Hmm, seems like either a win lose or a lose win…
Wonder why lumens increased after dedome and optic change. Perhaps you could measure the lumens with the dedomed xpls but with the modified optic.
Probably shaving off the inside of the optic meant that some light hits the shaved portion and then diffuses. Instead of reflecting off the edges of the optic and then out the front, some of this diffuse light goes out the bottom of the optic and is lost … absorbed inside the bezel.
Light will only reflect inside a TIR optic if it hits the edges of the optic at the correct angle. If it hits the wrong angle it will transmit through without reflecting.
Even though modding the optic doesn’t look like it affects the beam, perhaps it really is causing a loss of lumens.
About what I expected.
Firelight2 is probably right, this is why I ordered the ball nose end mill bit...to get a clean cut and a rounded cut in the hopes of keeping the internal reflectors maximized.
I charged a 20R and ran the light in the light box with the bored quad optics on the now de-domed XP-L V6 2C emitters and got 2387.4 lumens.
Then I ran the light through the box again, immediately after and on the same cell, with the stock optics and got 2849.7 lumens.
So the bored optics lost 462.3 lumens. Fairly substantial, on top of the losses already accumulated from having to de-dome the XP-L's. Previous reading with the altered optics and domed emitters was 2742.75. So de-doming cost some 355.35 lumens.
In theory then, if domed emitters could fit under stock optics they should be making something in the 3200 lumen range in this light. Only theory of course, as the domed emitters just don't fit.
I worked pretty hard on this light, carving a solid copper pill/heat sink out of a bar of copper with basic tools. The BLF driver was already in this light before and running an XM-L2 just fine. So I kept it and put it in the copper pill with the quad re-flowed to the top of it.
I'm getting some flickering issues. The output in the light box is not steady, runs up and down. And I can see occasional flickering especially in low but it's there in every mode. I drilled a hole and forced a screw into the copper at the ground ring. It broke off. I soldered it to the ground ring in the light, with the solder in contact with the side of head. Same flickering.
Much as I hate to do it, I'm probably about to tear it apart and rebuild it with a new driver. Will more than likely put 4 XP-G2 S2 1D emitters in it when I do.
A Sony C5 fresh off the charger allows 14.61A to flow, while the lightbox says 2925.6 lumens. ?
That's 731.4 lumens per emitter from 3.65A to each emitter. That amperage should have the emitter up in the 1100 lumen range, being de-domed, 1300 with a dome on.
This is with a stock quad optic over de-domed XP-L V6 2C emitters. What am I missing? (Aside from 369 lumens 4 times?)
I’m not sure.
- How well does lumens/amp scale downwards? If you use a lower-drain cell and get current input down to 5-10A or something does it give you the expected output?
- How about trying the LEDs, MCPCB, and heatsink without an optic, lens, head, etc?
- Check with welder’s goggles, are all the emitters flickering or just some?
If I put a single XP-L in an Eagle Eye X6 and bump the resistors to 3.5-3.6A it makes around 1400 lumens out the front with a reflector.
Looking at the light while on, it will run stable for a while then drop, a semi blink, and be right back to stable again. All the emitters at the same time are affected. It's a sudden blink in the total output.
Heading off to rebuild it, putting a ramping firmware in it this time but I'll use the XP-L's so when my ball nose end mill comes in I can check with it and see how that works out. ;)
Took the light apart, pretty much started over.
New Quad board, 4 new emitters, new driver, ramping UI. Soldered a short wire to the extra ground via on the BLF17DD Ver 3.1 board and cut a notch in the copper pill so this wire could come through to the outside. Once the driver was pushing into it's bay this short wire was soldered to the outer body of the copper sink/pill, then filed smooth so the sink fits into the machined ring at the base of the head on this light. So the ground wire is soldered in and sandwiched between the pill and head.
Still has a flicker. :P Go figure!
The bored optic works ok as far as looks go, it has a bigger "hot spot" with bunches of artifacts around the periphery, like many petaled flower. Only really noticeable if you're close, like a wall in a room 8-10' away. With the domes on the XP-L V6 2C the color is white. Much better. Losses or no losses, I'm liking it just like this. The ramping firmware works well, can be turned off from any place you stop along the way and comes back on where you left off. Or a press and hold from off gives low. A double click from off goes to High.
Lotta light from this small no name cheap side clicky. Not even sure where it comes from. Manxbuggy1 gave me 2 of em and they're pretty dang nice little lights, with a very strong magnet in the tail that will still hold the light horizontal even with this chunk of copper in it.
Bottom line is, this method of fitting the optics works. There may be some caveat's, but it works. The better tint makes it well worth any trouble it is to make the bit from a torx 25.
Edit: Forgot, it's now making 2988 out the front with a low of 10.5. Same Samsung 20R I've been using.
Does the flicker happen when you bridge gate to bat+?
No doubt anything we do to these optics, polished or not, will hurt output. What we really need are some purpose built XP-L optics, but for now this and dedoming will have to do.
BTW, did you compile that firmware with the red/green indicators enabled? It’s possible that they’re getting toggled and diverting some of the power to somewhere else.
Just a thought; might be worth rebuilding with that part disabled (if you haven’t already), in case that might explain the flicker. I don’t think it makes any difference if nothing is connected to those pins, but it’s hard to say for sure.
Sorry if this is covered elsewhere. But where can you get the bits to build a triple S3?
Post #3249 through 3253 or greater. AFAIK RMM carries everything you need in his store - http://www.mtnelectronics.com/
If the alteration of the optics in this manner costs us output, then I wonder…
I got the ground issues solved in this PH-08 side clicky FastTech light. And now it’s seeing a little over 15A from an Efest 35A. The lightbox shows me 4147 lumens, from altered Quad optics! And it’s hotter than a $2 pistol on Bourbon St.!!
The 3/4” x 1” copper pill in the head gets the light too hot to hold in about 45 seconds on high. 
So how many lumens would the 4 XP-L V6 2C emitters be putting down range under pristine optics? Interesting…
If I understand correctly, dedomed XP-L has roughly the same beam pattern as domed XP-G2?
With the loss of modifying the carclo optics in mind, what would give more light at about 6A total for triples? Dedomed XP-Ls in unmodified optics, or XP-G2s?
Cool!
Some tweaks to making the drilled out holes on the emitters even more polished.
Tools you need:
- T25 bit
- Hardened steel file
- Clamp
- Ultra fine sandpaper. P2000 or finer
- Fine/Very fine sandpaper. Around P150
- A piece of round stick at approximately same diameter as the T25. You could use the tapered half of a ballpoint pen in stead of the stick.
- Glue
To file down the T25 Torx:
- Attach the steel file to the table using a clamp.
- Plug the T25 into your screwdriver, and run it at an angle on the steel file, to get the right tapering.
To polish the T25 even further you could run it at an angle on a piece of ultra fine sandpaper.
Now you could glue some ultra fine sand paper to the tapered T25 (or the ballpoint-pen) and use that for polishing the emitter. Or you could taper the round-stick:
- Lay out a piece of fine sandpaper on the table. You could wrap it around a book or something of similar size, and fix the book/sandpaper to the table with your clamp.
- Plug the round-stick into your screwdriver, and run it at an angle on the fine sand paper, to get the right tapering.
- Glue some ultra fine sand paper to the tapered round-stick and use that for polishing the emitter.
… - Observe
- Rejoice :-p
Are people still using this T-25 torx bit method to open up XP-G optics for XP-L’s? I’m considering taking a Revtronic (Nitefighter) BT40s and swapping the stock XP-G2 emitters for XP-L2’s (and I don’t think replacement optics are available for this light). Originally I was going to go with XP-G3’s, but I’m not tempted to try the new XP-L2’s.
Thanks,
-Garry
Hi Garry, yes I'm still using it as I am sure many others are. I would be cautious trying to modify any non-replaceable optics. The XP-L2 will just fit into a few of the TIR optics without any modification, so you might be able to give that a try. Of course, the XP-L2 will give you a floodier beam than XP-L HI or XP-G2/G3/Nichia 219.
Not tempted or now tempted? It’s a good idea for starters but I think a means of polishing or at least getting closer to frosted than a coarse grind will be less costly in lumens mislaid. I’ve had reasonable results attaching small pieces of fine grit to dremel bits for lapping recesses in pills so maybe a small drill bit with some tape and 600 or better w/d paper could improve things after the torx opens it up. Final polish with a q-tip mounted in a drill and some headlight lens clarifying compound. I’d give something along those lines a try.
Plastic heats up quickly under friction and you lose everything gained and have to start over if it melts so go slowly and lubricate/clean the paper frequently. You need to be very careful of bottoming out in the recess as that is a*much* harder spot to polish. I did a few optics to fit the old P7’s back in the day and that’s all I remember of the pitfalls.