New Osram Oslon Square, testing a 4000K 92CRI typ. emitter

I don’t need 5 more medium sized LEDs just now, but this looks like real progress. Hooray!

Thanks for the test, djozz, excellent work. These new Oslon Square look very promising. I hope we’ll see a dtp-copper-MCPCB for them, especially a triple-version.

+1 :bigsmile:

As expected dedoming fails: much of the phosfor is mixed with the silicone. I tried the hot method but usually then the gas method does not work either. You could try slicing the dome to close to the die, but I'm not sure if you improve the illuminance with that...

(stereo image)

I have a headache trying to find one of the new square LED’s on digikey in a cool white. If somebody can find it and post a link I’d appreciate it. Djozz, I couldn’t even find the one that you listed. I found a lot that had similar part numbers but not the same. Like you said, the Osram part numbers make no sense!

I think dedoming will always increase the luminance, unless there is an air space between the die and the dome, as with the XP-L. The total internal reflection at the air where the dome used to be reflects light back into the phosphor, making it brighter. The dome in your picture looks different shaped, but that outer surface is just passive optics.

Pretty cool LED. It looks like as long as you're sticking to 3A or less the gain isn't huge (except for the lower vF), but you can definitely push it harder! Thanks for sharing.

What I was referring to (did not myself very clear, I admit) is slicing the dome off, but leave a thin layer over the die. Any illuminance increase relies on internal reflection against the cutted surface, I'm not sure how much gain that is going to give.

According to James who sells dedomed and sliced emitters here at BLF, a sliced emitter should be in between a domed and a dedomed emitter, hotspot more intense than a domed one but less intense than a dedomed, but with no tint shift.

I must say that Digikey does not seem to have the nicest emitters in stock, in a glance Mouser's selection seems better. I think for the new gen. Oslon Square you have to look at the datasheet:

-If it is from 2013 it is the old one, is it from 2014, it is the new one, I saw datasheets from feb2014, but the led of the OP has a datasheet from sep2014.

-In the introduction to the new one it mentions 'this is the second generation Oslon square', and when you look at the Vf, the old gen says min2.80typ3.05max3.25, the new one says min2.70typ2.90max3.20.

How are things going with your X6 triple square? Still planning to do this? I’d love to see beamshots of the square vs a 219b.

hmmm. I really wanted to try the GW CSSRM1.PC-MUNQ-5F7G-1. It is a 5700k LED and is rated at 259lm but nobody has it in stock.

I did find GW CSSRM1.PC-MTNP-5H7I-1 which is a 5000k at 240lm (700ma). data sheet is here

comparing it to the XP-G2, its close to the R5 but I don’t think it will beat the 5000k S2, but I don’t have all the fancy testing equipment to say for sure. The S2 is 271lm (700ma) vs the 240lm of the Square. How that looks once the current is turned up and the FV increases….who knows. The fact that it can’t be dedomed and I can’t find the 5700k version may kill me buying a Square. :frowning: I may still toy with doing a triple in something, but I don’t see using it as a single emitter right now. Not to mention it is 75% of the price of XP-G2 LED’s

I'm sorry, the idea was cancelled for now, I used one up for the dedoming experiment and now I only have two leftover. But when I do a rsonline order again (can't do that too often because I can only order with them via my work) I will include a few of these again and make the triple.

I spent a bit of time looking at the Olson spec sheets and did learn a few things

CSSRM1.PC-MUNQ-5F7G-1.

MUNQ is the luminance binning, OSRAM uses letter coding the first two letters are the min Lum. bin, the second two letters are the max lum. bin. MU (min) 259, NQ (max) 330

Same way for the tint bin

5F7G means emitter tint is 5F to 7G (look at the chromacity chart, similar to Cree chart but different tint codes) indicates 5700K - 5500K which is a pretty tight tint grouping.

and the M1 designation is the Vf

Hi Djozz, thanks for finding these leds. I have a few on order from RS, but was wondering how easy they were to reflow onto xp mcpcb’s as the dimensions dont quite match up?

Cheers Gurthang for helping to decode the part numbers, its easy when you know how! :slight_smile:

Woody

Reflowing on xp-boards works, but you have guide the led more and use not too much solder paste because it is way more difficult to squeeze excess solder out from under the led (see also post #10)

thanks to Gurthang as well for deciphering the osram code!

It seems that a bit of practice is required….good job I ordered more than I need (at the moment!)

It’s not that tight, you must have gotten something wrong when reading the chart. 5F to 7G is more like 5700K - 5300K.

I picked up a few of these and the tint appears to be a little bit green. I think I prefer the Nichias and 2700k xm-l2’s, but I will give this one a little more time.

I got the following Mouser part:

720-GWCSSRM1CCMQMS5L
MFG Part No:GW CSSRM1.CC-MQMS-5L7N-1-700-R18
GW CSSRM1.CC-MQMS-5L7N-1-700-R18
Osram High Power LEDs - White
US HTS:8541402000 ECCN:EAR99 COO:MY

Sinkpad has triple stars for these, as well as 20mm stars… No 16mm, and they are DTP, but aluminum….

I have put a couple on XP Noctigons, but the triples I cannot get all three aligned enough to do correctly, one is always off…

If there is enough interest, we could do a bulk buy of the sinkpads as they have a MOQ of 40-50 pieces…