XinTD X3 - a close-up look (Updated with mod details)

I'm not sure if anyone has mod'ed one yet, but I started on a mod to change out the emitter/star to possibly U2 1A or 1C on a copper star, boosting up amps to 4.5A, and custom programming the driver to modes I want. I felt compelled to post what I found so far.

Update 08/18/13: See post #5 for details of the mod and the test results!

Here's the product listing:

Here's a review in progress:

I'm impressed with the light, the look, solid feel. I'm not happy it's not just a standard 26650 form factor, but has extra width with no provided spacer for a 26650 cell. I ordered mine with the T3 warm white LED, which I really like. First thing I did was put in a King Kong ICR 4200 and did measurements:

694 lumens at start, 666 at 30 seconds, throw of 36 kcd

The throw # is impressive for under 700 lumens, so this is why I decided to pump it up and see what I can get out of it. As leaftye pointed out in his review, it's a unihead construction, making working on the star difficult because you have to work down into the head. Coming stock with an aluminum star is a shame for a light of this caliber. Here's some pics of the break down.

Here is a view of them empty pill, driver removed:

Notice the depression in the center - was thinking of adding full width copper discs, but that is a big hole:

Here's where it gets strange. The pill top is nice and thick. The surface under the star is polished. I found a iPhone/iPod repair kit that had the perfect size star driver to fit the screw heads that secure down the star. For some unknown reason, they used an epoxy instead of grease under the star, Below is the removed star. It took some coaxing to get it out, specially at a angle with no leverage (straight down) but eventually gave way.

Here's a view of the head:

The bump of epoxy is caused by this:

Note in the photos above, I cleaned up the surface of the pill top by removing the epoxy (maybe fujik), then cleaning with iso. alcohol. The scratches are from the screw driver and pliers I was using to try to remove the star.

Now that hole is directly under where the emitter is mounted on the star. I measured it roughly at 1.2 mm deep, which is deeper than the hole on the underside of the pill. Why?? Why would they kill such a good design? Is it easier to machine this way?

I'm sooo disappointed now, but my plan is using a 20 mm Noctigon, though I'm considering mounting the Noctigon on a copper disc to get a bit more thermal transfer out of it, though the height may be an issue. I'm thinking it will take a lot of AS5 to fill that hole, so maybe for a filler, I'll use some copper filings I've been collecting, mixed in with the thermal grease.

Another thing is they obviously didn't use enough thermal epoxy to even completely fill up the hole, or for that matter, didn't use enough to cover the full alum star contact surface.

For a good quality, not so cheap light, I'm a bit disappointing finding this kind of deficiency. I'll pm Hank with a link to this thread so he knows. Maybe it's not that bad once you fill the hole and use a Noctigon - I'll find out.

Tom E. Thanks for starting this thread on the mod. Its almost laughable the way this has been machined. I’d suggest that the light was designed however and given to the machine shop to build without them having proper instruction. They have machined the light to the best of their ability and tooling they had to get the job done. How thick is the material the star sits on?

that big divot is disappointing to see...

I'd say the pill top is about 4 mm thick. Overall, the pill has a lot of mass for the size of this light.

If you had access to a lathe I would machine it flat and fill the gap with a piece of copper. Good luck with whatever you decide.

Finished the mod. here's some steps along the way:

I took a small piece of household copper wire and flattened it to fill the void. Thinking it's better than wasting all that AS5 in the hole

I piled up the AS5 in the center, hoping it will fill the void and spread out under the star - I think it worked based on how much came out after it was screwed down - nott too much but a little sign of it coming out.

Here's the Noctigon secured down with the stock screws

Here is a view showing the size compared to some popular lights, including it's little brother:

Closer view with it's brother:

Here's a close-up with my 2 favorite cells at this time:

Sorry, I don't have pics of all of this, but here's the full details of the mod:

  • IS XM-L2 U2 1C on a Noctigon
  • sanded pill top, sanded w/2500 grit the Noctigon bottom, AS5, copper to fill the hole, stock screws to lock it down
  • Qlite stock 3.04A driver with 4 380's added, total of 4.56A
  • kept stock 22 gauge wires
  • copper braided both the driver spring and tailcap spring
  • the tailcap spring is small, looks cheap, and spring was not hard connected, so soldered the spring to the switch contact
  • Nyogel on the threads, may have used CRC 2-26 in contact surfaces or points
  • re-programmed the QLite driver to have 8 modes (based on Dr. Jones' luxdrv):

turbo(100%), high(56%), med(32%), lo(6%), moonlight(value of 5), low moonlight(value of 4), strobe, battery check/beacon

Test Results

  • freshly charged King Kong 4200 ICR, amps: 4.26A 4.56A measured at the tail (Note: new 14 gauge DMM leads)
  • Light Output in lumens: 1,469 at start - 1,448 at 30 seconds
  • Beam Intensity (Throw): 76 kcd


  • now the lumens #'s seem high, higher than my fully mod'ed HD2010's I believe, and I re-tested the HD2010 and got same results as previous
  • the throw # is somewhat expected, I suppose - mod'ed out C8 I get 45-50 kcd, HD2010's 90-100 (re-measured mine at 92 kcd)
  • also what should be noted is how well it holds the lumens during the 30 secs. This light has excellent heat sinking with a thick pill top, thick amount of aluminum in the critical pill area, and it's one piece. So my guess is the dimple under the star didn't hurt much, whether because of the AS5/copper fill, or that there's enough surface contact between the star and pill and secured down by screws, making a very good bond.
  • it's not the easiest light to mod because of having to work on the emitter/star by going down from the head, but wasn't as hard as I thought it would be


I don't know but I think this light is a lumens monster with a nice throw for it's size. I hope someone else could substantiate this. The KK 4200 (or 4000) flat tops seem to fit fine, but I tried a KK ICR button top and it was too tight - couldn't screw down the tailcap, and took some playing to get the light to even work.

08/25: Edited amps measurements done with new DMM leads

those are some very nice numbers for a single cell 26650 light

Nice mods! 8)

Do you know if a KK 4200ICR delivers the full 4,56A, and how long it will be able to stay above 4A?

Well, throw seems right there, but lumens - I checked, re-checked, checked with another light.... Now I have a lot of confidence in the PVC lightbox, built by rdrfronty and manxbuggy1, and used by several others. TurboBB seemed to be the first with this particular PVC design, and in his tests, he got good results across the spectrum of flooders and throwers, so I'm thinking these #'s are pretty good.

It is high #'s but pretty high amps as well - I have to re-check readings, but I believe they are 4.2-4.5, in that range, which is very efficient for an XM-L2 on copper. I really, really like the KK's over any other battery because not only does it deliver high amps, but it keeps the amps high much longer than a IMR or high current 18650 cell. This is the main reason why I prefer single cell 26650 lights over a 18650 single cell.

Well, funny because I always said you can't get 4.5A on a 7135 based driver, single cell, for an XM-L2 or XP-G2 on copper. Let me re-check tailcap readings -- it was difficult on this light because of the anodizing on the threads, and the outer shell is very thin, so I've been using the top edge to connect to with the DMM lead.

nice job Tom, I may get another to play with but this one is staying stock, its too good at what I want it to do to play with.

my stepson has taken mine away loaded with rechargeable Panasonic aa’s for his week long scout camp. hope he looks after it.

It's just begging to have the head drilled out and filled with a press-fit plug cut from copper bar stock.

Nice work as always Tom! This torch is definitely moving up my list…… but I really should limit myself. :wink:

Ok - got a 4.26A measurement at the highest with the cell at 4.00v, so I'm thinking that's pretty up there. Sprayed CRC 2-26 on my 12 gauge DMM lead ends and that helped get a higher and more consistent amp reading (before I was getting only 3.0-3.7A), so, there's lies my amp measurement problems.....


XintTD X3 w/KK 26650 cell: 400 grams

HD 2010 w/KK 26650 cell, copper added to inside of pill: 396 grams

You can tell just by the feel it has weight to it. It's got a thick alum tailcap too, but the smallest, cheapest looking switch around.

Thanks for the info. I have not pushed the XM-L2 beyond 4A from a single cell yet.

Ill stick to 4,2-4,3A as absolute upper limit. Small gain for limited time in going higher.

I usually go to 3.5 or 3.85A for C8 size lights, and 4.2A for HD2010/SS T08 sized lights, but this hunk of metal I figured could do more - I don't think I've ever seen this amount of weight in such a small sized light and the uni-body construction must help. I ran it for 6 minutes straight on high and it holds up well - well warm but nothing crazy.

At 350g stock, no cells, it's heavier than a SRK or most standard 2 cell throwers like the STL-V2, TF X9, or bigger head single cell lights like the HD2010 and Small Sun T08, and all associated lights in the same form factor.

Nice job Tom. That is an impressive gain.

That looks like a nice light that didn’t quite get finished at the factory. As a stock flashlight, it seems lacking. After Tom’s work, dynoMITE! :party: It throws 551m?!!

always fun to hear your modded light Tom :beer:

Wow Tom, cool stuff! I'm satisfied with my stock X3. If it gives me a problem in the future, I'll mod mine just like yours. (I may need to get in touch with you on some of the details)