try:
    a) If the board has large ground plane where LEDs are to be soldered on, you should shorten the legs as much as possible to minimize the thermal resistance to the ground plane.
    b) If the board doesn’t have the ground plane, you may need to keep the legs as long as possible to maximize the surface area instead.
    c) Put the airflow on the back of the board to cool the ground plane.

Well, well...

I guess i need another package. Let's try those bada55 newcomers.

RWN05-45K-B (4500k 5mm) Tests by Djozz (20 mA)

The specs on both of those new emitters look great, especially that higher powered 3200k! Although the 3400K has been reported to perform well in the Nitecore Tube, this model should be slightly brighter and provide more assurance of a long life.

I wonder if Nitecore would be open to buying some to do a special edition Tube CRI, or Klarus on the similar Mi2…

Payment sent for 3200!


The US store Costco currently has 8-packs of “Life Gear” brand colored LED glow sticks for something like $20. They’re a bit larger than it seems like they need to be, but otherwise seem like a good value.

They are a bit unusual in that they have a flashlight on the opposite end of the glow stick. As expected, it’s a probably 6500+ K cool white, so I immediately started replacing them with warmer, high CRI LED’s. My kids are definitely enjoying them.

Here’s three of them with the flashlight end replaced with a rngwn 2300K, an rngwn 3400K, and a Yuji 5600K. There is a cheap reflector around the LED that causes a slight lopsided hotspot, but overall, the replacement LED’s are a huge improvement.

It took me a few minutes to figure out how to get them open. A screw locks the diffuser in place, so children won’t accidentally access the button cell batteries. The lens on the flashlight end seems to be glued in place, but it turns out a solid tap of the head against a hard surface will cause the entire light assembly to drop out the other end:

Blue spikes are high

These are normal blue 450nm based white leds so they have their 450nm spike which is needed as part of the visible spectrum to form the CCT/CRI, nothing extraordinary here compared to other leds, that is simply how they work. A few manufacturers recently came with new types of leds with 420nm base leds or with double dies, reducing the blue peak and filling in the cyan gap, but those are far from mainstream yet and none of the regular leds used by BLF have that.

I read about new Leds,and 420nm is near UV-A.
Marketing from bad to worse.

So what? It is the base led, the phosfor converts almost all of it to make the rest of the spectrum. Look at the spectrum of these leds and you see that the remaining small 420nm peak is hardly higher than the 420nm level of any incandescent or day light of the same colour temperature. Which implies that you are afraid of most natural light sources too (excluding maybe camp fire light) which is not very practical in life unless you are dedicated to spend your life inside under a edison bulb at 2500K.

I wish you will test one day these 420NM Leds
from unknown brand,so we’ll know if they’re safe.
Afraid? I already take enough natural daylight,don’t need more UV-A on night