I just tested the pressure required to activate the switch in some of my lights using a digital postal scale:
KR1AA V2 bronze (unmodified): 460g
Wurkkos TS10 (unmodified): 750g
FW3A (using early production run higher pressure popper, plus o-ring mod to switch): 1390g
KR1AA V1 black (modified with 9x1.5mm o-ring added to switch): 1590g
KR1AA V1 blue (modified with 9x1.5mm o-ring added to switch): 1690g
I like the pressure required for my modified KR1AA v1s and modified FW3A the most. All 3 require a lot of pressure and give the loudest clicks when activated.
The higher pressure of the last 3 lights on the list is nice, because I donāt have to worry about them accidentally activating inside my pocket or when pushing the light down over the edge of a pants pocket or sweats when clipping to my clothing. I feel confident carrying those three lights without using lockout as the pressure required is enough to prevent accidental activation.
In contrast, I experienced 2 accidental activations when trying to clip my KR1AA V2 to my pocket on just the first day. The activation pressure isnāt enough for my tastes.
Note: All 5 lights I tested have fully reliable switches with no failed activations (where there is a click, but the light fails to activate).
Not sure whatās going on. True that itās too low as Toykeeper noted, but this (the underlined part) does not seem to be the case with my 2 Emisars with Lume X1:
DA1K w/ NTG50 4200k
D4K w/ 519a 4500k.
Am I testing this correctly? 1 click and hold and quickly release right? And if click and hold again it comes out of moonlight? Total: 7 lower levels + Turbo = 8 .
Noon now in LA, indoor w/ indirect natural light through windows, I could see very clearly LEDs are on at an angle 2 meters away.
Moonlight is not set to the lowest level after a factory reset, itās something like level 10/150.
I set mine to 7 IIRC, thatās about the ZebraLight 0.01L low.
My KR1AA has now failed completely. The switch had a few misfires and late activations but now the light shows blue aux upon tightening then the red it was set to. If I repeatedly press the button itāll sometimes fire up or jump into turbo (reading a double click). A misfiring switch on an Anduril light isnt great.
Hank needs to be given his due here, I said Iād have a replacement if he could find a TIR that didnāt have such a large chunk out of it as mine has, but he got back to me and said they all have a mark where the moulding has been broken off the stalk. Heās going to refund, and I may or may not add a tailcap to my next Intl-Outdoor order to see if it solves the problem.
When the teething problems are ironed out, or he reverts to the V1 design and gets those reflectors nice and round, then I think itāll be one of the nicest little EDCs ever.
If youāre using the default config and youāre in stepped ramping mode, or in simple mode, the bottom step is level 10 of 150. Try switching to the smooth ramp in advanced mode to go lower.
The ramp has 150 levels. The driver hardware has 3 āgearsā, like the low/med/high gears on a bicycle or car, for different power ranges. The ramp is divided among gears like thisā¦
1 to 5: low gear
6 to 50: medium gear
51 to 150 high gear
So if youāre at level 10/150, the bottom default step in stepped ramp mode, youāre not even in low gear.
Internally, those 150 ramp levels control the power for each gear. If I recall correctly, the top of low āgearā is about 0.01 lm or soā¦ and it has 2500 different internal power values per gear. The bottom 5 ramp levels use these internal power values:
1/150 = 100 / 2500 * low gear
2/150 = 359 / 2500 * low gear
3/150 = 790 / 2500 * low gear
4/150 = 1436 / 2500 * low gear
5/150 = 2500 / 2500 * low gear
So the lowest ramp level is about 1/25th of the 0.01 lm brightness the low gear can produce. In a perfect circuit, that would make it about 0.0004 lm. And if you reflash it with a different ramp table, the driver can still go 100X lower than thatā¦ like 0.000004 lm. Thatās four millionths of a lumen.
Thatās for a perfect circuit though. In practice, individual torches have different brightness levels at the bottom of their range, because the components involved simply arenāt precise enough to be accurate to within millionths of a lumen.
So on one light I tried, the lowest ramp level was like 0.0001 lmā¦ and on another it was more like 0.005 lm.
Either way, itās really, really low. And the driver is capable of going even lower (by two orders of magnitude)ā¦ but I didnāt enable anything lower by default because it seemed a bit pointless. I like low modesā¦ but damn. Loneoceans really overengineered this thing.
I donāt know if I asked this before, but is it possible that the low gears were intended to work with the candle mode for even higher resolution āflickeringā, to make it flicker more realistically?
**particularly in a triple, potentially having each LED emulate a different part of the flameā¦although independent channels might be necessary
The current state of the art in flashlight drivers and firmware is really awesome.
I definitely focus on the āreally nice that it CAN go this lowā perspective rather than the lack of apparent need that might make it seem ācompletely uselessā perspective. A flashlight is first of all a tool, and itās not a problem if a tool is even better than what is needed in some metric, unless that results in a compromise in another metric.
No one complains about a wrench being uselessly strong, or a microscope having uselessly fine resolution. If there were a complaint, the real reasons would be considerations like the wrench being too heavy, or the microscope being too expensive. But Iām not aware of significant drawbacks to how low the Lume X1 or TheFreemanās D3AA driver can go. Lower than was practical with a single sense resistor was something many users definitely wanted, justifying the original changes. My understanding is adding the extra resistors more or less inherently opened up a range beyond the known needs. Admittedly, drawbacks on the user side are a different question than on the developer side, where there was some extra work.
So my thoughts on a firefly mode that goes lower than is needed? Great! Then I can be confident I wonāt ever be stuck with more light than I want in some situation.
I donāt yet have one of these lights, although certainly will have to get one, and I will be looking forward to see just where the limit of what I find too dim to be useful really lies. I do know that when Iām camping in the wilderness and get up in the middle of a moonless night, my current headlampās 0.1 lumen firefly mode is brighter than I want to use in the tent. Itās really startling how bright light levels we barely even perceive in other situations can seem with completely dark-adjusted eyes.
In those situations, I suspect even the aux lights on low might also be brighter than I want, and the beam pattern is usually very busy. I also am inclined to stick with white unless there is a specific reason not to do so. While some here are likely aware that color perception is reduced at very low light levels, it is still necessary to have a spectrum that overlaps the reflection spectrum of objects being illuminated to see them.
Plus, Anduril gives us much more control over the main lights than the aux.
Perhaps an oddly-specific question about the dual channel / tint-ramping drivers:
Is the rated output current of the different versions determined by the hardware (eg, different sense resistor), which would mean that the lower current versions would also have a lower moonlight; or by the firmware, which would mean only the maximum output is limited, but the moonlight is the same between them?
Alternately, does anyone happen to have minimum output measurements for the E21A version compared to one of the higher rated dual channel versions?
As reference material for others, since I had difficulty finding this information again, the different channel driver outputs are:
Currently Iām leaning toward E21A for a roughly 2000K + 4500K dual channel light that stays closer to neutral tint, but havenāt fully decided. If that also gives me lower moonlight, thatās probably the direction Iāll go. Iām not really concerned about max output for this light.
For the customers with KR1AA switch or other issues, please contact us, we will not only send a new switch, but a complete working light with new switch.
Just please give us 2 weeks time since we have run out of stock some parts, Iām sure we will find solutions for all the addressed issues, and we will keep improving.
Excellent, Hank!
I know you said the clip is designed to rotate, but honestly that makes the flashlight feel like it has a cheap finish. I would much prefer it to stay firmly in place when tightening the tailcap.
A sturdier, more robust clip would also be very welcome.
Yes, this issue has also been solved by us recently. The KR1AA tailcap has been modified a bit, so that it will secure the clip much more firmly, it will also move if you push it hard, but it would not move freely.
For those customers who are bothered with the rotating clip issue, please also contact us, we will send an updated / complete working light as well.
Copper spacer in the tail might be the way to go to reduce slop? Similar to button top spacers. Unless the spring can be stretched enough to snug up 26650?
The lower current items should have lower moon too.
BUT
That is not the main factor which determines the moon brightness on these. Itāll emit like 1/16384th of the maximum power, in theory, meaning if turbo is half as bright, moon should be half as bright tooā¦
ā¦ but in practice it depends mostly on the LED Vf, and battery level, and how hot the regulator is at that moment, and sheer random variation.
At the lowest modes, it does not have sufficient precision to be able to guarantee any particular lumen level. It tends to vary even from moment to moment, depending on environmental factors.
This is a stark contrast to thefreemanās high-dynamic-range drivers or loneoceansā ultra-dynamic-range driversā¦ which are both very precise at low levels.
Thereās nothing wrong with the 2-channel drivers though. Itās just a consequence of having 2 channels. Because on thefreemanās drivers, for exampleā¦ it uses 2 channels for one set of LEDs ā one for a high āgearā and one for a low āgearā, so the low modes are precise. But on a 2-channel driver, it uses one for high āgearā on LED 1, and the other for high āgearā on LED 2. So there is no low āgearā, and the low modes are like putting a car into high gear and trying to drive very very slowly.
