Three Oslons tested: 1) latest gen. Oslon Black Flat 2) SSL80 4000K 92CRI latest gen. 3) SSL80 4500K 96CRI 1 gen. before latest

So even though digikey has a 2000 minimum order I found it on mouser for $5 a piece.
http://ca.mouser.com/ProductDetail/OSRAM-Opto-Semiconductors/LUW-HWQP-8M7N-EBVF46FCBB46-8E8H/?qs=sGAEpiMZZMt82OzCyDsLFM%252b7yORtb%2FhA55GIye8lCf8%3D
It doesn’t look like you can order a specific bin.

MTN electronics hasn’t updated with any info about stocking these yet , so if anyone wants to reflow it on their own and test it out go ahead :slight_smile:

You never asked for a link, I could have told you that.

Osram almost never sells specific Bins. This explains the high variance between the LEDs.

It’s been a while, I have recently ordered 10 of the Black Flats for my next project.

Mountain Electronics doesn’t seem to want to stock them, so I will have to reflow them myself.
I’m going to use Noctigon XP-32 because of the thickness and area, it will help with thermal transfer from the MCPCB to the heatsink.
Depending on how easy they are to reflow, I might test two, or all 10 of the LEDs I get. I didn’t have enough money to buy 10 of the MCPCBs and reflow them all at once.

Not sure yet if I want to use a 240mm liquid cooler or a regular air CPU heatsink to do my testing.
The air cooler is what will be used in the final build, but at the same time I’m going to be doing my testing with regular thermal paste instead of liquid metal so maybe using the liquid cooler will balance that out a bit.

I probably won’t be measuring lumens unless I get access to the integration spheres at my university, but if they are easy to reflow I will be binning all 10 to find the best performing one for my build.

It would be nice if you post the results of your binning process. I would test each LED at maybe a low current and a high current (4-4.5A). Additionally you could also check what each LEDs maximum brightness is and at what current.

Yeah, it won’t be super useful without lumens but it could tell us where the peak current is with better cooling.
If I do all 10 we should be able to see the difference between flux bins.

Yeah, it won’t be super useful without lumens but it could tell us where the peak current is with better cooling.
If I do all 10 we should be able to see the difference between flux bins.

Do you have a spare (small) smooth reflector? You could measure the luminous intensity and we could calulate the luminance and approximate the lumens from that.

No not really, just lenses.
If I have time this coming year then I’ll ask if I can use an integrating sphere, will be much more accurate than anything else.

Looking forward for this test.
Djozz recently build another flat black in Supwildfire reflector and he got 350kcd while I with old xp-g2 s4 2b(previous throw king) in my best Supwildfire build got around 300kcd (when I calculate lux meter difference data that I was given from Djozz). But I probably did not had the best xpg2s42b here.

BTW where did you bought them? Can you share us a link? Maybe I would like to give it a try in a smaller aspheric and of course with precoolimator.

http://ca.mouser.com/ProductDetail/OSRAM-Opto-Semiconductors/LUW-HWQP-8M7N-EBVF46FCBB46-8E8H/?qs=sGAEpiMZZMt82OzCyDsLFM%252b7yORtb%2FhA55GIye8lCf8%3D
Pretty sure mouser.com has them too if you’re in the US.
HWQP is the one you want, and no place sells individual bins, they’re all a group from 8M to 7N so to get the most out of the LED you need to buy a lot and bin it yourself :frowning:

Thanks Mr. Enderman.

“They’re all a group from 8M to 7N so to get the most out of the LED you need to buy a lot and bin it yourself ”
Yes this is certainly disadvantage… Especially for guys like myself that are selling their mods.
I know Osram has great stuff like 4715AS emitter(with that emitter they listened to customers and made superior product to previous 4715S) , and if they will further develop this led and maybe even enlarge it will be a winner.

I read this part in the datasheet:

The above Type Numbers represent the order groups which include only a few brightness groups (see page 5). Only one group will
be shipped on each packing unit (there will be no mixing of two groups on each packing unit). E. g. LUW
HWQP-8M7N-ebvF46fcbB46-8E8H means that only one group 5N, 6N, 7N, 8M, N6 will be shippable for any packing unit.

If I interpreted this correctly, I think this means that for a single reel of 2000 LEDs, they will all be the same bin.
So buying 1 or 10 LEDs won’t matter because chances are they will all be the same.
Since the distributors don’t test the LED to find out which bin it is, even if you could afford it ($4k per reel) you would be buying 2000 LEDs of a mystery bin, and they will all perform the same.
:frowning:
Assuming that the distributors only order in quantities of 2000, it may be possible to wait until their stock is near a multiple of 2000, and then purchase a bunch of LEDs at once and have a chance of getting LEDs from two reels.
For example, mouser (canada) has 8036 in stock, so maybe if I bought 72 LEDs I would get 36 from one reel and 36 from another reel. This would effectively double the probability of getting a good bin.
.
On the other hand, I have seen stuff like 6N7N instead of 8M7N so I think certain places can sell a smaller range of bins. 8M to 7N is from worst to best, however 6N to 7N are the two best bins.
https://octopart.com/luw+hwqp-6n7n-ebxd46ebzb46-osram+opto-74739048
It says 0 stock, but maybe there are other places that sell only the best 6N or 7N bins, and then you have higher chances of getting a good LED bin :slight_smile:
.
As of right now though, pretty much everyone who has tested a flat black has found it to give higher intensity than an XP-G2 though, so I still have high hopes of great performance even if I get a bad bin.

Maybe someone should tell Hank from IO store(or any other led store owner here) to get them. They probably have good connections and can demand the best of manufacturers and if they’ll do that I don’t even doubt they will sell 2000 or more pcs in no time.
This emitter is good even if it draws only 3,6A with fet driver while giving same or higher numbers than old XP-G2 cause battery will last longer and light will run cooler in small and not so chunky aspheric lights.
So Djozz with one mod got 285kcd (Brinyte B158 fet driver modded, 3,6A draw) and unbeatable 350kcd record with Supwildfire, direct drive mod and 4,5A current draw. I got around 320kcd with good old XP-G2 S4 2B after adding Djozz lux meter measuring parameters for same budget lux meter we both have.
So this is emitter with great potential. Really wish that Cree starts producing super throwy XP-G2(or however will they call them) flat emitters. That would shorten mod work time for us.

The guys from flashlight stores simply buy their LEDs from places like mouser or digikey or some cheaper chinese places.
The manufacturers only sell in bulk to distributors, and I don’t think any store like mountain electronics or elsewhere is going to buy 2000 LEDs of one bin.
I asked (richard I think?) from mountain electronics if he could get some a while ago, but he never did :frowning: so I bought them myself and will attempt to reflow it.

Anyway, I’ve started my project, I just need to reflow the LEDs now, find out what the max current they can take is, and order an appropriate driver for it.

Did you used slightly press from above emitter to push all excess solder out and for better heat transfer?

Edit:
Upss… You did not started yet with reflowing… Be careful. Eagle eye and steady hand will help and I would really go with Old Lumens method on those tiny ones (benchvise+soldering iron).

Thanks for the tip.
I was thinking of using a heat gun instead of a soldering iron so I don’t get bits of solder stuck underneath the MCPCB.

But you will not have any bits of solder beneath if you use very thick and clean flat soldering iron tip.
And even if you’ll have some use fine steel wool 0000 grade or fine sandpaper to remove that, and that will even help for better sticking of isolator thermal compound like artic alumina. When I do stuff with AA (mostly when I am using osram IR leds) I fine sand noctigon, if I use artic silver thermal paste I leave noctigon as it is (gold and clean).

Heat gun? Good luck with that cause you’ll need it.

Luck?
I thought most people used hot air when reflowing SMD components?

Just do it the way you like and if you fail try mentioned OL method :+1:

I am looking forward for tests.

BTW Heat gun and hot air station are not the same to me.

Yeah I meant the hot air thingy that’s smaller than the gun.
I will try one the OL method first, if that doesn’t work I will try with hot air + solder paste.