Thrunite TN30 developed a yellow XML2 LED

Wow that’s crazy. I’ve never seen anything like it before.

I have seen XML2 turn yellow from too much heat, too. The old XML didnt have that problem. Can be really annoying if you notice a tint shift after reflowing..

Mine has 1 led that has the same thing too. :_(

In these pictures it looks like the "borders" of the yellow areas on the die are the same as the vertical pwm lines.

Do you really see that a sharp bordered diagonal half of the LED has turned yellow, or is that just the pwm messing with the camera?

So it seems like I’m not the only one. That’s annoying.

My point of annoyance is that I actually sent it for replacement and it came back worse! (I had one before but now 2)

Well time for me to open it up

Those are good pictures for a camera phone! It looks like the silicone of the dome detached partly from the die, which results in the same effect as dedoming: tint shift and some output loss. Interesting that it happens easier on the xml2, this will also be the reason that the dedoming is easier with those newer leds.

Anyway, that is very annoying if it happens in such an expensive flashlight, I hope they come up with a proper solution for you.

They told me to ship it back to them!
$$$$ disappearing fast!

What do you guys think?

I know this may sound like a kind of dumb suggestion, but maybe do a full dedome on the emitters? At least that way the color would be uniform.

Is this going to be a problem with ALL XM-L2 emitters/lights? Or is this only a problem with the TN30?

I dont think I wll de-dedome.

I like my lumens too much :slight_smile: and not too bothered about throw.

Perhaps I will send it back :S

I hope that they can fix it, and esp., I hope that this is not a latent defect in the XM-L2 emitters, because I have several lights incoming that have XM-L2 emitters in them, and I’ll be ticked off if they all start doing that (dome de-laminating partially) :(!!

If it’s only happening in TN30s then it probably has something to do with the way Thrunite reflows the XML2s on. They might be using temperatures way too high. Perhaps a worker slipped and pushed a knob?

I don’t think it’s a problem with the XML2, but rather the manufacturing process of the TN30. Otherwise we’d be seeing other XML2s having this effect.

I’m not at all familiar with the TN30, but how long have the ones with XM-L2 been out/available?

They came out very quickly after the XML2 was released (December 2012), probably around March or April I think.

I was just checking, and one of the Amazon reviews was from Nov 2012, and the listing was under XM-L U2, vs. XM-L2, so it seems like it’s not just with the XM-L2 emitter versions, so maybe (hopefully) you’re right, that it is just a TN30 problem, rather than a generic XM-L2 problem, since the XM-L U2s have been out for awhile already?

If it happened with the original model of the TN30 with the XML, then it’s definitely not Cree at fault here but Thrunite. XML U2 came out when XML came out in 11 Dec 2012 Source

I took apart my TN31 (xml2 version) this weekend and was surprized to find just a small dap of thermal paste under the emitter base. Less then 30% of the base had thermal grease contacting it. The bigger problem though was that there was a big bump in the center of the pill. This lifted the led base off the pill so that only 3 small spots were touching the pill. The center bump and the points were the 2 screws hold the base down. Furthermore, the led itself was not fully seated. It was tilted up and has a partial gap between the base’s thermal pad and the led. I bet if I would have got the yellow led too if I kept using the light like this. Maybe, there is a similar QC issue with the TN30.

Ah good point, it might not be manufacturing but rather QC issues. I wonder if this is how they brought their price down by so much.

If you take a look at THESE pictures you can see a couple of serious QC issues in a TN31 with direct effects on thermal characteristics. The LED was reflowed on with not enough solder on the middle heat dispersion pad and kapton tape was even left attached to the base of the PCB.

It really looks like a overheated emitter to me, I was thinking that from the start then I read this whole thread where you say there is only a tiny, inadiquate amount of thermal compound and it makes me think that even more. From someone who has burned up multiple multiple emitters that’s exactly what it looks like to me.

Do you have a way to measure lumens/lux/cd? I would be welling to bet the yellow one(s ) are lower output than the others.