I have question for you folks: do these Osram leds work “better” (meaning: to generate more throw) on larger or narrower reflectors?

I saw this flashlight that has a long and relatively “narrow” reflector :

I was wondering if it might have a good effect or not (let’s say with an Osram 2mm2).

Thanks in advance for your input :+1:

Width, wideness is the biggest factor for good throw. The depth will keep the spill area narrower. I couldn't see any measurements on that light, but looks like about a C8 in width. I got a coupe lights with reflectors about the same dims, and no more, or much different than shallow reflectors. The old HD2010 was a good example of a wide reflector, shallow reflector but had good throw.

This led in a looong reflector will not likely do anything different than other leds: producing more or less the same throw as a shorter reflector of the same front surface area.

I’m a fan of those deep reflector lights though :slight_smile:

Took a nice bunch of lux measurements with my smartphone in front of the above white flats from Kaidomain (≈11 second averages with a luxmeter application, 0.1A current steps).

I finished quickly with the CSLNM1.TG (1mm²), I measured output at a few figures up to 5.1A (where I obtained its maximum) so I decided it was good to be driven at ≈4.7A with the Convoy ∅17mm SST-40 5A driver.

The 2mm² CSLPM1.TG was another story, noted down quite a bit of figures and while I could have gone a bit further (got a maximum at 7.4A, but given the previous and subsequent values I'd say that maximum was probably a tiny bit over the top already), I already had what I needed. No way 8 - 10 A, definitively. Between 6.0 and 6.7 A, I'd say.

Thanks TomE and djozz for the explanation about the reflector shape and its influence on the beam/throw! :+1:

Cheers for the info.

Did you notice any tint shift on the 2mm²?

Wellp, while I oriented the beam output in a direction more or less opposite to that of my eyes, I'd say there was no tint shift I could realize.

Uh! Oh! The back surface of the MCPCBs from Kaidomain was far from flat, being the 2mm² the worst offender:

Notorious ridges all around, and a tiny bulge you can see above which didn't go away with flux, solder braid and iron treatment.

I fixed the mess with 2000 grit sandpaper.

Barkuti, nice work! I bookmarked the post. We get spoiled by Hank's QC/QA on his Noctigons. KD just isn't there (yet I hope), but actually Hank has always been keeping achieving quality/reliability as best he can. I find stuff like this though in stock flashlights as well, Lumintop recently, but other brands as well. I always try to post the details of the hidden defects to make everyone aware.

Hank used to contact me occasionally if I'd be interested in a batch of reduced price rejected MCPCB's. I'd bout always buy them - the defects were such minor things they made no difference, but sure enough, they'd be marked with an X as quality rejected.

Barkuti. Thanks for the testing and info. Always appreciated!

Are planning on rerunning the 2mm test after fixing the MCPCB issues? Very curious how much peak current may have changed. It seems like the KD batch has a really low current Imax limit.

May you also remove the LED from the star and photograph the solder coverage as well? It seems there is something off… 5.1A/mm on the NM and 3.7 on the PM seems like too big of a gap IMO

It also doesn’t appear to have any clamping force which I think tests have shown is the most important factor to dissipate heat?

It doesn’t appear to, and yes, that would be an obvious bottleneck combined with the poor flatness. I guess I assumed it was epoxy since there weren’t any mechanical fasteners… 5.1A on the NM1 also suggested the setup was at least reasonable sufficient.

Both MCPCBs were already fixed with 2000 grit sandpaper, no need to re-run.

Now, yesterday I had a @#$% day. Emitters were thermal glued to the heatsink (with Tian Mu grey). Got really mad at me trying to forcefully remove the MCPCBs from the heatsink (alcohol, turpentine, pliers, …) slightly damaging the CSLNM1 board, and literally ruining the CSLPM1 with a blow. Ended up carefully sliding a razor blade underside, this worked pretty well. Lesson learned. :((

The performance of the CSLNM1 was definitively within margin (I didn't try anything higher than 5.1A). Concerning the CSLPM1, can't really say. djozz could only measure a maximum of 7.5A in his early test, beware so. Some more formal output tests are required, imho.

I'll order some CSLPM1 from led4power for my build. I will keep the driving current definitively under 7A anyway, there's really not much to gain near the limit.

Usually we screw down the MCPCB's for testing, but you might not have the tools or experience for that? I've done a bunch of tapping holes for screw mounts, both on a test alum block and in lights, but it's usually a pain, but this way you can use the better regular thermal greases.

Although I don’t do some powerful throwers as you folks, I like to try some things.
Last weekend I picked my Convoy S2 w/ Osram KW CSLPM1.TG 2mm2 (w/ led4power driver set on 6A,SMO reflector), my EagleEye X3R w/ OSRAM KW CSLNM1.TG 1mm2 (w/ led4power driver set on 4.5A, SMO reflector) and my Convoy S2 with XP-L HI U6-3A+ (w/ 8*7135 Biscotti driver, SMO reflector).

Went outside and compared some beams!
Convoy S2 vs Eagle Eye X3R vs Convoy S2+

Convoy S2 vs Convoy S2+

Convoy S2 vs Eagle Eye X3R

The beam of the KW CSLPM1.TG 2mm2 (W2) is definitely wider than the KW CSLNM1.TG 1mm2 (W1), so it doesn’t do the pencil beam. Still, it is very intense I am curious to see how it would act on a zoomable flashlight like the Odepro KL52.
Do you consider it would be a nice option?

The EagleEye with this led (W1) is still a nice pocket thrower, as I previously shown here: OSRAM CSLNM1.TG & CULNM1.TG 1mm², CSLPM1.TG & CULPM1.TG 2mm² - #956 by MascaratumB

Very nice MascaratumB! Lot of effort here :+1:

Yes it would be good in 50mm aspheric but only with precoolimator lenses. I think it is more suitable for smaller aspherics smaller than 40mm.

But for playing why not? It should throw around 400 kcd if you decide to mod it.

Just tested the Osram KR CSLNM1.23 Red and it maxes out at 3.5A which is disappointing.

2.8A - 233 lumens
3.15A - 240
3.5A - 250
3.8A - 240

taken at 5 sec

I will add that I’m not using a dtp board so maybe it could be pushed a little harder

Thanks for posting the results. Any idea what Vf was at those levels? Thanks

I wouldn’t know. I just used my lumen tube each time I added an AMC7135 chip with a fully charged 30Q

I don’t know how to measure vf