Seoul P4 Die degrading

[roguesw]

Hi guys, I just noticed this in my lights. I had bought 2 P4 from Fred earlier this year and installed them into my lights. One is an upgrade, Pr-T LuxIII to P4 driven by DB 750.
The other is a P4 on ArcMania Tower module driven by SOB 1000.
I noticed the P4 Die on the Pr-T is beginning to degrade.

This is the P4 in tower module. Notice nice clean P4 die area.



This is the P4 in Pr-T. Notice the P4 Die area is smaller than the previous and there appear
to be oxidation inside the die area.



The P4 in the Pr-T is being driven less than the P4 in the Tower module, 750 mA vs 1000 mA and in fact, i use the tower module more than i use the Pr-T.
Any ideas as to what is happening?

 

[LukeA]

The lower current doesn't mean that it's cooler. It might not be heatsinking as well, and is getting hotter. If it's not that, :shrug:

 

[tebore]

How's the heat sinking on the Pr-T?

You should look at the analysis that member Newbie did on the P4, before he "retired".

The P4 uses a less stable phosphor than Cree, you can say Unstable. The LED is rather cheaply constructed compared to the Cree or old Luxeons.

The other thing I can think of is the LED is sucking in moisture from the air and it's degrading the LED. You can try baking it for a while. I had one of my P4s get a bit foggy and the shiny silver area get a tinge of brown.

I love the P4 LED because it's an easy fit for old Luxeons but once you know how they are made you question their reliability in the long run with abuse.

 

[roguesw]

Thanks for the comments, i am thinking along the lines of moisture too. As i said i have not used the light much, only a handfull of times since i built it, its been in my bob bag.
Its summer here and for the last month and a half, especially where my bob bag was stored, its been humid, humidity in the high 80% for the whole day.

Also maybe heat can also be a contributing factor too. When i assembled the light, i put a thin layer of AA epoxy on the base of the slug to isolate it. Then i used AA compound instead of epoxy to heatsink the LED. I did this so that i could easily upgrade the LED in the future and not destroy it removing it.

I didnt realise the SSC was cheaply constructed, maybe its time to upgrade to the U2 bin that Fred is selling.

Thanks again for the answer.

 

[tebore]

Thanks for the comments, i am thinking along the lines of moisture too. As i said i have not used the light much, only a handfull of times since i built it, its been in my bob bag.
Its summer here and for the last month and a half, especially where my bob bag was stored, its been humid, humidity in the high 80% for the whole day.

Also maybe heat can also be a contributing factor too. When i assembled the light, i put a thin layer of AA epoxy on the base of the slug to isolate it. Then i used AA compound instead of epoxy to heatsink the LED. I did this so that i could easily upgrade the LED in the future and not destroy it removing it.

I didnt realise the SSC was cheaply constructed, maybe its time to upgrade to the U2 bin that Fred is selling.

Thanks again for the answer.

I don't know of the U2 or if the new P4s use the same methods of construction. I really suggest looking at the data Newbie put together.

The die of the P4 is epoxied while Cree and the old Luxs used a metallurgical bonding method. And the Phosphor was tested and showed that it was extremely temperature sensitive.

To avoid sealing in water what I do with P4 equipped lights is I run them on max in an unsealed light to burn away the moisture in the LED. Another approach is to bake the LEDs according to the spec sheets in an oven. The oven is for LED's I've stored for a long time.

Depending on the thickness of the AA epoxy that method sounds like it would of acted like an insulator. Based on my experience using it on computers you don't want to have it set before attaching it to another surface. The heat transfer becomes really really bad. it bonds and transfers well when the surfaces are mated wet but when it's set it doesn't transfer heat so well.

 

[roguesw]

I just did a quick check to see if my method of construction with the AA is conducting the heat to the Pr-T head. I turned on the light and within 1 minute the head was warming up and within 3 minutes the head was warm to the touch. The temperature is similar to the L4 head I have. I dont have any tools to measure the temperature but
am i safe to assume that if the head can heat up sufficiently fast that its conducting the heat away from the LED and heatsinking is ok?

Thats an interesting idea about running the LED without the head sealed as a burn in time to remove moisture. I heard of this in the Incan forums for high temp incans but never considered it for LEDs.
I did read the datasheet about baking the LED when stored for a long time. The next time i do a build, i will bake the LEDs before bonding it.

Thanks again.

 

[tebore]

Sounds like the method is working.

It's possible you just got a bad LED.

There's another number of reasons for burn-ins. Usually if the LED is a bad it'll fail early. Also it'll let you see if the LED is properly bonded as it'll discolor as it overheats. I also do burn ins to see how the vf's shift. Usually drop a bit within 100 hours then rise back up and settle a bit lower than initial.

 

[darkzero]

I don't think it's because of lack of heatsinking. In my experience lack of heatsinking causes the dome to turn yellow or brown.

 

[TexLite]

Ther is a post somewhere about this,its not a defect and shouldnt hurt anything IIRC.

The thread was in the LED section I think.

Michael

 

[Black Rose]

I didnt realise the SSC was cheaply constructed, maybe its time to upgrade to the U2 bin that Fred is selling.

I have some new U bin SSC P4s I got from DX and a couple of U2 bin SSC P4s I got from Fred.

Visually, they look the same (construction wise, not illumination wise) to me.

I've put a couple in some lanterns I've modified, and those gummy domes are a pain in the butt.

 

[tebore]

http://www.candlepowerforums.com/vb/showthread.php?t=146607

Here enjoy boyos. It's kinda nice to know how the SSC is put together.

There's a similar thread for the Cree P3, once you compare both threads you'll see what I mean.