TOPIC: Reflective coating for lens

New thread to pick up from the topic in my Revan hilt build thread.

OVERVIEW:

I coated the lens used in that build with a reflective coating to improve brightness and it appears to me at least, to have worked.

Why'd I do it?

There has been plenty of debate and back and forth over the years on who has the brightest LEDs, brightest blades, etc etc. SFs method of attack on this was to create the quad XQ-E LED to bring some brute force to the party in the form of 4 LEDs vs the standard two or 3 usually seen on TRI-Cree setups.

A lot of things go into the end result of a bright looking blade, the LED, lens, blade type, blade diffusion - all play a part. But I started asking why it was that some other companies had some TRI-CREE setups that even when using the same blade were pretty close to the same level of brightness seen in the SF Quad.

When I got my ASP parts for the Revan build I noticed that the SF heatsink was powder coated black, inside and out, including on the base. My background in extreme computer cooling told me this was a big time no-no on a heatsink where you would mate the heat source to the sink. I promptly sanded that down for good contact with the copper LED star. I got to thinking.

Since the inside of the SF heatsink is black, it would have VERY bad reflective properties at best. Looking around at pretty much all the other Heat sink lens holder combos used I could see that they were all at least raw aluminum, some were polished. They would have better reflective properties than the SF heatsink, allowing them to return more light back into the lens. Then I started looking into the Carclo lenses used.

SF uses the Carclo 20 mm narrow spot TIR single lens for their Quad LED
Compared to the Carclo 20mm 18 degree narrow spot TIR lens used in most TRI-Cree setups.

Both lenses have close to the same efficiency ratings when compared with the same LEDs, low to mid 80% That does mean there is some light loss just by the nature of the lens material.

What stood out to me was that the lenses are TIR (Total Internal Reflection). I had to take a step back about 20 years and remember some things from Physics class regarding optical properties, but I remembered that TIR requires specific angles of light to work.

I got to thinking, since the SF quad LED doesn't have the LEDs sitting directly under the centerline of the lens, this meant the angles at which light coming form the LEDS would be different and could possibly not be past the critical angle for TIR, resulting in additional light loss.

So, couple the non reflective nature of the Heatsink with the LED grouping and you end up with less total efficiency. My first thought was, "We need to reflect any stray waves back into the lens".

How I did it

I had some aluminum duct foil tape laying around from some past projects that I knew was a highly reflective film with adhesive on one side. So, I cut very small triangular shaped pieces of the tape, peeled them, and covered the outside of the lens with them so that the entire lens had a reflective coating.

I opted for this so that I could nullify the TIR boundary in the case of any "stray waves" and direct them back into the lens right at the boundary point.

What did I notice?

The blade seems brighter. Side by side, just two LEDs lit in the Quad LED under the treated lens lit up an Infinity V4 blade almost as well as all 4 LEDs lit up in the non treated lens. This was a non scientific eyeball oberservation, but my camera also seems to agree, washing just as bad in either case.

I compared pictures to some other sabers and found that as far as the camera and my eye was concerned, you could barely tell the difference between all 4 LEDs or just 2 of them lit.

What does it mean?
I'm not a physicist, but it seems to me that the SF optical system CAN be improved, and in the case of a build that uses 2 LED mixes off the Quad Cree that applying the reflector to the lens appears to increase the blade brightness.

Nice breakdown! Do you think the brand of the tape matters?

I don't think brand matters as long as it's a metal tape that's highly reflective and the adhesive side uses a clear adhesive and maintains a high level of reflectiveness.

I ordered some model glue from Duratrax to test it out on a lens. Supposedly it's for use on polycarbonate model car bodies, so I'm gambling that there aren't too many chemical differences between model grade PC and lens grade PC. I'll keep my distance when applying just in case and won't check results for at least 48 hours to ensure there's no reaction.

I checked the spec sheet on the other model paint I was planning to test (I think for metal cars) and saw that only one of the solvents had notable reactions with PC. The exception was Xylene, which causes PC to swell and would ruin the optics.

On the off chance the Duratrax wrecks my lens, I have a few spares on hand.

Ideally I would have 3 of the same color LED modules so I could compare paint, tape and a standard methods, but since I don't have quite that many modules I guess I'll stick to before/after photos.

I'm curious as to how paint will affect the TIR boundary. I'm wondering if it may nullify it or reduce its effectiveness. Forcing the reflection properties of the paint to be the primary method of reflection.

Kinda reminds me of how, at work, we paint the inside of our channel letters white to help the LEDs bounce as much light out the translucent face as possible.

Guess I'll be experimenting with gloss white paint on my next install.

It would be great if someone had a light meter that could test multiple lens setups to provide some data on what type's of lens modifications do what.

I'd be hesitant to coat the lens in paint, since the TIR boundary works because of the difference in the refractive properties of the lens and the air outside it. If you create a new boundary between the lens and the paint, it may nullify the TIR boundary and force the light to reflect off the paint, which depending on the paint might be less effective than an unmodified lens.

I think the bigger difference would be a mirrored surface vs a white surface, as the mirror will bounce light in predictable angles based on origin, while a white surface will scatter and diffuse the light - though without knowing all the math, I can't say which would result in a better blade. For all I know, one might result in more base flare or something.

True. that's why I'd like to see some side by side comparisons of different setups either with the same camera under the lighting conditions, or using a light meter to measure different points along the blade.

At this point, all I know is putting the reflective material on the lens make the blade look brighter to my eyes, and my camera.

Dusting this off to see if anyone has had a chance to try this and what your results were?

I have the paint all set to go, just need to find the time to test it out. Deadlines are starting to set in again so I'll test it as soon as I'm able.

Gonna pull a Plagueis on this thread. Current status of my test (since I've been busy til now, not a ton of progress).

-Ordered 2 lime green LED modules from SF on the 30th. This was a color I haven't had before and wanted. Ordered 2 simultaneously to limit the differences in LED module design. One will be used as the control for side-to-side comparison of the final brightness.

-Received a 0.5oz jar of Duratrax polycarbonate RC car paint (Silver Streak) around the time this thread was last replied to. I double checked the spec sheet and solvent and I believe there should be no reaction between the lens material and the paint. Still need to test in person though. I haven't seen too many pictures of the paint when applied, as it seems most people use it for the underside of their RCs, but the ones I have found have varying degrees of gloss and reflectivity. Hoping it turns out reflective; if not, I may be able to use another type of paint on top of the polycarbonate safe layer.

I have a bunch of spare batteries and batter holders, but for the sake of controlling as many variables as possible I'm thinking of wiring the modules in parallel off of the same source. The discharge cap is 8-10A for the Panasonic 18650s so this should work (See also: crossguard sabers which can simultaneously support one 12W main and two 6W quillions). If I leave some wire slack between them, I should be able to place some physical barrier between the two modules to provide a side-by-side brightness comparison.

Just a quick question, Jas - can you be more specific on the areas of the lens you covered? Just want to make sure I'm doing this right if I only have one shot.

The cone section of the lens, basically the outside layer base to top.

FYI, I've also found that simply putting the SF star in a silvered heatsink/housing helps too.

Cool, I'll note the color of the heat sink finish when I'm doing my write-up. I should be able to get it done this weekend (or earlier) since Etsy says my package of LEDs was processed in Springfield, MA early this morning.

I got one of the lenses all painted this weekend. From a side-by-side comparison I can see that the optics have been changed greatly; the one with the reflective paint appears to be much more 'mixed' and focused to the central area of the cone. Because the optics are so different I'm not sure a proper comparison can be made without lighting a blade. Since I have two short blades I have everything to do that comparison, I just need to get it all sorted. Probably going to tape the LEDs directly to the blades so we can eliminate the hilt itself as a factor.

Also of note, the LED modules were both black. I added a little bit of paint to the inside edge of the heatsink north of the star to increase reflection, since honestly I didn't feel like pulling out the sandpaper this weekend. Shouldn't affect coupling too much where I put it, but unsure if the paint will remain 'stuck' once the metal heats up.

Here's a few pics of the test in progress. Haven't gotten one with the LEDs lit up yet, though.





I'm probably going to add one more coat of paint tonight before rigging up the comparison shots.

Hi Snakeeyz99,

If the paint doesn't work out have you looked into the possibility of using Specular Film to line the reflector? (I'm not sure if it needs to be on the inside surface of a reflector or if it can be applied outside like the paint).
I think you guys are trying to recreate a TIR module LED right?

kind regards,

For Tyeth

If you read my first post, I used a reflective tape originally on this project.

The idea was to improve the brightness of the SF LED. It always bothered me that the X4 star was at best on par with a tri cree, and generally about as bright as two Dice on a Tri-Cree XP-E2

Some basic physicals later I came to the conclusion that there was excessive light loss beyond the normal 80-85% efficiency of the TIR lens, likely due to the 4 dice not being perfectly centered. So coating the TIR lens in reflective material allows that extra light to bounce back into the lens and be focus back out. I saw an improvement in the brightness doing that.

Hi Jas-Ot,

Yeah sorry bout that but I wasn't around when you originally posted, I had an issue to sort out so wasn't on the forum for a while. I caught this thread when it started talking about painting the parabolic cone.

I'll let you carry on as you were.

For Tyeth

Yeah, I thought it might be easier to use a reflective paint instead of the tape to create a more seamless reflective surface. Then the issue of solvents came up, so I tracked down some polycarbonate safe silver paint to use. It's not as reflective as the duct tape or specular film, but it's easier to apply and completely smooth (though you do need multiple coats).

Here's a better look at the paint.