The evidence is not on the CPF thread, but here from everyone who has tried it: however botched the job, usually (or always) resulting in less total lumen output but always more intense hot spot. Looks like usually a smaller hotspot as well. This seems to be the case with a wide variety of reflectors.
I have zero experience with attempting to focus a de-domed XM-L, but may consider it on a SMO P60 I have with a XM-L T6 3C and pathetic multi-colored ringy beam profile. (currently covered with DC-Fix diffusion, which makes for a wonderful smooth even light.
My experience with the mule light de-dome is that it definitely reduces the light ‘focused’ straight out the top and increases the light spread at greater angles. This is quite evident without precise measure.
Cree provides the Data on what the beam intensity profile looks like out of the dome. What would help to answer some of these questions is if someone were to make those same measures from a de-domed XM-L.
Haha! Yeah, there seems to be a few different methods for final polishing. The felt wheel on my Dremel cleared mine up, and then I suppose I could do some more hand polishing with a glass cloth as well.
Um, I would advise anyone with a light they care about, that works fine, NOT to do this!
Try this with stuff in your spare parts bin first. …and have a replacement handy if things go wrong. All jokes about bricks and butter knives, if you do go for it, increase your chance of a successful outcome, take the time to do it as precisely as possible.
For those who have made a rough frosted surface and trying to polish it out;
Try this:
Clean the polish rouge and sanding grit off well with a Q-tip and alcohol.
Mix a small amount of clear epoxy and carefully apply a tiny drop with a toothpick to restore a nice smooth glossy surface. I did this with mine a few hours ago. Looks good.
Edit: make sure your clear epoxy is temp rated; Loctite makes some good to 149C.
You are probably right that there is far more complex interaction between the dome optic and the reflector contributing to the beam profile. If anything, the bare emitter makes a simpler optic.
If we make the observation that in practice lensed lights tend to throw much better than reflectored ones for similar diameter, even though theory claims otherwise, then it reasons that the reflector designs are flawed. It might even be an inherent physical limitation with how reflectors can be constructed and not just incompetence/laziness; we just don’t know why. It’s obvious that high power led emitters were not designed for throw in reflectors.
It would be interesting to compare dedomed emitter + reflector to aspheric of similar size, and then dedomed to domed in aspheric. I suspect all of these would be similar.
Alright, I just couldn’t help myself after seeing more info from LowLumens and HAL, I decided to go ahead and de-dome one of my UF 3-mode XM-L T6s (I have three, so…)
As I noted above with my first successful effort, I did the same here, but the LED retaining ring in this drop-in actually provides somewhat of a solid guide that kept my blade from slicing the dome too deep. So, I got a more even cut, a bit closer to the emitter. Used the same dremel methods to polish: first the polishing point stone to even it out and remove imperfections, which left the frosted surface. Then switched to the felt wheel and buzzed it til the frost was clear. This time, I also hand rubbed the surface with a glass cleaning cloth, and it appears to be completely clear.
Here are the preliminary indoor results.
Spots on the white wall from 1m, both identical MF UF 3-mode XM-L T6 drop-ins running a current of close to 3A. De-domed on the left, unaltered in the right.
Can’t wait til it gets dark out. Thanks again guys. :bigsmile:
My aspheric-lensed single mode Q5 has a tight spot and no spill, and has a decent throw, albeit without the power I wish it had. XM-L in the same set up is just a bigger spot. Based on what I’ve learned here in this thread and otherwise, I suspect putting a de-domed XM-L in the same setup will reduce OTF light with a blacked-out reflector, and be messy as hell with a normal reflector. Doesn’t that sound about right?
With a lens though, all of the emitter light can be brought to focus, where the reflector can only focus the light that strikes it. Only light significantly off axis from the emitter touches the reflector; Most all of the ‘spill’ light never touches the reflector.
Note; The carclo optic (or a camera lens) can project a perfect image of the LED die with the de-domed LED. The domed will project a round dot.
Jeeze, man. Why bicker over the details to the point it changes the mood of a thread? This is supposed to be about de-doming XM-Ls and the actual results of doing so… not all the intricate details of what causes what. I don’t care how smart anyone thinks they are or how bad they need to prove it. I just like to play and learn.
I’m gonna skip out on this thread too. Thanks guys. I liked my little project and the results. Hopefully someone dug the beam shots. Later.
I just did the test myself and observed this with my own eyes: first time I saw an XM-L clearly projected on my wall. sections, bonding wires and all. Really cool.
Hey JaffoAZ stick around. I know a few folks asked about the impact on Luminous LED. Please note my procedure was taken from somone who knows much more than i do. Im familar with the end results only. I can say the differnce is somewhat dramatic on de-domed SST-50. Once the lens is and even a mild driven SST-50 will display a much tighter hot spot. I havent tried them with a TIR lens but i understand its an entirly
different beam. Not sure about the lower output but it appears logical that some loss
will result from such an operation..
I will try the same method on an XM-L and let you know if my assumptions are
indeed correct..Both LED lens use the same method to mount a lens and should
be the same to remove.
