Both of my KR4s exhibit the same behaviour, it isn’t a fault. It is a limitation of the driver type which the vastly more knowledgeable Toykeeper has better explained before. By default the bottom of the ramp (smooth, not stepped) is set to 3 due to the driver.
I set mine to 2 around the house and back to the default level 3 on night time dog walks.
Refer to post #1895.
Thanks Jasetaylor for your quick reply, and for the link to the other thread. I will follow up on this topic in this specific thread.
Yes I indeed noted that this behaviour shows up only if the Floor level is set to 1/150 or 2/150.
I'll try to set it to 3/150, and see if it is low enough when I wake up at night.
The MCU on that driver can’t run Anduril. It uses an attiny13a, but Anduril requires attiny85 or attiny1634.
If you swap the MCU though, it might work. IIRC, the t85 can fit if its legs are bent underneath the chip.
Second measurement:
Does this mean Anduril measures the voltage under charge?
The good news is the conservative approach of the Anduril battery check, 0.2V less is safer than 0.2V more, WRT the actual voltage.
The issue with moon turning off is a side effect of the constant current power circuit. Instead of using a digital signal to turn on and off quickly, it uses an analog control voltage. And analog circuits are typically pretty sensitive to environmental factors, so it can behave differently depending on things like temperature and which level it was at a moment earlier.
Steps at the bottom end of the ramp don’t have great spacing on this driver, because the output resolution isn’t very high. It could work around that by repeating each ramp level quite a few times in the internal 150-step table, which would help the stepped ramp spacing… but then the smooth ramp wouldn’t look as nice. So there are tradeoffs. I’m hoping a future version of the hardware will add more resolution at the bottom end.
About voltage calibration, that was added in Anduril 2. It’s not in Anduril 1 though, so you’d have to reflash the firmware to get that function.
If it helps, the way I set up my D4v2 with E21A is:
This gives me consistent and predictable output every time I pick up the light, defaulting to my most-used level. But it also gives me the flexibility of the smooth ramp, including moon. So it has some benefits of both ramp styles.
As for reaching moon from off without accidentally going through a bright level, what I find works is: Aim the light so I can see the aux LEDs. Hold the button. As soon as the aux LEDs turn off, let go of the button and turn the light toward what I actually wanted to see.
On most other lights, it’s easy to time the button release by watching for the main LEDs to turn on… but since this driver responds slower, I watch instead for the aux LEDs to turn off.
It mostly just means the attiny chip and other components don’t have very precise calibration from the factory. So some lights read high while others read low. That’s why I eventually added a calibration function.
While the light is running, it reads under whatever load it’s currently using. In battcheck mode, it stays mostly off, so the load is pretty low. However, it uses a rather strong lowpass filter, and the battery itself may take some time to recover after being used at a high level, so battcheck can take a couple readings to converge on a stable value.
Many thanks ToyKeeper for your detailed reply.
I’ve done something similar, which allows me instant access to #1 my most-used level, #2 Moon, #3 High. No other UI allows this AFAIK, because they all allow only automatic memory (which I hate, I don’t want my most-used level superseded by a Moon or a High because this is what I last used).
But I chose to set Floor to 3/150, hence avoiding the LED extinction after going down from High to Moon.
A bit risky if you don't react quickly enough, and your next level after Moon is a bright one (e.g. few steps in stepped ramping mode). I think I'll get used to make a very short "Hold", hence going to Floor and staying at this level before the ramping up starts.
A nice addition to Anduril 2 (maybe it is already planned?) would be a third menu option for the Ramp Config Mode / smooth ramping: ramp speed adjustment. N clicks ==> N seconds between Floor and Ceiling when ramping up.
One more reason for me to plan a future upgrade. I've entered the Anduril world only a few days ago, and I already feel addicted, this UI is so great and configurable that I don't feel anymore like buying lights with a different UI (apart from tiny AAA or 10180 lights). I regret I haven't ordered the reflashing kit from Hank, together with the light. BTW, which lights (apart from Emisar and Noctigon) have a board compatible with this reflashing kit?
Here are the results of a few more measurements:
The latter value is surprising: I remember doing a battery check immediately when I received my D4v2 with a full battery, and Anduril measured 4.1V.
You are right, for some reason I thought that it use the same MCU…
Anyway I have A17-DD driver with attiny25, so at least swapping it for 85 should be easy 
I don’t know if this is right topic for my questions, but it seems like, so here they are.
1) I got D4V2 with KR4 CC 5A driver (LED - Nichia 219c) and have noticed that it seems like main channel and FET are active together when FET is active (config file - cfg-noctigon-kr4-219.h) - why it is like that? I may be wrong, but shouldn't it be like FET only?
// don't turn off first channel at turbo level
#undef PWM1_LEVELS
#define PWM1_LEVELS 0, ... ,1023
2) I have compared config files for kr4 (cfg-noctigon-kr4.h) and original driver d4v2 (cfg-emisar-d4sv2.h) and found those difference, which makes me question why it's like this (exept differences in configs that I can undestand, like AUX LED in button, soft reset with button and so on - wich root cause are hosts possibilities)
D4V2:
// stop panicking at ~30% power or ~1200 lm
#define THERM_FASTER_LEVEL 105
KR4:
// stop panicking at ~25% power or ~1000 lm
#define THERM_FASTER_LEVEL 100
#define MIN_THERM_STEPDOWN DEFAULT_LEVEL
#define THERM_NEXT_WARNING_THRESHOLD 16 // accumulate less error before adjusting
#define THERM_RESPONSE_MAGNITUDE 128 // bigger adjustments
So, I see different thermal settings and the question is - this difference due to bigger kr4 thermal mass, or the driver itself? In other words, should I change them for my light like in original d4 config (remove them, so ones from fsm will be taken)?
Why 51 for 100/150?
Because 1 is 150.
For 1), this is a totally different driver than a FET+1 - the main LED PWM is not PWM'ing output directly, like it does on 7135's. The very last level has both turned ON probably for a good reason that TK would know best.
For 2), I assume she custom tweaked the settings for the KR4, as she did for the Emisars. Driver and light are both important
If it was me, my best guess would be for your setup of a KR4 driver in a D4V2, I'd go with the KR4 settings.
Thank you for your answer!
For 1 - I tried it myself (disable 1st channel when turbo on and get very weak turbo - so that’s the reason
For 2 - I’ll tried both config entryes and see no difference, so I continue with KR4 config
Hi, i have not found another place for my question. If it’s wrong place i am so sorry, please move. So…
(Fireflies E07)
I measured my room temperature with Fluke… 21C°
I want to calibrate temperature so 21 clicks in current temperature mode.
I left the light overnight.
Next day i want to measure temperature and this is happen…
The first cycle of blinks are 21C° , the second cycle are 16… between the first cycle and second i do not shut off the light… the room temperature is 21
(FW3A)
Now I have set the temperature to 18 (room is 18C°)
Shut off the light…3 clicks than temperature mod… 21 blinks 
Why?
is correct?
The thermal sensor must give me the room temperature or the light temperature?
I don’t understand so ask to community
Thank you
Before you set room temp in the flashlight, you must do your best to get the flashlight at room temp - that means leaving it sit off for maybe an hour or 2 before setting the temp, and even when you do be sure the light never accidentally goes to a hi mode. What I do is to be sure to press&hold to turn it on at the lowest, ramp up very slightly - turn it off. This assures a lo output when doing the 3 clicks to get into voltage display, then 2C's from there.
Also the first blink of temp is not accurate - thought TK fixed this in a more recent Anduril, maybe just in Anduril2? Not sure. Let it blink out temp 2 times to get a good reading.
I leave overnight… maybe 8H
Why?
Not mine I think
The sequence you described was leaving it over night "after" it was calibrated - you need to stabilize the flashlight temp before calibrated.
Yes, because I have already calibrate one time before i write the first post
I am so sorry for my English… I will try another time and upload thread.
When I was calibrating my SC31pro, just holding it in my hand had the temp increase (on blinkout) 1° and then 2° in just a minute or two, so you need to do it fast. And keep the light overnight right next to a thermometer so you can see at a glance the temperature and let both equilibrate.
Next day or however long, go immediately to the temp blinkout and verify what it tells you vs the thermometer temp. If too different, then again, immediately do the calibration, and another temp blinkout to verify that it stuck.
21°C, holding it in a 37°C hand, it’ll warm up a few degrees very quickly.
Ahh, ok. I forgot the details but I think the first reading was perhaps a bug.
As LB said, it's very sensitive. All these calibration issues go away with the Atmel 1-series (1616, 3216) since they are factory calibrated.
Hi! Could you tell me, https://bazaar.launchpad.net/~toykeeper/flashlight-firmware/fsm/view/head:/ToyKeeper/spaghetti-monster/anduril/cfg-ff-rot66g2.h applied for any ROT66 II on factory (with any Leds?)
And please, could you tell me config for EDC18 and Astrolux HL01
Thank you!