I think we’re on the same page here Jason, but I have the tendency to express a bit differently. I’d say with both spring stock and bypassed, the driver works equally hard. But with a spring bypass you have a tad more voltage available for the driver, resulting in more turbo (max output with intermittent cooling) sequences with depleting cell voltage before reducing in output.
Tailcap spring bypassed even the stock Acebeam 21700 battery manages five full turbo cycles. Did only one before.
Maybe I should change “harder” to “less efficiently”.
This is actually something I’d like to explore in more detail. If a boost driver produces the same output at 3.5v and at 4v, is it working more efficiently or harder?
When I tested the Lumintop ODF30 I tested Turbo at three different battery voltage levels (Liitokala 5000mah 26650 black).
4.2v = 8.35A
4.0v = 9A
3.6v = 10A for one second then drop down to next level.
When testing a protected KeepPower 5200mah at 4.2v it drew 10.6A for a second then dropped down to next level.
I found all this pretty interesting, but I’d still like to learn a bit more in depth as to what’s going on especially in relation to heat build up.
Hopefully I’m not going too far off topic.
Elementary, my dear Watson. Lol
Thanks for trying it out Maukka.
Thanks for sharing the results with the 30T. 
I think it’s important to understand what is going on. With a lower input voltage you need more current for the same output. More current means more heat for the same output, therefore lower efficiency. But the intrinsic efficiency of the driver could be just a tad higher or lower at that current level, so you need to superimpose the efficiencies to get the total efficiency.
To state that the driver is working harder or less efficiently is merely a definition issue I think. And since there probably is not a well defined definition, miscommunication can occur easily.
For those curious about how the different batteries can have different amounts of voltage sag, here is a comparison of the Samsung 30T to the Samsung 48G (I’m not sure what 5100mah cell Acebeam is using, but it seems similar to the 48G)
You can immediately see why the “high drain” 30T allows for more turbo run time due to its minimal voltage sag compared to a “high capacity” type battery.
as usually - awesome review.
the only thing i can add to this are pics of how this light performs, for the rest - you did a great, great job.
thnx
It is sadly very difficult to make beamshots during the summer here. There’s only a couple hours of darkness during the night. The need for flashlights severely decreases for a while
I'm a little late to the the party, but this review is so good, I had to give you a shout out. Thanks, maukka. Love all the technical data, especially runtime charts and color rendition info.
I am impressed by the flat output levels once this flashlight steps down to about 800 lumens. Predictable, flat output is the main reason I prefer boost (and boost/buck) drivers over linear and buck.
The high-CRI Nichia version must be fun!
Buck drivers are actually great, and are actually better if you have a higher voltage source, as that means the contacts don’t need to be as thick to carry as much amperage than at lower voltages.
You just need 2x 21700 instead.
It looks like we will have some aftermarket boost drivers to play with soon. 3A@12v and 6A@6v. :+1: 
Yeah, I get that. And I know buck drivers can be more efficient. They don’t fight the battery as the voltage wanes.
Nevertheless, I like the perfect staircase-like runtime charts produced by buck/boost drivers such as the ones ZebraLight used in its older lights (old, ha, ha, like 2 years ago). If a mode activates at all, then I know I’m getting its rated lumen output. (That is, so long as thermal regulation hasn’t kicked in.)
I think the new 2018 EU headlight regulations for Class-3 e-bikes require this kind of flat output. Slow fading is illegal. Obviously, that’s good for rider safety in bicycle lights, but I like it in handhelds, too.
Just my bias.
Huh? Slow fading?
Why would that be illegal?
I’m guessing there is a legal minimum brightness.
In California, they are called Class 3 E-bikes. In the EU, "speed-pedelecs." These fun e-bikes have motors that provide "assistance" up to 45 km/hr (about 28 mph). The new regulations are designed to treat them more like scooters/motor bikes, so, yeah, light output must be consistent. No gradual dimming allowed.
Regular "pedelecs," which top out at 25 km/hr in the EU, are governed by less stringent regulations. In most jurisdictions, they are treated the same as regular bicycles.
I don't know much about these rules, except what I have read in passing. As I write, it occurs to me that I may be wrong in attributing them to the EU. They may be limited to Germany.
Great review OP . One of the most thorough I have read , most of it over my head but that is good to learn from.
Took some beamshots for the Fireflies ROT66 review and included the EC65. OP updated with comparisons to other TIR optic lights.
Very pity they consider their customers a sh!t.
I contacted them via their offsite email and asked about difference. How to spot the new batch from the first batch, which had problem of using high drain batts. Asked some other questions about charging etc. They answered that the light goes with their acebeam battery in package. Said not a word beside that.
Now, a rhytorical question.
Why to bother having deal with such morons.
They do not even respect the time of person who is contacting them.
Amazing review Maukka. Thank you for all the hard work you put into it… 