Strange Cree MC-E failure with Fenix TK40 -- is this how an LED fails?

energythoughts

Newly Enlightened
Joined
Sep 16, 2011
Messages
66
So after a great 2+ years with my TK40.. I'll be sending it back to Fenix tomorrow. over the course of a few weeks I've noticed the light just wasn't as bright as I remembered it.. and this past week the color has really shifted to yellow. I've always used Sanyo Eneloop batteries, and I never really run it more than a few minutes at a time.. It's been warm.. but never hot to the touch.

I finally realized something was really bad when I looked at the LED and noticed it was turning black.. like it was burning.. As of today this is what it looks like:

JDR_120801_7345-2.jpg



I also posted a video so you can see the crazy Yellow tint it has now: http://youtu.be/RadqNYiIfaA

Is this a typical failure mode for LED's??

Luckily I called customer service at the US site fenixlighting.com and supposedly they are shipping me a TK41 U2 to replace this TK40. I was very happy to get to talk to a real person, and I'm shipping the TK40 to Arkansas and not China which is nice. I will be very happy with Fenix support if things go as smooth as planned!!

I'll miss the TK40 since it was a great light.. but I'm sure I'll like the TK41 even better!
 

Up All Night

Enlightened
Joined
May 29, 2012
Messages
487
So long to the mega-thousand hour lifetime! Must be an anomaly.
Glad to hear Fenix is taking care of you! I like my TK41, plenty of useful spill for a thrower. I don't think it will be quite the wall of light that the MCE provides. You're getting the U2 version! Good score, I didn't think it was shipping yet.
Good CS for Fenix!:twothumbs
 

ErConos

Newly Enlightened
Joined
Jul 13, 2012
Messages
2
Plis can test the new fenix tk41 u2? Any significant problem? thanks for video : Twothumbs
 

fiberguy

Newly Enlightened
Joined
Feb 23, 2012
Messages
188
It looks like you drew the short straw and ran into the VOC discoloration.

http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/XLamp_Chemical_Comp.pdf

It's covered from page 5 to 11.

From what I got here, the VOC discoloration is caused by a chemical or gas, no?

And "clearing the air" fixes the problem..

How did the OP draw a short straw and what exposed it to the gas? Could this be something that came sealed in at the factory? Could this be something all Fenix users should be somewhat concerned with in long term, high use lights? It took 450 hours in the lights that were linked. How many people actually have 450 hours on a light? Could this be the beginning of a potential issue? I'm just curious as to if I'm interpreting what I read correctly. I lean pretty heavily on Fenix lights for work..
 

Gunner12

Flashaholic
Joined
Dec 18, 2006
Messages
10,063
Location
Bay Area, CA
If the LED still lights up, it hasn't failed yet. The blue high power LED behind the Cree will keep running for quite a while.

From reading the Cree Datasheet, it seems like the VOC discoloring will clear up if you leave the LED in an open environment for a few days. The stuff that caused the discoloration might have been from the solder flux, any epoxy, thermal paste, or who knos what. Try opening the bezel ald letting the light sit for a few days. If that doesn't fix the issue, then look into modding the light.
 

HighlanderNorth

Flashlight Enthusiast
Joined
Sep 15, 2011
Messages
1,593
Location
Mid Atlantic USA
He said he is shipping the light back in return for a TK-41, so modding it wont be possible, but I have a 700L MC-E light and I havent had any problems whatsoever. I've owned it for 10 months, and it has a very different beam profile than all the other Cree LED's I own. I really like it. Its got a crazy wide spill, and yet its still got a fairly concentrated, medium sized hot spot, so it floods and throws pretty well. I can light up my entire living room from the balcony above one side of it, and its not a small living room! None of my other lights can do that, with the exception of the SC600, but the MC-E light has better throw.
 

energythoughts

Newly Enlightened
Joined
Sep 16, 2011
Messages
66
It looks like you drew the short straw and ran into the VOC discoloration.

http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/XLamp_Chemical_Comp.pdf

It's covered from page 5 to 11.

Great document! thanks for sharing it. This definitely matches my TK40 exactly. Very interesting that if the encapsulant is vented, the VOC's can outgas from the encapsulant and the discoloration will clear. I didn't realize the LED encapsulant is gas permeable either... I wonder if the source of the VOC's were really from inside of the TK40. I'm guessing they must be, but seems odd it would take 2 years before this got triggered. IPX-8 more specifically relates to water.. but what about gasses? Are all of these quality lights we have really hermetically sealed? If not I guess I shouldn't use Gorilla Glue (on the Cree incompatible chemistry list) too close to any of my lights... I wouldn't want to trigger VOC discoloration accidentally!

Well.. I guess if I didn't ship back my TK40 today I could play around and experiment.. maybe I'll save that project for later if another one of my lights fails. Maybe if one of you are bored out there... stick one of your lights in a sealed container with a bunch of Gorilla Glue.. and let me know the results :)
 

enomosiki

Flashlight Enthusiast
Joined
Mar 13, 2011
Messages
1,109
From what I got here, the VOC discoloration is caused by a chemical or gas, no?

And "clearing the air" fixes the problem..

How did the OP draw a short straw and what exposed it to the gas? Could this be something that came sealed in at the factory? Could this be something all Fenix users should be somewhat concerned with in long term, high use lights? It took 450 hours in the lights that were linked. How many people actually have 450 hours on a light? Could this be the beginning of a potential issue? I'm just curious as to if I'm interpreting what I read correctly. I lean pretty heavily on Fenix lights for work..

Glues, conformal coatings, o-rings, gaskets, and potting compounds are materials frequently used in the construction of luminaires, bulbs, or other solid-state lighting products. During the operation of the lighting fixtures, heating takes place, and some of these materials are known to emit volatile organic compounds (VOCs). In any type of sealed environment (such as beneath secondary optics or fixture covers), the VOCs will surround the LED and diffuse into the porous silicone encapsulant. There in the silicone, the VOCs will occupy a free space within the interwoven siloxanes chains. With subsequent exposure to heat and high-energy photons emitted from the LED, the volatile compounds can discolor and block the light emitted from the LED. This discoloration usually occurs just above the top surface of the LED chip since that is where the highest temperature and flux density is located.

Impurities in certain materials necessary for the manufacture of the emitter are the cause.

As noted, the discoloration of the encapsulant above the LED chip is due to a darkening of the VOCs that have diffused into the silicone from outside of the LED component. In most cases these volatiles, occupying a free space within the encapsulant, do not cause damage to the silicone itself.

In such situations, if the optic or cover above an LED that has discolored is removed, and the fixture continues to be operated, then the VOCs can outgas from the encapsulant and the discoloration in the silicone will clear. As with the initial discoloration of the material, depending on the nature of the volatile compound that has diffused into the silicone, this clearing can occur in a matter of hours or could take several weeks. The photographs shown in Figure 11 are examples of this type of outgassing.

There are, however, some VOCs that can damage the encapsulant, causing it to swell and crack, rendering the LED unacceptable for future use.

Basically, in order to reverse the effects of VOC, the emitter has to be de-domed and allowed to run, so that the heat can build up and vent any further VOCs that may be still present. De-doming, unfortunately, is not an easy process and there is a high probability that an inexperienced person will damage the emitter. Not to mention, once de-domed, it's impossible to re-dome the silicone without specialized equipment. Once de-domed, the emitter's tint will shift, while producing less output and decreased lifespan.

To the OP: just send the light back. It's not worth toying around with.
 

lightfooted

Flashlight Enthusiast
Joined
May 6, 2010
Messages
1,017
@Enomosiki: What I understand of the document from Cree is that the silicone dome covering the emitter chip is what is discoloring, not the chip itself. On pages 7 through 9 they are clearly talking about removing any kind of gas tight seal around the LED package not removing the silicone dome from the LED as you are suggesting. You can plainly see in the photos on page 9 that the LED getting "fixed" still retains it's dome.

Thank you for posting the link...it had not occurred to me that this was even a possibility before today, I certainly will be more careful of what types of chemicals get inside my lights from now on.
 

energythoughts

Newly Enlightened
Joined
Sep 16, 2011
Messages
66
Well.. Fenix did send me a brand new replacement light.. the TK41 (Not U2!).. They said they were sending the U2 so I might call them back and see if they will still swap it out. Even if they don't I'm very pleased with their customer service.. after owning my TK40 for over 2 years, and no questions asked they send me a newer model.. I can't complain really. One strange thing with my new TK41... There is NO serial number on it. Every Fenix Light I own has a serial number stamped on it. Is this something with all new Fenix lights? Manufacturing saving money? or is this some knockoff? hmm.
 

Up All Night

Enlightened
Joined
May 29, 2012
Messages
487
What about the sleeve the case came in? Mine has a silver upc sticker on the sleeve with a serial number below that matches the serial # on the head of the light. Perhaps in their haste to get the U2's out they've neglected to stamp/etch the heads. Maybe a U2 in old packaging? I'd email them about the lack of a serial # on the light and segue into the U2 issue. Just to be clear, mine is not a U2 version.
Good luck!
 

energythoughts

Newly Enlightened
Joined
Sep 16, 2011
Messages
66
energythoughts,
Any update with your quandary?


Well.. I'm a little late to respond.. but Yes there is a serial number on the sleeve... but when I look it up on the "serial number inquiry" tool on the website.. it returns "Item Not Found". I've been too busy to call them back and ask about it. I'm still very happy with the one they sent me.
 
Top