PK Design Lab
Page 1 of 2 12 LastLast
Results 1 to 30 of 43

Thread: dedomed a xm-l with success

  1. #1
    Flashaholic*
    Join Date
    May 2007
    Location
    Greece/Hellas/Crete
    Posts
    730

    Default dedomed a xm-l with success

    I don't have much time, i will write a few comments, later will post more details.
    The led is a xm-l U2 2S from cutter, pure white color, was damaged during soldering, but dedomed with success. Can tell you how to do it with a good success rate. Same reflector, 7135 based driver same current, in a xeno F8V5 host:

    Was a normal domed led in the flashlight,a xm-l U2 1A:
    580 OTF and 9500 lux
    now the dedomed xm-l,
    488 OTF and 19450 lux!!!!!

    Tint shift to ~4800K-5000K(!!!!) (as i said it is a U2 2S from cutter)
    Looks slightly more yellow than 3S tint that i have, i think it is now in this category.

    lost a 15% of light(assuming both were performing about the same,can't remember well, but i think they did perform the same in lumen testing before dedoming),got a neutral tint and got more than double increase in lux!

    Now this is WOW!

  2. #2
    Flashaholic* Walterk's Avatar
    Join Date
    Jan 2010
    Location
    Netherlands
    Posts
    739

    Default Re: dedomed a xm-l with success

    Lame teaser, tell us more !

    Did you remove the jelly dome in full?
    Did you cut the dome off? How ?

  3. #3

    Default Re: dedomed a xm-l with success

    How does this affect the color separation?

  4. #4
    Flashaholic* Morelite's Avatar
    Join Date
    Nov 2005
    Location
    Central PA
    Posts
    2,237

    Default Re: dedomed a xm-l with success

    Please show and tell how you dedome it.

  5. #5
    Flashaholic*
    Join Date
    May 2007
    Location
    Greece/Hellas/Crete
    Posts
    730

    Default Re: dedomed a xm-l with success

    I will try to dedome a third led in a while and i will give you all the info needed. The main idea of the method is with an oven or something to produce heat under the led. Not very high temperature, just the one needed to solder it to a pcb. Then the core becomes soft and delicate, with some press in specific points it is possible to dedome it with success.
    I removed the whole silicon core, it pops out like in sst50/sst90.
    I don't know why the color changed, we will have to ask someone more expert or try to find some more information about it. Will post some photos too soon, sorry for not having the time currently to complete all the details.

  6. #6
    Banned richpalm's Avatar
    Join Date
    Jun 2003
    Location
    Central Pa.
    Posts
    965

    Default Re: dedomed a xm-l with success

    Had a T6 that got dedomed when the GD file slipped. Still works but got horribly green hotspot.

    No more filing stars!

  7. #7
    Thread Killer Illum's Avatar
    Join Date
    Apr 2006
    Location
    Central Florida, USA
    Posts
    12,450

    Default Re: dedomed a xm-l with success

    The dome of certain LEDs do assist in shifting the color a bit. I once sheared off the dome by accident on a CREE XRE CW and the output turned NW

  8. #8
    Enlightened bmmeadors's Avatar
    Join Date
    Mar 2011
    Location
    Central Arkansas
    Posts
    66

    Default Re: dedomed a xm-l with success

    Wow, with those lux numbers, this might be a candidate for a monster XM-L thrower! Cool find!
    -When life gives you lumens, make lumenade.

  9. #9
    Flashaholic* Walterk's Avatar
    Join Date
    Jan 2010
    Location
    Netherlands
    Posts
    739

    Default Re: dedomed a xm-l with success

    Maybe you could measure the lux@1 meter of the led only (without reflector/lens) before and after?
    Now that would give me a lot more insight...

  10. #10

    Default Re: dedomed a xm-l with success

    Pardon me for not knowing better but I don't quite understand the excitement over building a light with a de-domed XM-L. The issue for me is the ~ 20% drop in output as measured in lumens (OTF). The higher lux readings are likely because the light from the die will be more directional. The LED is reduced in output due the loss of out-coupling efficiency for the light emitted at a high viewing angle from the die.

    I would be interested in measured spectral data gathered using a photogoniometer (as opposed to an integrating sphere) to understand the color shift. The XM-L has a lot of color separation over the viewing axis so it won't be a big surprise that the color temperature would shift with the dome removed.

    We have de-domed a few LEDs ourselves including the XM-L except that we want everything off including the phosphor layer. Several hours soaking in IPA does the trick along with a couple of hours spent picking chunks of silicone off the die. The real challenge is not scratching the die surface or damaging the wire bonds.

  11. #11
    Flashaholic*
    Join Date
    Jul 2010
    Location
    Sydney, Australia
    Posts
    995

    Default Re: dedomed a xm-l with success

    Quote Originally Posted by Harold_B View Post
    Pardon me for not knowing better but I don't quite understand the excitement over building a light with a de-domed XM-L. . . . . . The higher lux readings are likely because the light from the die will be more directional. . . . . .
    I think you've answered your own question !

    For some people the sole aim of building lights is to have the brightest spot, no matter how tiny it is.

  12. #12

    Default Re: dedomed a xm-l with success

    Well, perhaps so Mike. I understand the desire for a bright spot. That one is easy. Getting a brighter spot by making the LED more directional while making it less efficient (lowered output) might be more economical than capturing the low angle output and directing it to the spot using appropriate optics. It seems like a waste of potential output to me so I thought I'd ask in case I was missing something. Not that one approach would be absolutely right while the other is wrong, just asking a question.

  13. #13

    Default Re: dedomed a xm-l with success

    Quote Originally Posted by Harold_B View Post
    Pardon me for not knowing better but I don't quite understand the excitement over building a light with a de-domed XM-L. The issue for me is the ~ 20% drop in output as measured in lumens (OTF). The higher lux readings are likely because the light from the die will be more directional.
    The excitement is about increasing the throw in a LED that's more floody most of the time.

    (Oops - too late!)

    I have the need for both types of light. During the day I need to light up shadows 50 - 100 feet away so a small but very bright spot is exactly what I need for that light. It ends up being the single light I use most often but only because I can always choose from several lights for any night time task.
    Last edited by RCantor; 09-04-2011 at 06:59 PM.

  14. #14
    Flashaholic*
    Join Date
    May 2007
    Location
    Greece/Hellas/Crete
    Posts
    730

    Default Re: dedomed a xm-l with success

    Guys, the challenge is that you can get in a P60 host 20.000 lux with quite a good number of lumen. Ok, you get a loss, but still enough lumen out there. In these small tactical flashlight hotspot in needed to be bright.
    Harold, can you describe the procedure you did more analytically?
    I will write about mine after i finish my third good success.( i have already 2 fails in the beginning, but i saw that heating the led to a high temperature helps the core to pop out)
    My next step is to put the led@4amper in a 73mm reflector.

  15. #15
    Enlightened bmmeadors's Avatar
    Join Date
    Mar 2011
    Location
    Central Arkansas
    Posts
    66

    Default Re: dedomed a xm-l with success

    If we're looking for more lux in a P60 with little / no spill then maybe try an Ahorton's Aspheric.
    If you are looking for less throw but with some spill then maybe try the Manafont (Ultrafire and sku: 6047) smooth parabolic reflector. There's a thread about it being used with an XP-E here. But I would think that it would give better throw than most P60 reflectors.
    Whatever you decide to do with it, good luck, and please keep us updated.
    -When life gives you lumens, make lumenade.

  16. #16
    Flashaholic* Walterk's Avatar
    Join Date
    Jan 2010
    Location
    Netherlands
    Posts
    739

    Default Re: dedomed a xm-l with success

    I was thinking of a dedomed led AND aspheric, the best purpose of a dedomed led.

  17. #17
    *Flashaholic* easilyled's Avatar
    Join Date
    Jun 2004
    Location
    Middlesex, UK
    Posts
    5,314

    Default Re: dedomed a xm-l with success

    Quote Originally Posted by Walterk View Post
    I was thinking of a dedomed led AND aspheric, the best purpose of a dedomed led.
    Now that's an interesting idea.
    Change is inevitable, except from a vending machine

  18. #18

    Default Re: dedomed a xm-l with success

    Quote Originally Posted by ergotelis View Post
    I don't know why the color changed, we will have to ask someone more expert or try to find some more information about it.
    There was a thread a while back that touched on the subject of color shift and lumen loss in an LED with its dome removed. The dome of a white LED reflects some of the light back onto the surface of phosphor covering the blue emitter. This reflection causes a secondary yellow or orange fluorescence (cool or warm) that increases the lumen output. The secondary fluorescence has no blue component because it is absorbed by the phosphor. The color shift is caused by the absence of this extra yellow or orange when the dome is removed. Or more precisely a higher ratio of blue than an LED that has a dome.

    If you look at the typical cool white LED's relative spectral power distribution graph in the data sheets, there is a transmission peak at ≈460nm that is about double the intensity of the level at 555nm. Significant enough to cause a color shift in its absence. The eye is much more sensitive to changes in color than intensity.

    Someone in this other thread argued a reflection couldn't possibly increase output because there are no new photons being created by the reflection from the dome. Fluorescence is never 100% efficient. It can't create a beam of light more energetic than the original excitation source. It would be more beneficial to multicoat the dome and let the would be reflection pass right on through.

    What they didn't to take into account is the lumen is by definition an adjustment of radiant flux power levels against the human eye's sensitivity to different wavelengths of light or luminous efficacy. The eye is ≈17 times more sensitive to 555nm green (peak photopic sensitivity) than it is to 460nm blue. The secondary fluorescence isn't creating new photons but it is converting them into wavelengths easier for the eye to see.


  19. #19

    Default Re: dedomed a xm-l with success

    I would be inclined to argue that the color shift due to the back reflected photons from the dome is not correct. GaN materials (InGan, P-Gan, ...) all have a high index of refraction and absorption coefficient in the visible wavelengths. Basically, anything that is a good emitter is a good absorber. The reflectivity of the die can be calculated using the refractive index which is about 2.45 at 450nM and then generating a best fit curve based on the real and imaginary components of the Reflection Coefficient. At the angles the photons would be back scattered to the die the reflectivity is about 6 - 25%. Mostly less than 10% for any visible wavelength striking the die at 0 to 45 degrees.

    The purpose of the dome is light extraction. Making the ray path from the die encapsulant exit surface perpendicular at as many points as possible increases the out-coupling. Removing the dome creates loss at the surface due to TIR. If the light should make it back to the phosphor layer and be reabsorbed by a phosphor chrystal it will be scattered isotropically. That is most of it, if not all will be scattered into the die. Some remote phosphor designs use light recycling schemes but there is typically a gap between the phosphor layer and the die or there is a filter that prevents the scattered light from being sent back toward the die surface.

  20. #20
    Flashaholic*
    Join Date
    Jul 2010
    Location
    Sydney, Australia
    Posts
    995

    Default Re: dedomed a xm-l with success

    If light from the phosphor crystal is scattered isotropically, then the percentage of light going outwards would be the same, regardless of whether the blue light came from the chip - or reflected back from the dome.

  21. #21

    Default Re: dedomed a xm-l with success

    Yes, exactly. I'm probably not being very clear by trying to be brief. Approximately half of the light from the die (before or after down conversion) will be back scattered and re-absorbed by the die surface. Whatever else makes it back will suffer the same odds, plus the light is likely to be down converted to a wavelength not absorbed by the phosphors so it will be scattered (minus Fresnel loss). There are a ton of papers describing the efficiencies of phosphor down conversion and die modeling. This is a quick one that talks about back scatter and loss: http://lightingresearch.org/programs...u-SPIE6337.pdf

  22. #22

    Default Re: dedomed a xm-l with success

    Clearly you guys have a better understanding of the nature of LED's than I do. Here is a picture from one of the threads I was thinking about. Saabluster had taken a picture of a Luminus CBT-90 at the same current level and exposure. The two pictures are with and without a Wavien recycling collar. The CBT-90 has no Lambertian dome, so the increased output and shift in color temperature must have been caused by some other factor(s). If not secondary fluorescence from a reflected blue ray of light, what is the cause?


  23. #23

    Default Re: dedomed a xm-l with success

    Thanks for including the links in your post Al. Very helpful. Seven page thread, 11:30 at night. It's going to take a bit to digest all of that! At a glance I am only familiar with the Waiven optic via their patent info. I've not had a chance to mess with one. My first impression is that the optic is used to capture the emission for increased brightness (as opposed to increased output or making a change in the numerical aperture). Photos are nice but there's no substitute for hard data when it comes to assessing what's going on. I would be interested in data from a spectrometer and output measurements gathered using an integrating sphere to gage the extent of whatever changes were made.

  24. #24
    Flashaholic*
    Join Date
    May 2007
    Location
    Greece/Hellas/Crete
    Posts
    730

    Default Re: dedomed a xm-l with success

    I think it is good to post here some first results with other reflectors.
    With trustfire x7 @3,5amp on the led, i got ~172,000 lux(!).Led is one of my crappy t6 with one of the 3 wires of the core cut.But still works!Before,on this host it was a xm-l U2 2S performing ~100,000 lux.

    By the way, what is the record of lux number with a led?

    On a trusfire C8 classic host with the 41mm widex31mm height smooth reflector, i got@ 2,7amp about 48000 lux. Owned my catapult v2 xm-l(!!!). Led is T6 3S with very few defects on the core.
    Till now, i have only dedomed 3 leds, 2 were ok on the one i got a wire cut, but still works,don't know if and how it affects its performance. Will see. The method as i said is at a good heat level of 150-180 C, to press gently the dome with tweezers from left and right of the wires and not close to them(we should not press the dome on the point where the wires are on the core).It will pop out after some time almost easily. But remember, this only for real crazy modders and those who don't have problem to waste their mine(well it is not fine to waste your money but i am maybe a lot more crazy)


    Currently, trying to measure the lumen loss.
    Last edited by ergotelis; 09-08-2011 at 09:09 AM.

  25. #25
    Flashaholic* Walterk's Avatar
    Join Date
    Jan 2010
    Location
    Netherlands
    Posts
    739

    Default Re: dedomed a xm-l with success

    Quote Originally Posted by ergotelis View Post
    I think it is good to post here some first results with other reflectors.
    By the way, what is the record of lux number with a led?
    I got 2.100.000 cd with a single led.
    But it means nothing for comparison, thats why I hope you would measure the Lux@ 1 meter without any reflector, tir or lens.
    Because only with that measurement you can calculate the surface brightness of the modified led.
    (If you have measured same set-up @1meter without optics before modding, then you really can tell its effect on throw....)

  26. #26

    Default Re: dedomed a xm-l with success

    Hello ergotelis! How well would the phosphor hold in the long run? Isn't its thin layer easy to scratch or to shed itself?

  27. #27
    Unenlightened
    Join Date
    Sep 2011
    Posts
    1

    Default Re: dedomed a xm-l with success

    Quote Originally Posted by ergotelis View Post
    I don't have much time, i will write a few comments, later will post more details.
    The led is a xm-l U2 2S from cutter, pure white color, was damaged during soldering, but dedomed with success. Can tell you how to do it with a good success rate. Same reflector, 7135 based driver same current, in a xeno F8V5 host:

    Was a normal domed led in the flashlight,a xm-l U2 1A:
    580 OTF and 9500 lux
    now the dedomed xm-l,
    488 OTF and 19450 lux!!!!!

    Tint shift to ~4800K-5000K(!!!!) (as i said it is a U2 2S from cutter)
    Looks slightly more yellow than 3S tint that i have, i think it is now in this category.

    lost a 15% of light(assuming both were performing about the same,can't remember well, but i think they did perform the same in lumen testing before dedoming),got a neutral tint and got more than double increase in lux!

    Now this is WOW!
    can you share the pic?

  28. #28
    Flashaholic* Walterk's Avatar
    Join Date
    Jan 2010
    Location
    Netherlands
    Posts
    739

    Default Re: dedomed a xm-l with success

    How did the experiment work out?
    You still like the de-domed XM-L?
    Does it hold, or is the silicongel or phosphor acting up?

  29. #29
    Flashaholic* saabluster's Avatar
    Join Date
    Oct 2006
    Location
    Garland Tx
    Posts
    3,545

    Default Re: dedomed a xm-l with success

    Quote Originally Posted by Walterk View Post
    How did the experiment work out?
    You still like the de-domed XM-L?
    Does it hold, or is the silicongel or phosphor acting up?
    The phosphor seems to hold up for me. I dedomed the LEDs in the latest LR+. I have destroyed a little less than half of the LEDs that I attempted to dedome. Sometimes it is the bond wires and other times it is the phosphor pulling up. Can be quite a nightmare. It does drastically increase performance though.

  30. #30
    *Flashaholic* easilyled's Avatar
    Join Date
    Jun 2004
    Location
    Middlesex, UK
    Posts
    5,314

    Default Re: dedomed a xm-l with success

    Quote Originally Posted by saabluster View Post
    The phosphor seems to hold up for me. I dedomed the LEDs in the latest LR+. I have destroyed a little less than half of the LEDs that I attempted to dedome. Sometimes it is the bond wires and other times it is the phosphor pulling up. Can be quite a nightmare. It does drastically increase performance though.
    To be pedantic, I'd say that it drastically increases throw rather than performance.
    Since the total output is actually decreased after dedoming, I don't think its correct to use the word performance.
    Change is inevitable, except from a vending machine

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •