OSRAM CSLNM1.TG & CULNM1.TG 1mm², CSLPM1.TG & CULPM1.TG 2mm²

I measured 950-980lm for white flat if that helps, but neither of us have proper integrating sphere.

Boost HX is just CSPM1.TG in 4040 package, so you can expect similar difference - 5~10%, I don't know why people are hyping boost HX so much, like it's gonna perform 50% better than CSLPM1, it's the same die in bigger package.

BTW, I have limited qty for sale (100 something pcs), I'll put listings tonight.

This LED is compatible with 16mm 4040 Cu DTP board, 20mm Cu mosDTP board and BLF Q8 4040/5050 Cu DTP board.

According to the datasheet it offers the same lm/A with lower Vf - so it’s more efficient. And thermal resistance drop is huge, to the point that K/W/mm² is better than that of White Flat.

Considering that thermals used to be the most important performance predictor for overdriven LEDs - yes, Boost HX looked revolutionary.
Your Boost HL results do curb my enthusiasm. Though I still expect it to be the best LED for large throwers.

Hmm, exciting times for my Z1. Maybe I'll throw this in it. I'd been planning on reflowing a white flat from 16mm to 20mm board, since I damaged the one on my 20mm board.

Some efficiency chart:

Sample 1 of CULNM1.TG is clearly less efficient than White Flat but it holds up longer to achieve ultimate win.
Sample 2 is slightly more efficient and keeps its own early on while pulling farther ahead at high currents.

Nothing surprising.

BTW, sample 1 didn’t reach 1000 lm/mm². The die is 1.069 mm² so it’s a tad under 960 lm/mm². Sample 2 exceeded 1048 lm/mm².

Yes, HX is more efficient than CULNM1.TG - reason is 2x larger die, so you need 2x lumens for the same throw. I agree HX would be better for bigger lights, but for small and medium lights CULNM1.TG/CSLNM1.TG is better for maximum throw at still relatively low current of ~5-6Amp.

No, Boost HX was supposed to be more efficient than CSLPM1.TG. See the datasheets. Both have about the same output at any given current, from 1A all the way to 6A, both with the lowest and the highest bins.
But Boost HX has ~2.89V at 1A and White 2 has ~2.94V. Small but grows with current. At 4A it’s ~3.2 vs. ~3.29. At 6A ~3.35 vs ~3.52. That’s where the datasheet ends, about 5% difference. If the trend continues this should be about 7% efficiency boost at high current. And considering that peak output for all new Osrams is just below 50 lm/W - this lower Vf should push the peak intensity up by about 7%.

Actually even if the Vf difference is not there - it should still gain more from the bigger package than the 1 mm² die. Because it was more thermally limited at 3030 format in the first place.

I expected Boost HX to seriously undermine the XHP35 thrower market. Now….I still think it will make a splash. Just not that big one.

Datasheet Vf curves are usually given for some typ. Vf bin, what I see in real life is CSLNM1 and CSLPM1 come in lowest "15" Vf bin, while CULNM1 (my samples) are middle "B5" Vf bin.

This means you can't really make predictions using datasheet numbers, Vf values of "15" binned CSLPM1 are considerably lower that what datasheet suggests.

Another picture, rescaled each curve so they start at the same point.
This way we can see clearer how does performance scale with current, abstracting away efficiency of individual LEDs. Here some samples are clearly better than the others, such test hides this difference.

The curves for Boost HL nearly merge into one, that’s good…much better than I’d expect…this hints at very good test setup.

I would expect Boost HL and White Flat to score pretty much the same at low currents but pull ahead later on.
That’s not the case, they start the same, White Flat is actually better at middle currents and Boost HL pulls since 4A (which is about 6.5 at the X axis).

I would think this win in the middle to be a test fluke but the consistency of Boost HLs makes me wonder it’s actually real.
I don’t know and I don’t have an explanation.

Thanks for these graphs Agro.

Which numbers did you use for cslnm1.tg? You have mentioned Enderman got 1115 lm, I got considerably lower max. (950~980lm) for the same 6N bin, djozz' numbers seem to be also <1000lm from last test.

Also it's worth mentioning that 6N bin is second highest datasheet bin fom cslnm1.tg, while 5Q bin is second lowest datasheet bin for culnm1.tg.

We don't know what bins are practically possible, but we can realistically assume culnm1.tg could come in one bin above 5Q.

I used to use your lower numbers because you highlighed them.
But since you mentioned getting 980 lm a couple of days ago I went with that in these 2 charts (though in the latter it doesn’t matter as it’s linearly scaled anyway).

Have they been put up for sale yet? Or already sold out?

Added to site, sorry for delay, busy as usual:

CULNM1.TG

I’m surprised but quite large price hike.
It is the throw king and for quite a few uses - well worth the price. But I’m still wondering.
Is it that Osram charges much more? Or is poor availability the reason? Can we expect the same hike from Boost HX?

Price of reels is significantly higher than white flats, and reel is 2000pcs, I guess higher price is mainly caused by the fact that LED is not mass produced.

I see, thank you. So…you’re not short of them yet :slight_smile:

This is going straight into my L2. It should centre a lot easier than it’s 3030 brothers. With the Luxeon V’s, I reamed a 3535 gasket with a 5.5mm drill bit and it fitted perfectly. It should work with this one as well….

I didn't buy full reel, far form it, I had ~190pcs before sales started, and I don't know when/if I will have new ones.

I secured 3 pieces, but am not sure yet where to apply them.

Thanks l4p for making these available :slight_smile: