LED test: Osram LE UW S2WP (3V, quad die)

Flashlight Modding and DIY Parts
1

Salvaged this Osram S2WP from a Nightwatch NI40, which had some issues. Thanks to Neal for sending it in for review, but other than the nice beefy host, the driver acted quite erratic and I never wanted to touch it.

The emitter is interesting. It’s a quad die LED which can was configured for parallel 4x 3V operation on the Nightwatch MCPCB. However, the forward voltage is very low, under 3V at most operating currents, so for optimal performance it needs preferably a buck driver for use with Li-Ion. A linear driver would waste a lot of power, since the voltage difference is large. Series config at 12V would be best, but at least this MCPCB didn’t seem to offer that possibility.

I don’t think this emitter fits on any ubiquitous solder pads as it’s smaller than an XHP70 but larger than an XHP50. This is more of a curiosity without many flashlights applications that I could think of.

Test was done on an Al heatsink. MCPCB was put on top of a graphite thermal pad and clamped securely down with two screws. Tests were done in an integrating sphere.

A single emitter is rated up to 5 amps of current at 25°C so four in parallel could theoretically go up to 20A and still be in spec. My PSU on only does 10A and since 4S config wasn’t doable, there is still lots more headroom above what I was able to test.

According to my results and the datasheet, this sample is flux bin SA, which should do 1800-2240 lumens when driven at 1400mA per emitter (5600mA total). This LED did 1830 lumens at 5600mA with an efficacy of 111 lumens per watt.

Datasheet: https://dammedia.osram.info/media/resource/hires/osram-dam-7378054/LE%20UW%20S2WP_EN.pdf

Voltage: 3V (four dies in parallel)
CRI: not specified (this usually means CRI Ra 70 at most)
Tint: 0.321x 0.327y typical on CIE1931 which equals 6073K with a duv value of –0.0020 (although can go anywhere between 5650K and 7200K with a duv spread of –0,0126 to 0.0088 as per chromaticity group BQ)
Flux bin: SA (1800-2240 lumens at 1400mA per emitter, 25°C)
Emitter size: 5.75 x 6.38 mm

Doesn’t look too good. Huge tint bin with probably atrocious CRI. Not for color critical applications :smiley:

Output and forward voltage

Table with all the data at various currents

Measured integrated in a 50cm sphere, the tint on my sample lies below BBL at all times. CCT ranges from 5880K to 6400K. Efficacy is still fine even at 10 amps.

CRI at the rated 4x1400mA=5600mA current.

CRI at various currents
200mA
1A
2A
10A

I didn’t measure it, but after trying the S2WP in the NI40, as is typical for a domeless emitter, the tint shift within the beam angle is very small. This will probably play nice with a wide range of optics and reflectors. Works fine as a bare emitter too. Other than being cool white and low CRI that is. I don’t really know who would want to use it.

edit: fixed a typo in the lumen calculations, which skewed the results (output is actually 6.4% lower than previously shown)

Added 8.12.2019 Data for a single emitter up to the maximum rated current of 5A

2

Would be great in a motion-activated light. Hella bright, and would make potential baddies run for their lives.

“Aaaauugh! Low-CRI!! RUN!!”

3

Now I wanna see Dale throw a 30T at one…

4

Thanks for the test data! Would you say it is better than a shaved dome xhp50.2?

5

Fixed a mistake on the lumen calculations (actually 6.4% less than originally posted) and added data when only a single emitter is driven. At 5A there’s still headroom for more output.

6

Would be interesting to compare it with SBT90.2… :slight_smile:

Thanks for the test maukka. :slight_smile:

7

Has anyone tied this LED in a large SMO reflector yet?

8

Thanks for the test. It is a powerful LED for sure, but with the SBT90.2 around it’s already obsolete IMO. There will be a donut hole with the S2WP and the beam intensity will suffer from the dark cross. Based on my testing of the smaller S2WN.

9

I’m on a quest for a brighter 4mm^2 LED. Getting 3+ A/mm^2 with the WFs has me spoiled. I wish osram would make a single 4mm^2 die rated for 12A.

The 4-die cross really kills the beam profile and throw, but I wonder if using just 2 of the dies side by side in the S2WP would work a bit better. Maybe even stuff some phosphor in the gap to try to fill in the dark area some.