I think the "what did you break today" thread is super important for newer modders to see, especially if people like CRX, djozz, MRsDNF post there. When they bust something, it helps them feel better that these well-known modders break stuff too.
I managed to get a picture of my bare XHP50.2 while it was on. This involved shining it through a slide projector lens onto a big sheet of paper, and taking a picture of the projected image.
When I take a picture of it directly, with the LED off, it looks like this. There’s also another flashlight shining into the reflector to illuminate things better.
If I use only overhead light, it’s less colorful… just some green shining in from a nearby Kermit-green wall.
I actually was looking for that thread and couldn’t find exact keyword to search the link.
Will bookmark the link instead for future reference :person_facepalming:
I opened my D1, to try a slice & dice on my 3V XHP50.2…
But as soon as I desoldered the LED wires, I saw that the dome was slightly lifted
I think I damaged it when I removed the “butterfly” centering gasket that I used.
Now I have a UV/blue thrower that heats stupidly fast…
I measured 19A at start !
On your first picture it really looks like the die spacing is very uneven.
On the second and third picture that does not seem to be the case. Though the bottom dies don’t seem to line up…and the top ones don’t either.
In my fairly elaborate Ham’r build the center SST40 has already had to be replaced once, the dedomed die surface got something on it and made a pile of ash, after which the light it produced was on the green side. I had put a domed SST40 in it’s place at a loss of throw but small lumens gain. Now, in testing my rebuilt light box, I found out Ham’r is reading significantly lower than expected, “only 20,000 lumens” in the now calibrated box. As it turns out, the SST40 in the middle isn’t working once again.
As this emitter is run directly off the original Q8 driver as the Master of 4 Slaves, it has to be a wire came loose to the SST40. Bond wires appear intact, no sign of burned phosphor, just nothing from the center emitter.
While this may sound comparatively easy to repair, it will not be… I sealed connections in epoxy then embedded them under Sugru, a silicone compound. The easiest repair would be if one of the leads on the MCPCB desoldered due to heat (those contacts are sealed but would be the least obtrusive to get at) If it’s at the driver it will require hours of work to get at to fix and reassemble.
It’s less uneven than it looks in the first pic. There’s still a bit of phosphor between the top two dies, and between the left two… so it lights up and looks white in the picture, which makes it look like there’s more space between the dies.
It’s definitely better than other SST-20 lights I’ve tried. It could probably stand to have more of the 3000K + 6500K emitters and fewer 4000K + 5000K though. Farther apart should help pull the Duv down more.
But then: 5000K and 6500K are low CRI. Maybe it would be better with just 3000K and 4000K mix as a base with few 6500K emitters to make it cooler and rosier?
No. At ~4200K, I find it’s already about as warm as I can deal with. If I made it warmer, it would look even less white. Even at 100 CRI, it would still look worse. Warm incandescent-style lights make my eyes hurt, literally.
I’ll enjoy my medium-CRI neutral white instead, thanks.
I did not mean to lower the total CCT. Rather - use 6500K exclusively as the cooling agent rather than the mix of 5000K and 6500K. It should arrive at about the same CCT with fewer low-CRI LEDs.
I view it like: mix 4000K and 6500K only, with a good majority of 4000K. If that’s not rosy enough, replace some 4000K with a mix of 3000K and 6500K.