Hi, I found a bug in the build for D11 B35AM (stock, I haven’t reflash it yet). If I recall correctly, there was similar thing discovered some time ago. I locked the flashlight, configured aux to blinking and white. After couple minutes it unlocked itself.
Just wondering, I just got a Fireflies T9R (SBT90.2), doing the 15-clicks from Off version check, it blinks out:
2020-12-24 model code ‘443’
The driver board indicates “LUME rev B 11-20’.
The model code 443 does not seem to be in the ‘MODELS’ file when I checked now.
Can’t seem to find Anduril 2 firmware with date code 2020-12-24 in ToyKeeper’s repository either.
The triple-click-from-Off sequence is: BattCheck -> TempCheck -> Beacon -> SOS
So this is another flashlight that has ‘SOS’ mode in the BattCheck sequence.
From Hank’s thread a little while back. TK, have you seen this stepped ramp idea? I would love that on my tint ramping D4V2, as I only ever use mid-ramp, one channel, or the other channel.
Also, would it be possible for you to add jump-start functionality to K1 SBT’s firmware?
Thank you for all your hard work. I know you are super busy.
I’ve started with it a week ago, but it’s not finished and I’m not happy with that implementation: GitHub - SammysHP/flashlight-firmware at stepped-tint-ramping
Btw here’s an overview of my modifications to anduril: Modifications Overview · SammysHP/flashlight-firmware Wiki · GitHub
Would it be possible to have a timekeeping feature in Anduril 2? It’d be great to have tint ramping lights automatically adjust CCT based on time of day.
No, there is no real-time clock on the drivers and the MCU’s internal clock is not precise enough to be useful more than a few minutes.
There is a feature that does that… (not by time of day, but by output)
“”there is an “auto CCT” mode.“:http://toykeeper.net/torches/fsm/anduril2/anduril-manual.txt This automatically chooses a CCT based on the brightness level”
(Tint ramping is a misuse of terms, it is actually Color Temperature Blending, so I edited out the word Tint and replaced it with CCT in the above quote from the manual)
Is there an official definition of “tint”? I’ve been using it as a term that means any deviation from pure white, whether it be CCT or DUV.
color temperature (CCT), expressed in Kelvin, ranges within the orange/blue spectrum
Tint, expressed as Duv, ranges within the green/magenta spectrum
.
”Tint” is not used in lighting standards, it’s a colloquial term, there is no official definition, just the one people use.
In standards they speak of light chromacity (color of the light), expressed by x,y coordinates in CIE 1931 color space for example, but also CCT and duv for white light because it’s more intuitive. ”above” ”on” and ”below the Planckian locus” are terms often used when talking about duv.
In my opinion tint in general refer more to CCT, as in ”I like warm tints”, while talking about tint for a certain CCT means duv, as in ”this 3500K LED has a good tint”.
Also AFAIK there is no mention of ”pure white”, maybe that’s a thing in the printing or other industries, but not in lighting.
Any idea how much it would vary over time?
I’ve been interested in a sunrise mode, where you tap the number of hours and tens of minutes that you want to sleep, then it slowly ramps up in the morning.
Each flashlight would require calibrating on first use (eg teach it how long 60 seconds is). I would have thought it would be accurate once calibrated to at least 5% (30 minutes overnight) but I haven’t tried it yet.
I haven’t measured it, but it depends on temperature, voltage and maybe some random factors. I guess you could achieve ±1 h/d after a proper calibration (factory is ±10, after user calibration up to ±1 with fixed voltage and temperature per datasheet).
Voltage is constant in the (not so) new Noctigon drivers, temperature isn’t though 
I don’t think there is one, singular definition of tint for lighting, but most frequently when used in the flashlight community, we’re referring to DUV. I’ve seen it used inconsistently by lighting manufacturers, sometimes referring to DUV, but probably just as often referring to any sense of non-white.
CCT of light isn’t really treated as a deviation from white, per se, as there is no color temperature defined as pure white. This is significant since our color perception has a relative component. 5700K may look pure white in one setting, while 3000K may in another. Of course, go much lower than 3000K, and there is almost no situation where the light looks pure white, so there are limits to this viewpoint.
Color theory can get really complicated, and numerous different color models have come into use over the last 100+ years as a result.
Rather than get too deeply into all of that, I prefer to just continue as we have been, informally using the CIE chromaticity diagram, and describing color temperature as one variable, and DUV or “tint” as another.
What’s the difference between the different curves (cube, fifth, 5.01, etc) in level_calc.py, and how are they chosen for the different flashlight configs? I’m thinking of adjusting the pwm’s for a 9a d4v2 and a kr1 with sbt90 to make the ramping visually linear.
If I were to order a custom Emisar D4V2 with SST-20s, an illuminated switch and the KR4 5A driver ... which Anduril 2 firmware would I use for that?
Does the KR4 fimware automatically support the switch LEDs?
If you want only 5A: kr4-nofet.hex
Full power with FET: kr4.hex
Yes, it does.
Great, thanks!
@SammysHP
Quick follow-up question, I just learned about the 7.5A+FET and 9A+FET linear drivers, do they also work with the normal kr4.hex firmware?
Yes, firmware is the same.