You donât need special equipment. A stove top and a pie tin will do. Pre tin the star using a soldering iron and heat in the pie tin until the solder melts and the led settles into place.
Well, I think someone whoâs done if for years will be able to do it better than meâŚ
I donât want some ugly burnt piece of copper, I like how clean the ones I buy are 
Yeah I toasted a few but the learning curve is pretty steep and if you practice on the ones you want to replace it doesnât come at any great price. Hint: it doesnât take enough heat to burn the silk mask. Waiting sucks, I have a bunch of LEDs and mcpcbâs and the hardest part is remembering which direction to install them.
@ Enderman & The_Driver: in Germany the flat was tested at 250cd mm^2, but that was including optics. Without it should be even better.
@ Djozz: no worrie, no hurrie mate. Just make sure you get enough of them to satisfy my modding needs 
It looks as if this Oslon will be able to take the pencil beam throwers a good bit further the the XP G2âs. :+1:
and without tint shift!! :+1: :+1: :+1:
Grtz
Nico
I wouldnât count on them reaching much higher values. Photographic optics like the one he used in the test have very little transmission losses (should be in the single digit percent range). The problem with his measurement of the LED without optic is that he didnât account for any of the stray light from the LED. It bounces off of the surroundings and also hits the sensor of the light meter which makes the measured value higher than it really is.
Using an optic reduces the amount of stray light and the the much higher measured lux values greatly reduce the effect of any remaining stray light.
I would use 250-255cd/mm^2 as the baseline for any calculations.
I emailed MTN electronics, he said he will look into it, try to get a few for himself first and then see if he will sell them.
Hopefully he will
Very good news, thanks!
Great! Thank you, Enderman.
Youâre welcome
Hey, interesting emitter!
Just requested a quote from RS (UK) for 5 pieces of Black Flat v1.2 5500K, but i read Djozz is waiting for v1.3 high CRI?
Either way, subbed to this threadâŚ
What does âv1.2â mean?
Ehm, not sure what you mean, Jerommel. AFAIK the current Black Flat is the latest one, it only comes in one tint and there is not a new version announced.
I am waiting for the high CRI 3rd generation Oslon Square (is out now but no one sells it in small quantities yet), if that is what you refer to?
Ah, yes, that must have been itâŚ
I sometimes mix up the types, names and numbers in my head with all these emittersâŚ
LUW HWQP is the 2nd version of the Black Flat, âversion 1.2â as they call it in the data sheet.
LUW H9QP is version 1.1
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 doesnât look like you can order a specific bin.
MTN electronics hasnât updated with any info about stocking these yet , so if anyone wants to reflow it on their own and test it out go ahead
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.
Itâs been a while, I have recently ordered 10 of the Black Flats for my next project.
Mountain Electronics doesnât seem to want to stock them, so I will have to reflow them myself.
Iâm 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 didnât 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 Iâm 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 wonât 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 wonât 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 wonât 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.