XR-E failure mode

[saabluster]

I thought the community might benefit from some of what I have been finding in examining old XR-Es that have some service life on them. It probably is no shock to many of you that I love the XR-E package as it is what I have used for years to make the longest throwing flashlights in the world. For my application they are simply optically superior as well as better thermally than the newer XP-E package design. Many people over the years have noticed that the hard driven XR-Es however can burn up and reduce the output significantly making some appear to be amber LEDs.

Here is an example of a really bad one.

Conventional wisdom has always said that the phosphor was burnt. In dissecting these I have found that it is not the phosphor but the silicone gel that is burning. Look at that picture again and you can see that the burnt part extends out away from the chip.

Removing the dome this is what we see.

And now with the burnt gel removed.

And yes these pictures were made using my iPhone along with an assisting lens.

This is looking back through the dome with the gel still attached.

I have done extreme tests on these newer style LEDs that have a solid silicone construction(XP-E/XP-C) and they seem to last much longer as the silicone can take higher heat.

 

[ahorton]

Interesting. Do you know why this particular one burnt out? How hard were you pushing it?

I have a similar one which was never overdriven and always had a good thermal path (soldered directly to brass). I put it down to the fact that I'd reflowed (on the stove) it almost 10 times onto different MCPCBs, brass and copper strips.


If it's the silicone that's burning, does that change anything in the way we should treat them?

 

[saabluster]

Interesting. Do you know why this particular one burnt out? How hard were you pushing it?

I have a similar one which was never overdriven and always had a good thermal path (soldered directly to brass). I put it down to the fact that I'd reflowed (on the stove) it almost 10 times onto different MCPCBs, brass and copper strips.


If it's the silicone that's burning, does that change anything in the way we should treat them?

I must admit I'm not entirely certain why this one cooked. I have received a few lights before for this but they were all the economy model FTP which used an mcpcb. This one was to solid copper. My running theory on this particular example is that it was a low forward voltage LED combined with the older driver setup that could have allowed a higher current to the LED than normal. Not much you can do I'm afraid other than keep the junction temps down.

 

[JohnR66]

Interesting. I'm not sure if that is really damage from heat or from the intense light radiation that causes optical defects in the silicone. Since silicone polymer is not based on carbon, I'd expect different results from it burning. Yellowing, or in severe cases, browning in clear materials (even glass) can be caused by intense or long term irradiation by electromagnetic radiation of certain energies.

 

[bshanahan14rulz]

Thought I'd post this link here. It's almost relevant, and I can't find the other thread that I was looking for regarding something similar..
http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/XLamp_Chemical_Comp.pdf

Apparently, the "burnt phosphor" effect isn't even burnt silicone either, but rather, the stuff that gets absorbed by the silicone is burning.

 

[saabluster]

Thought I'd post this link here. It's almost relevant, and I can't find the other thread that I was looking for regarding something similar..
http://www.cree.com/~/media/Files/Cree/LED Components and Modules/XLamp/XLamp Application Notes/XLamp_Chemical_Comp.pdf

Apparently, the "burnt phosphor" effect isn't even burnt silicone either, but rather, the stuff that gets absorbed by the silicone is burning.

Almost relevant?!! Man this is completely relevant to this thread. Thanks for posting that. Now I am going back in my mind trying to remember all the potential culprits.

 

[yellow]

for the non-technicians:
what does this found mean?

* that, by removing the gel (and then putting the ring+dome on again) a higher current can be used safely?
or
* by "cooling" the Ring/dome (say, with an additional metal "ring" to put away some heat) the gel can be safed ... (when running higher power through the led)

:thinking:

how about putting the Ring/dome over an XP-G / XM-L?
Would this lead to the optical pros of the XR-E, with the improved brightness of the newer emitter plates?

 

[wquiles]

Almost relevant?!! Man this is completely relevant to this thread. Thanks for posting that.

+1

What a great article :D

 

[Canuke]

Thought I'd post this link here. It's almost relevant, and I can't find the other thread that I was looking for regarding something similar..
http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/XLamp_Chemical_Comp.pdf

Apparently, the "burnt phosphor" effect isn't even burnt silicone either, but rather, the stuff that gets absorbed by the silicone is burning.

Not just that; if you get this issue, according to the report, the LED's can be cleared by simply running them in open air so the VOC's can leave just the way they came. Worth knowing, as I can easily see anyone assuming that the emitters were a goner once they went dark.

 

[wquiles]

for the non-technicians:
what does this found mean?

* that, by removing the gel (and then putting the ring+dome on again) a higher current can be used safely?
or
* by "cooling" the Ring/dome (say, with an additional metal "ring" to put away some heat) the gel can be safed ... (when running higher power through the led)

:thinking:

how about putting the Ring/dome over an XP-G / XM-L?
Would this lead to the optical pros of the XR-E, with the improved brightness of the newer emitter plates?

What this means to me:
- try to stick with "approved" chemicals/glues/etc. when doing LED projects (Cree has the list in that article)
- if you can, test it before you go to production - to make sure things will be OK longer term
- it would be good to operate the LED/system in the open (instead of an enclosed setup) for a while to cook/bake off the components so that the VOC's can go outside of the LED setup, instead of being absorbed by the LED's porous dome. Once compounds are "cured" then you can go back and enclose the LED as necessary. Of course this implies waiting, which is always expensive in any production environment!

Will