My wifes LED flashlight suddenly shines rainbows

[rockz4532]

maybe some oil on the reflector?

 

[Monocrom]

alright...got a picture uploaded before bed....

oo:

Now that's one helluva pic.

Does your wife like the rainbow beam?

 

[d1337]

Wow that looks just like the beam on my LRI proton

 

[PhotonWrangler]

Are the rainbows still there without the glass lens? If yes, there's got to be some moisture or oil comtamination in the emitter.

 

[ambientmind]

My guess is something shifted inside the emitter from the heat, then from the bumping around in the purse. I'm not totally sure, but maybe there's some part of the led that acts as a polarizer for light. Maybe that got moved and is affecting the way light comes out of the die. If you shine polarized light through molded plastic, those are the exact patterns you will get. I'm guessing the lens is plastic, or maybe the dome of the led is plastic causing these colors to be projected out of the flashlight. It would also explain the loss of output.
That, or maybe your flashlight finally came out of the closet after living in a purse for so long....:huh::shrug:

 

[MrGman]

The multiple rainbow color patterns is the repeating refringent pattern of light through plastic that is under stress. There is actually equipment that is used that can tell you how much microstrain exists from going through the rainbow pattern. If the lens or "window" in front is glass it can't be that. It is most likely that the heat has stretched and then shrunk the plastic lens bead sitting on the die itself. And only because it was well bonded and has not peeled off do we now see the stress pattern in that plastic lens sitting on the LED die. If it peels off and relieves the stress the rainbow pattern would go away. It may have partially detached from the die and the air gap is causing the reduction in total output.

I have actually worked with equipment to do this intentionally to look at the stress pattern in plastic models of mechanical equipment to study where the stress is.

to prove the point you can take any piece of clear plastic and look at it through polarized sunglasses and bend it and flex it in a light source and watch the diffraction of light come and go with the bending stress. Polycarbonates are great for this, but its not the only plastic will do this. I have done this many times. In fact I have inspected plastic samples of material that was going to be used for lenses before and after thermal cycling to see if they would exhibit this phenomenon and take pictures of them through a polarizer to see what happens.

The reflector is acting as the polarizer of sorts but should not be causing this effect directly.

The plastic bead that is the lens over the die got too hot sitting some where, probably the van as you said.

G

 

[orcinus]

You usually need two polarizers to show the stress patterns. One for viewing, one for the light source. And their polarization axes should be perpendicular or close to perpendicular.

These look like Newton rings to me. Meaning something probably got detached from something, or the coating on the reflector has gone a bit nuts.

Have you tried shining a light on the reflector from various angles and checking what it looks like?

 

[Ryanrpm]

Bet there's a break between the phosphor and the emitter, or something's pulled open making a wedge/prism gap.

+1

I vote for this.

 

[hank]

> a wedge/prism

elaborating slightly -- if there's a lens (bump of clear whatsit) that was glued over the LED with something like the clear glue typically used to assemble lenses, and that glue softened while you were baking the light on the dashboard, it could've slipped a bit leaving a prism-shaped space still filled with the glue.

Got a way to do an extreme closeup of the thing? You might want to patent it ....

 

[clg0159]

The multiple rainbow color patterns is the repeating refringent pattern of light through plastic that is under stress. There is actually equipment that is used that can tell you how much microstrain exists from going through the rainbow pattern. If the lens or "window" in front is glass it can't be that. It is most likely that the heat has stretched and then shrunk the plastic lens bead sitting on the die itself. And only because it was well bonded and has not peeled off do we now see the stress pattern in that plastic lens sitting on the LED die. If it peels off and relieves the stress the rainbow pattern would go away. It may have partially detached from the die and the air gap is causing the reduction in total output.

I have actually worked with equipment to do this intentionally to look at the stress pattern in plastic models of mechanical equipment to study where the stress is.

to prove the point you can take any piece of clear plastic and look at it through polarized sunglasses and bend it and flex it in a light source and watch the diffraction of light come and go with the bending stress. Polycarbonates are great for this, but its not the only plastic will do this. I have done this many times. In fact I have inspected plastic samples of material that was going to be used for lenses before and after thermal cycling to see if they would exhibit this phenomenon and take pictures of them through a polarizer to see what happens.

The reflector is acting as the polarizer of sorts but should not be causing this effect directly.

The plastic bead that is the lens over the die got too hot sitting some where, probably the van as you said.

G

Wow, there is always someone who knows the answer to even the most random questions. Excellent post MrGman!

 

[MrGman]

> a wedge/prism

elaborating slightly -- if there's a lens (bump of clear whatsit) that was glued over the LED with something like the clear glue typically used to assemble lenses, and that glue softened while you were baking the light on the dashboard, it could've slipped a bit leaving a prism-shaped space still filled with the glue.

Got a way to do an extreme closeup of the thing? You might want to patent it ....

And seeing the stress pattern through the plastic lens and the combination of the reflector/window without a set of polarizers is very possible. So this stretching out of non separated "semi" solid adhesive is probably the most closest to the truth. Needless to say that the LED assembly although still making as much light as it probably ever was, is ruined for the sake of being a lightsource and not a light show.

It would be interesting if you could open it up and press down on the plastic bead "lens" directly on the LED and while doing so if the rainbow pattern would disappear.

I am thinking it would mostly disappear but not completely. It might just distort and change shape while being held down.