I sanded my MC-E dome

I've read in a few topics about MC-E dome removal and how it reduced the overall brightness and what not. No one ever really compared how the LED works in a reflector before/after removal of the dome. Tonight I set out to answer the question myself.

The problem with the MC-E and using it with reflectors is the dome magnifies the image of the 4 dies. This makes the reflector think the light source is relatively huge and makes it very difficult to focus into a tight spot.

I went ahead and sanded the dome on my MC-E to reduce this magnification and let the reflector "see" the dies as they truly are. I figured sanding was the best way to remove the dome (if sanding was even possible) without disturbing any of the LED phosphor and potentially degrading the LED in any way.

I used 60 grit sand paper to get it down and then used some finer grits to polish the dome. There still is roughly 1mm of glass above the plastic base...I will possibly remove this as well but I didn't think much light was going through this sliver to hit the reflector. The dome isn't perfectly polished, but it is pretty clear. I'd say it is 90% clear.

The LED
The LED I'm using is a MC-E M-bin WC tint from Cutter. It is attached to an Aleph LE and driven by a Bi-FluPIC.

The Beam
Before the mod, I tested the MC-E in an McR38 and an McR27L. Both produced a large hotspot with large doughnut hole...not exactly pleasing to the eye. My main goal with this mod was to tighten up the beam and possibly remove the doughnut/increase throw. After the mod, the beam is significantly tighter. There still is a doughnut in the beam, but it is much smaller (as is the hotspot). The beam looks like it could be from an XR-E, albeit with a smal doughnut in the center. Looks like this was worth it so far...

The Output
Prior to the mod, I measured a "37" in my ceiling bounce test with my light meter. After the sanding, with everything the same, the light now measures a "25." As you can see, the LED produces 32% LESS light. How can this be, you ask? The Dome of the LEDs are very carefully created to allow for maximal light extraction. Once you mess with the dome, light doesn't hit the dome at the correct angle to leave and gets reflected. These reflections cause a decrease in output. Lux@1 stayed relatively the same.

Conclusion
It is NOT worth it to sand/remove the dome of the MC-E in the hopes of getting a tighter/longer throwing beam. While the beam does tighten up, the loss in output is significant and cancels out any potential gains.

I'll post up some pictures if anyone is interested.

Well first off I am impressed you managed to sand the dome down without knocking it off. Second, I could of told you what the final result would be. Been there done that. Still fun to experiment for oneself I suppose. A far better approach is to etch the dome with glass etching chemicals. Works great for me. You don't have the losses associated with a change in dome shape(although there is still a loss in intensity) and you can get rid of the doughnut.

Once you mess with the dome, light doesn't hit the dome at the correct angle to leave and gets reflected. These reflections cause a decrease in output.

Click to expand...

The domes are obviously optimized for the best possible light dispersion, but sanding it down shouldn't result in this much light loss. Obviously there's some first surface reflection going on, but it shouldn't result in that much.

My experience with flat top style 5mm LEDs is that they are more efficient when it comes to a smaller emission cone than their domed cousins. So, your idea was sound at least. Durn LEDs' though...just don't want to do as they are told.