TK's Emisar D18 review -- 3x18650 photon grenade

Does anyone know what the sustainable output is on this light?

Starting @ 14k lumens for the 5000k, what will it drop to but hold? 5k lumens?

~2,000 probably based on the charts in the OP

Photos don’t attach for some reason.

Got a aux LED board from Lexel and got it to work after 6 hours of struggling.

You don’t want to let anyone know what you learned?

Here you go.


I was tired last night and just wanted to share my results.

The + goes to I+ on the driver, the - goes to G. I tried to bridge R4 but I believe it may be possible to solder the + directly to the furthest pad of R4 (from I+) and have it work.

The steps for replicating my mod are:

Buy driver from Lexel

1. Disassemble the front of head- remove optics, desolder the big red/black wires connected to the mpcb that poke through the middle hole from the driver

2. Remove reverse polarity protection ring by prying and then unscrew the driver.

3. Solder wires to the lighted board/your aux led. Solder those wires through the hole in the mpcb shelf and to the board in the aforementioned positions. Leave a couple inches of wire. (Too much and you’ll have to cram the wires in later, too little and they’ll be too short)

Test everything!

4. Pull the two original thick cables through the mpcb hole. I did this buy soldering a thin wire to the tips of both thick wires to help me pull them through. I can draw a diagram if someone wants it (send me a pm).

Test everything!

5. Reassemble everything.


There is an easy way to subscribe without making a post. Just look at the bottom of post #1. There is a button you can click.

I want to update Anduril on my D18. This red ring that covers the screws is glued or what?

Yeah, it’s glued. I pried mine off with the back end of a nail so I could remove the retaining screws and the driver.

It’s also pretty inconvenient to take out the optics and unsolder the MCPCB wires before reflashing, so I almost never do it. I hope if there’s ever an updated version, it’ll have convenient reflashing access.

My buddy says he is going to gift me his D18, the one I swapped in LH351D’s, so I will have the entire Intl-Outdoor collection. He says it can sit over here gathering dust as easily as at his place. Lol

Who knows, I might actually put it to use!

Lately my favorite way to use soup-can flooders is to remove the optics so it runs in mule mode, then ceiling-bounce it for photo fill lighting. I only run it at ~6000 lumens for a few seconds per shot, but it sure is handy having so much omnidirectional light to eliminate the shadows.

Here are a couple shots taken that way, at night, with a single flashlight as the only light source. I just grabbed a couple random items to use for test shots:

Just wondering how many people have measured the lumens of their D18s, perhaps with SSTs?

In my quasi-calibrated sphere I get 15-18k lumens depending on the cells I use (brand new VTC6 and 30Q or older VTC6) with mine which has 5000k SST-20s.

The VTC6 are clearly more powerful but just wondering if it’s likely this light is putting out slightly more than the claimed 14k.

Absolutely in love with this light. Insane output in a lovely compact form factor and terrific build quality.

No 18k lumens is not realistic. I think quasi calibrated answers your question, lol

What about 15-16k?

I say ‘quasi’ because it’s accurate with cool white lights, but I had to apply a different correction factor for neutral/warm white based on lights that I already have, so it shouldn’t be wildly inaccurate.

Maybe with VTC6 at 0s. It’s the worst way to try to guess calibration. You should use a regulated mode but I don’t have that LED in mine.

But again, is it fair to calibrate a light box with a small 300 lumen light and expect the box to be accurate in testing a 15-30,000 lumen monster? Radically different output between the calibration lights and real world lights.

My light box tested quite close when looking at the calibration light or similar small output lights, but correcting it to exactly match the calibration light costs several thousand lumens when testing a 23,000 lumen beast. Fair? Accurate? Who really knows… if I knew where to take Ham’r to test it on an actual sphere I’d do it and find out where reality is… but what company that owns a very expensive sphere is willing to test my creation? And so assumptions are made by necessity…

18k lumens would have to be pulling close to 60 Amps which is how I came to my answer. I don’t think 6 SST20 are going to pull 20A from a VTC6.

I have 17 emitters in my Ham’r build, based on a Q8 body. I’ve tested quite a few single cells powering the 17 emitters and have seen current draws well above 20A from a single 18650, even up to 28A. I didn’t have a VTC6 to test but a VTC5A pulled 26.2A. This was all 17 emitters though, don’t know what it would do with 6 SST-20’s.

I don’t know about this particular light and I don’t have one here to test but may soon. This one of my buddies down H-town way came with SST-20’s and I do believe I tested it before I swapped emitters, now if I can find that paperwork…

Floodier or throwier lights than you calibrated with, and perhaps even wider or narrower apparent sources (eg - multi-emitter) are known to affect measurements.

Also, dedicated light meters should be pretty linear, but I’ve seen data from tests of smart phone ambient meters showing they are usually far from linear.