Thanks for the pics Nicolaas 
@KKW, the reason i am interested in maintaining the CW tint is because if you use a collar, the colder the beam you start out with the more kcd improvement you get, because blue light can excite the phosphors more & you get more recycled light. And i hope to start experimenting with collars in the future.
I’ve got a heatsink prepped to burn in emitters. Spaced for 16mm boards on the side in the picture, 20mm on the other. Once I get some short 4-40 screws I’ll be testing it out. 1”x4”x8” aluminum block.
How are you going to do the burn in? How much current, how much time?
Here are some old but interesting threads from the other side, if you want to get a feel as to how much current and time is needed…
KKW maybe you can try to put some soldering paste on the bond wires on some of them to compare, but i don’t know if normal paste works for gold.
That’s a chunk of aluminum KKW
It’s one of a few I have laying around. My biggest piece is a 12”x30”x1” piece of cast jig plate. Mmmmmmmm, flat.
I’m planning on using around 3A per LED, I’ll probably start there for XM-L2, XP-L, and XP-G2. I can run 9 at a time at 3 amps off my power supply with them wired in series. It sounds like 24 hours is the time to shoot for.
I have a recipe that involves graphite, that is if I can successfully expose the bond wires without damaging them
Silver, like the silver coated copper braiding I use? Maybe? I dunno???
They sure do, i have few Led's here i might try and tin one :)
The burn in block has 9 LED’s mounted. I just need to check their starting Vf and then wire them up in series.
Tested and wired.
Making light. A lot of it. You can’t tell but it’s in a brightly lit place already, blows the exposure value it’s so bright. Starting off it only pulled 2.55 amps at 30.15 volts, but in 15 minutes that has already gone to 3A at 29.82V. Temperature of the heatsink at the hottest spot on the broad side has stabilized at 106°F. The hottest emitter measurement I could get with my IR temp gun was 250°F on one of the XM-L2’s.
Are the tiny wires in LEDs meant to act as fuses?
So as not to evaporate the emitter or phosphor with too many electrons too fast?
That’s what I’m going to find out. The going theory is that it will lower the Vf. Useful for all of the single cell lights I have to build if true.
No. They’re gold wires sized to handle rated current with a safety factor that we exploit to push more power than they’re rated for. They act like a fuse when we get overzealous with the current. A non-replaceable fuse unfortunately.
not likely, those wires are way sufficient for the current range that these leds are rated for (hotrod lights use 2.5 times the maximum rated current of the xml2). Btw, the same wires are used from xpe2 to xml2. The xpe2 die dies long before the bond wires go, the xml2 bond wires fry long before the die goes, with the xpg2 it is a close call: the die already turns very blue but still the wires go first.
Keep in mind that there are 64 square inches of surface area to radiate heat just in the two broad sides, and less heat to dissipate than a 100 watt incandescent bulb.
Temperature after an hour is still stable at around 106° at the hottest spot on the broad side.
Edit: And down to 29.25 volts to maintain the 3 amps. I like where this is going. Remember it started at 30.15 volts and 2.55 amps. Even after initial temperature stabilization it was only 2.75A at 30.15 volts.
I started to wonder about the discrepancy between the temperature the of the block I was getting from my ir meter and what it felt like, so I stuck a post-it note to the side and checked the temperature on that (non-reflective) surface. Yeah, that’s more like it.
240°F. 
On goes the fan.
Freaking AWESOME! :bigsmile:
Not if the LED’s get hot enough to reflow themselves. :_(
Fortunately all is well, they’re all still working and the temperature is on the decline.
Interested to see the results here, but I can’t help but wonder what the difference is between running them on the test bench, as opposed to inside a torch?
That’s where your surface emissivity comes in