[~25 Runtimes] – Emisar D3AA 4500K Nichia 519a vs Zebralight 4500K

[NightTime]

Before diving into the results, please take a moment to read through the following key points to understand the context of this comparison.

Introduction

I purchased two versions of the Emisar D3AA:

• 4000K dedomed Nichia 519a : My actual go-to EDC light.
• 4500K domed Nichia 519a : Bought for science

I was interested to find out whether the D3AA can rival what I consider the most efficient AA light ever: the Zebralight SC5w MK II. While I don’t claim one is outright better than the other, the SC5w has set a high bar for NiMH efficiency.

To level the playing field, I selected the most efficient 4500K emitter Hank offers: the domed Nichia 519a as suggested by TheFreeman. As far I as know, the XP-L2 found in the SC5w is a bit more efficient than the 519a so it’s not completely fair for the D3AA. I know the SC5w is discontinued, but I have no plan to buy a SC54c either.

Testing Methodology

Since sustained brightness is crucial to me, I tested both the host and driver independently. As you know, there are various strategies to help a host manage temperature, such as using specific materials, employing active cooling, adjusting the temperature ceiling, adding cooling fins, and of course the mass itself. However, not all drivers are created equal, so even if you cool down the host, a low-quality driver won’t sustain any level. By running temperature-sensitive levels with and without active cooling, I obtain valuable information about both the host and the driver.

For Anduril 2 (D3AA):

• Ambient temperature sensor calibration was executed
• Thermal ceiling was set to 55°C which is the very limit for my hands.
• Ramp style was set to 7 steps ranging from 1 to 150. Note that for some graphs, I played with the ceiling to compare with the Zebralight.
• Cells used:
  • Two Vapcell H10 14500 (<15 cycles)
  • Two Eneloop Pro NiMH (purchased January 2022). Note that these are not brand-new, so both capacity and resistance have changed over time. To stay fair with the D3AA, I re-tested the SC5w using those, instead of comparing with my older runtime results. The D3AA can do better using brand-new cells, but the SC5w also.
  • I added a bonus alkaline runtime, but only for the max level.

All tests were conducted in a rudimentary amateur setup, utilizing a shoebox, a cellphone and Zak Wilson’s Ceiling bounce app. I used an array of lights to calibrate the shoe box. Although, I am very happy with the results and get consistent results with other top reviewers, please don’t take these numbers as the absolute truth.

The contenders:

Emisar D3AA (4500K domed Nichia 519a)

Zebralight SC5w MK II (4500K XP-L2)

Runtimes

D3AA - Vapcell H10 - Level 7/7:

Appears to function as a FET-driven level.

D3AA - Vapcell H10 - Level 6/7:

921 lumens for 18 minutes is excellent for a 14500 battery.

D3AA - Vapcell H10 - Level 5/7:

Hold the light in your hand, and this would probably be the first level to be sustained by the host.

D3AA - Vapcell H10 - Level 4/7:

Just for fun:

After playing a bit with my spreadsheet, I found that the D3AA level 5 is roughly the same as level 10 from my Zebralight SC64w HI using the Cree 4500K XHP35. The D3AA is able to run for 46 mins at this level using a 1000 mAh battery. Scaling this to a 3500mAh battery means we could except 46 x 3500/1000 = 161 min and the SC64w is actually running for 153 mins. Great job Mr. TheFreeman !

D3AA - Eneloop Pro - Level 7/7:

D3AA - Eneloop Pro - Level 6/7:

Yes, level 6 starts at a higher level. Don’t ask me why. I tested a second time. Maybe brand-new cells would make a difference here between level 6 and 7.

D3AA - Eneloop Pro - Level 5/7:

D3AA - Eneloop Pro - Level 4/7:

D3AA - Eneloop Pro - Level 3/7:

Both lights on high levels:

The D3AA is really making a great job pumping all the juice an Eneloop can give.

D3AA vs SC5w Comparisons - Eneloop Pro - Level 91/150 match:

After 7 years, the SC5w finally meets its match!

D3AA vs SC5w Comparisons - Eneloop Pro - Level 83/150 match:

D3AA vs SC5w Comparisons - Eneloop Pro - Level 37/150 match:

Zebralight remains unbeaten at lower levels, as discussed in this thread by BLF/thefreeman

D3AA - Alkaline Battery - Level 150:

~100 lumens sustained for just over an hour. While the alkaline performance is respectable, the Eneloop Pro demonstrates significantly better capabilities, even when not brand new.

Conclusion

The Emisar D3AA is delivering exceptional performance across various power sources. While the Zebralight SC5w still excels at low levels, the D3AA showcases its strengths with a modern UI, a highly efficient driver and emitter pairing.

Cheers !

[jon_slider]

Thank you for the excellent info!

My conclusion from reading your post is that the D3AA has the same runtime as the Zebra, at a 200 lumen output level.

And the Zebra runs 50% longer than the D3AA at the 12 lumen level.

Advantages of the D3AA include, optional tail magnet, dual fuel, Aux lights, NoPWM, better LED options, and Anduril.

Advantage of the Zebra is potting, but for me, the Zebra UI is a total dealbreaker.

[NightTime]

I agree. I used to be ok with ZL UI for years, but now that I tasted Anduril, I’m like meh. Not having Anduril is not a complete dealbreaker, but not that far also. Total preferences of course.

[Neon_Sunset]

Thank you for taking your time to do this and posting the results.

[dmenezes]

Great job, thanks for spending the time to produce this and for sharing it with us!

One suggestion to make it even better, tho: if you can, please use more distinguishing colors (dark green and dark blue are similar enough that I have a hard time distinguishing one from the other on my phone’s screen with night-time mode on – red and blue, or orange and green, would make visualization much easier).

And finally an ask: would you consider doing a LiIon-only comparison of the D3AA vs a less efficient but light like the Wurkkos TS10 or HD10? This would be interesting to show how much better an efficient driver would perform, therefore making people more interested in them and helping manufacturers that don’t yet offer them to go in that direction…

Just my $0.02, and thanks again for what you’d already done.

[dmenezes]

@jon_slider

My conclusion from reading your post is that the D3AA has the same runtime as the Zebra, at a 200 lumen output level.

And the Zebra runs 50% longer than the D3AA at the 12 lumen level.

Agreed, and as per the " Both lights on high levels:" graph, I would add that at higher levels (~430 lm), the D3AA beats the Zebra both at endurance (total time on) and *lumen-hours" (total light emitted over time, ie area under the curve).

@jon_slider

for me, the Zebra UI is a total dealbreaker.

Never used a Zebra, but not running Anduril makes it a deal breaker right from the start (“no Anduril, no deal”)

@NightTime

Not having Anduril is not a complete dealbreaker, but not that far also

And that makes it three of us, in varying degrees of Anduril preference

[NightTime]

@dmenezes

would you consider doing a LiIon-only comparison of the D3AA vs a less efficient but light like the Wurkkos TS10 or HD10?

Good idea. Unfortunately, it looks like I’m the only flashlight enthusiast that does not own a TS10. Because sustained output is important to me, I never bought a TS10. Still, I do have two other (cheap) dual fuel lights using 14500:

• The SP10 Pro AA (5000K): bought for my son because at that time, it was the only AA light using Anduril and I wanted to limit the output.
• Convoy T2: bought to see if I liked 3500K lights (I do).

None of them are 4500K, so it’s not a fair comparison, but it would be interesting. I can’t play much with the T2 output (ceiling) so I might match the highest level of this one with the D3AA/SP10 Anduril.

[lg.tech]

You’re not alone, I never got the TS10 either, too floody for me, low sustained output and not efficient.

[dmenezes]

@NightTime

Good idea. Unfortunately, it looks like I’m the only flashlight enthusiast that does not own a TS10.

That’s too bad

Because sustained output is important to me, I never bought a TS10.

I understand the sustained output thing, but I myself feel safer with a light that will lower output in tandem with the battery’s charge, than one that will go all the way to almost zero charge with the same output (and therefore giving me next to no warning when I’m about to be left in the dark). But OFC that’s just me…

Still, I do have two other (cheap) dual fuel lights using 14500:

• The SP10 Pro AA (5000K): bought for my son because at that time, it was the only AA light using Anduril and I wanted to limit the output.
• Convoy T2: bought to see if I liked 3500K lights (I do).

None of them are 4500K, so it’s not a fair comparison, but it would be interesting.

Agreed on all counts, and if you could do that, it would be great as it would show clearly how much more efficient regulated drivers are (I have to confess that, after EDCing my D3AA for a couple of months, I can’t see much difference in terms of battery duration compared to my TS10v1 with the same battery).

[jon_slider]

@dmenezes

comparison of the D3AA vs a less efficient but light like the Wurkkos TS10

I have some data comparing a TS10 to a D3AA here

[dmenezes]

@jon_slider

I have some data comparing a TS10 to a D3AA here

Thank you for posting that and linking it here! Will comment further over there.

[jerv]

@NightTime

Good idea. Unfortunately, it looks like I’m the only flashlight enthusiast that does not own a TS10.

BLASPHEMY!!!

Seriously though, the biggest draw of the TS10 is that it’s under $20. And I feel it’s an educational light as there are many facts about enthusiast-grade lights that it does a better job of teaching than an objectively better light like the D3AA, and at half the price. Things like the relationship between output and thermal rampdown. Or output and runtime. You really don’t need more than a TS10 can thermally sustain (and have good runtimes on) for a lot of things most people do with their lights. Maybe for a lot of the things we do with flashlights, but if you are on BLF then you are not “most people”. It’s also a low-cost entry to Anduril. If you try a TS10 and like it, then you can use a ton of good lights. If you hate it, or simply cannot grasp that there’s no law saying that you must use the features that lights like the FC11C lack despite being basically identical in really real-world use, then you’re out only half of what you’d spend on a D3AA.

Then again, if you are on BLF then you are not “most people”. You probably already know the importance of a decent driver, and really don’t need to be beat over the head with the role thermal mass plays when it comes to sustained lumens. If you truly are a flashlight enthusiast, then you already know all that a TS10 can teach you. At that point, the price is the only draw. And also the main reason I have so many of them. Yeah, the D3AA is objectively better in most ways. However, if you need a few lights for around the house for a power outage, then compare the cost of a D3AA and one 14500 to the cost of a TS10 with battery included and a couple spare cells on top of that.

My last D3AA was one of the most expensive lights in my 40+ light collection; UV mules are not cheap! I could’ve got 4-5 TS10’s for what I threw at Jackson for that light. I don’t regret it as it’s an excellent light, but there is no denying that price is a factor in many people’s light-buying decisions. A lot of folks would not drop that sort of money on a flashlight. Or even half that for a more normal D3AA like the ones Hank offers. But $20 for a fancy UI and huge power (at startup) makes it a great gateway.

@jon_slider

I have some data comparing a TS10 to a D3AA here

And it’s that sort of thoroughness that makes me respect you despite our differing opinions on surpassing the thermal limitations of a host

Before seeing that, I would’ve thought that the limitations of the Freeman driver at low levels would’ve made the TS10 more competitive at lower levels despite it’s shortcomings past the 350 mA limit of a 7135, and here you are changing my views with empirical evidence. Now I feel better about ditching my TS10 addiction for the D3AA… that I already have half a dozen of. To be fair, I have not bought a single TS10 since the D3AA came out. Thanks to you , I now feel better about spending over twice as much per light.

[jon_slider]

@jerv

Thanks to you , I now feel better about spending over twice as much per light.

LOL, happy to help you spend

I otoh, learned from you, that my TS10 runtime is just about as good as my D3AA, at the brief and low output levels, that I use to look for stuff in my closets, or to walk through the house in the middle of the night.

Where the D3AA really outshines the TS10, is not in the fact they both have regulated output below 100 lumens, but rather the lack of PWM of the D3AA driver… and the very impressive performance of the D3AA on Eneloop.

Then again, the D3AA loses my respect when it preflashes at moonlight and blips crossing above the 10 lumen mark…

I try to love my D3AA, because I paid for it with my own money, but its form factor is just out of balance, compared the Sexy curves on my TS10…

Bearing in mind that my TS10 does not use stock LEDs, and the TS10 mods brought the price up above D3AA levels. While the D3AA comes with my favorite LED options and needs no modding.

But I just dont love sideswitches, not even nicely recessed and illuminated ones like the D3AA has. The TS10 tailswitch is my jam.

First world problems are nice to have. Should I carry a sideswitch or a tailswitch today? The answer consistently, is a TS10 in my pocket.

Plus the TS10 comes in MAO, the D3AA cant touch that…