Emisar D3AA is available now

:open_mouth: How did you get this to work?

It’s a battery tube (and protected 26650 battery) from a Thrunite TC20. It’s not a great fit, but it does work.

Thanks so much for the tip… it’s ugly, but the fit is surprisingly decent, and it feels great in the hand!

Sadly the DM11 tube won’t fit the TC20 head

Edit: Catapult V6 tube works great too

Now for the real question. Do we have any other tubes with usb charging that fits the head?! :smiley:

This should be added to the ‘lego’ thread

I’d ask over in the Anduril 2 thread: Anduril ... 2?

Thanks for all this 26650 tube testing, guys!

I would love a 26800 tube option for DM11, especially if it’s similar to D4SV2’s 26800 tube. The ergonomics are fantastic, and I was able to modify a KR4 clip to fit. That would bring DM11 into a different ballpark for me.

Acebeam E10’s body tube will fit DM11’s head too. Per Cheule on Reddit.

I think this light looks much better with a 26650 tube.

Any updates on 26800 colors? I just want a grey one to go along with the new flashlight I’ve been waiting to order


It’s completely infeasible on any currently-supported hardware. Keeping wall clock time requires hardware these lights don’t have.

They have some time-based functions, like the sunset timer and memory timer, but they use short, imprecise time spans between like a minute and a couple hours… However, if you want to see why they can’t keep track of the time of day, try configuring a few lights to auto-lock after ten minutes, then turn them off at the same time and measure how long they actually take to change the aux LED color indicating it has locked itself. The measured time will likely range from 9 to 11 minutes, an error margin of +/- 10%.

Then apply that error margin to an entire day. You could set the time on a bunch of lights at noon one day, then ask them what time it is 24 hours later… some would say it’s 10:00, while others think it’s 14:15.

Gaining or losing a second or two each day wouldn’t be so bad… but skewing by a couple hours per day would make a wall-clock function totally useless.

So, at minimum, they would need to get a RTC chip to keep time. And with many drivers already using up all available space on both sides of the circuit board, there’s no room for something like that. There’s also the question of adding code to talk to the chip, and adding UI functions to expose it to the user, both of which are complicated.

Wouldn’t 3500k to 4000k be much more obvious a difference than 5000k to 5700k? Presuming of course same optics in play.

What do you think about mix LH351D 2700/5500K in D4SV2 (+W1)? Very greenish?

That mix should be a fair bit below BBL I imagine. The further apart the color temperatures are, the lower the DUV of their mix will be, generally.

Yeah. I have a D4S tint-ramping between 2800K and 6500K with XP-L HI 8A+1A. It’s great, but it makes every other light look green.

So now I think I’ll have to order a KR4 with these LEDs. :smiley:

Also I have a XPL 2850/6500K mix in D4SV2 and it is terribly pinkish and distorts the colors, although I like pink there is too much of it here :smiley:

I have same problem with my KR4. Too much pink with those emitters.

It’s been a wild ride, but I’ve finally been able to thoroughly test what’s going on in with LH351D 5700K in the D4SV2. What I found can be summed up in one line: Hank sent me 5700K LH351D emitters twice, and they are actually 5700K, but appear much greener and warmer in the D4SV2.

I wanted to get to the bottom of this as much as anyone, believe me! Let me give you a little timeline to understand better:

1) I receive a KR4 W1/LH351D 5700K. It instantly is my favorite light.
2) I order and receive a D4SV2 and it looks very different.
3) I email Hank, and he asks about what it says on the order, and the box, then says he’s sending a MCPCB out to me. (please note, he didn’t ask me to test without optics at this point, and I wish we had thought of that)
4) I make a video about the two lights (the KR4 vs D4SV2 dual channels) because I think they’re awesome. Based on the email chain previously I mention that Hank sent the wrong emitters and he’s sending out the correct ones (He did, and I know realize the original emitters are actually 5700K as well)
5) People question my testing methods on BLF, and I try to explain that even though Opple is $20, it works enough for this
6) ToyKeeper and Hank mention that the D4SV2 optic could be the cause, I ask Hank if I should test without optic, but a replacement is already on the way, so I wait until it arrives to do full testing.
7) MCPCB arrived, and I do full testing that proves original emitters were in fact 5700K, and it’s a function of the D4SV2 optic that it’s warming and shifting tint towards green

About my testing:
I tested the original MCPCB and the new one in three configurations each:

- a bench test, MCPCB only hooked up to a bench PSU (3v, 150mA)

- installed in the DVSV2 without optics or glass (mule config)

  • installed properly with aux TIR, glass, and bezel (how they arrive from Hank)

For both MCPCBs, the results were within 50K or so on an Opple Light Meter Pro.
Bench = 5800K (rosy)
Mule = 5600K
Stock = 5100K (and greener)

There was some discussion on whether the Opple was “good enough.” I found it very good for this. I wouldn’t quote anything off it with any accuracy closer than +/- 100K or so. To get an idea of how it works, and how I tested it, here is a video I made: https://youtu.be/ZH2_gTmsxkE

So the summary here is that this community is full of a lot of smart people that know what they are talking about, Hank included. Hope this helps clear everything up!

Great post, Cheule.

I did a bunch of re-testing on my samples last night as well (my original D4SV2 with LH351D’s in 5700K, an ostensibly identical replacement, and a D4V2 with LH351D’s in 5700K). I don’t have the capability of bench testing the emitters with no external influence on CCT, but I was able to test with and without the optics. I put my Opple Light Master III in my white photo booth. and “ceiling” bounced the lights to test.

With optics, they all measured in the 5100-5200K range. Without optics, they measured in the 5400-5550K range. That’s satisfactorily within the 5300K-6000K range for Samsung’s “5700K” bin, and it looks like my results line up close enough with yours. Those numbers are cooler than I measured initially, but I think that was user error on my part by measuring with the lights pointed directly at the light meter and on a low mode where the LED’s are a little bit warmer.

Thanks to everyone here for your help. Sorry I doubted you. Now I can confidently recommend the W1/LH351D 5700K (5500K now, apparently) combo in Hank’s 2 channel lights. The tint match is not as close as I was hoping (nor as close as Cheule got with his first KR4), but these are some really incredible pieces Hank has put together.

Good on you, Cheule. My initial reaction was to think “I would’ve had a hard time owning that one”, but that’s not accurate. Reading back through everything, I don’t think you ever actually had bias toward one result or the other, despite what some folks’ straw man arguments would’ve had someone believe. It’s no wonder politicians pull that sh*t all the time. It’s effective. Objectively reporting back what you found was “on-brand”. It’s odd for that to have gotten so emotionally charged for some people.

Anyway, moving on, it’s crazy the optic can make that big of a difference. I don’t really understand how that works… is part of the spectrum getting trapped in the TIR or something?