NOT AGAIN!!!!

[PeterScowcroft]

Arrgh,

A few months back you would have read the thread regarding one of my LEDs yellowing.

I couldn't find out what caused it, assumed it was a dodgy chip and replaced it.

As the title suggests it has happened again!!!

What on earth could cause this?

The setup is 2x2 parralel/series run at a maximum of 500mA with a massively over engineered heatsink.

It was the same LED in the same position that went.

What can cause yellowing of LEDs?

Anyone has any good suggestions?:twothumbs

 

[RustyKnee]

what type of led?

i knocked the glass dome off a cree recently. it seemed to make the led go yellow.

Stu

 

[PeterScowcroft]

Seoul P4.
It ran fine for a few months. Then I left the battery connected to the driver from taskled for a weekend (normally used for 5-7hrs then battery disconnected) and it yellowed!

 

[barkingmad]

500ma - what voltage?

 

[carrot]

Seoul P4's have a tendency to tint shift after running them for an extended period of time at surprising drive levels (500mA would do it), it seems. Try some Cree XR-E's instead? The Crees appear to have better longevity.

 

[PeterScowcroft]

I am using the taskled driver to set voltage so probably providing exactly what it needs?

The chips have mostly be running 60-125mA in use and short bursts of 500mA.

The problem was the torch when I finished using it worked fine, when I turned it on a few weeks later it was yellow from the first second!!

The thing I am trying to work out is it is the chip in this one spot that keeps on going, there are no problems with the others.

 

[sortafast]

Ok so I am noticing this problem as well. Although I am using the EdisonOpto LED but they use the same Cree die. It started out with a very very slight yellow ring around the outside of beam. Then it slowly got a little more noticeable and thicker. So tonight I got some of that fancy glow paint to play with, so I thought I would put some on the optic holder and fire it up for abit and see how it glows. Well, i kinda forgot that the bike light was on and came back about 20 minutes later. when I did, i noticed that the the very center is the only spot that was not yellow, and the outer rings are yellow/amber/almost greenish yellow. Also the beam pattern is no longer round, and kinda lopsisded. I did notice that the light body had gotten very warm. A result i do belive, from me not tollerencing the heatsink to the body properly and not using a thermal compount between the body and the sink. The set up is my garage made bike light. Its got the Edison LED, a GD750, running off a 18650 2400mAh Li-ion batt. Battery probably was somewhere between 2/3-3/4 charge, and the light all together probably only has 1-2hrs of total run time, most of which was running DD. Only thing that I can think of is either a.) the Edison lights cant quite handle the beating I was delivering to it, b.) the tint shift and beam pattern degredation are a result of poor thermal management (kinda leaning this way on it), or c.) I like pie. Either way, I got some SSC p4's on the way to mess with and see if i can get a better result with them. I didnt have any tint shift issues when i had a cree P3 in a similar light housing, but running off of a 6v NiMH batt pack with a DB750 runnin the show. Ran that set up for a while with no problems at all, and no tint shift. I am sure had there been air flow around the light, the thermal stuff would have been a non issue. oh well, have 4 more edisons to play with so we shall see.

 

[nein166]

I wonder if its the posphor or the silicone?
This may sound funny but, wipe the LED with Isopropyl Alcohol.
I've done it and after drying it looks like glass. I used a Qtip. Its safe to do and even suggested in the SSC documentation.
Perhaps it will clear the yellow, at the very least the dome will be cleaner.