Here is the link to the 4K version I have. Notice the 95 cri rated leds 
After taking a few hours this afternoon I finally took my tutorials (thanks again Funtastic, your links were SUPER helpful). I reset the temp sensor which several users have mentioned is coming from the factory way off. I didn’t wait for mine to count it out though, I just reset it along with the thermal limit of 60 Celsius. I am wondering if I should have gone for 55, but we’ll see.
One thing that seemed to not work though was the upper limit for smooth ramping. I’m not 100% sure, but it seems that it was coming on at full turbo even though I chose the TWO (2) click limit. I also wonder if double clicking (for turbo) in smooth ramping ONLY goes up to the “set” limit. After I tried twice to set the limit a tad lower, it didn’t seem to work AND when I double clicked for turbo- it didn’t increase in brightness.
Because this light heats up so fast, I’m looking for the brightest setting it will run at (in smooth ramping) without getting ridiculously hot. I’m estimating that to be about 75-80% of it’s full turbo output after four times out with the light. The idea is to get smooth ramping set for a good every day use brightness being the top limit… BUT still be able to double click UP for turbo as needed.
I guess I have more work to do but one last thought is that although this color is REALLY nice (at 4k) when running it towards the top of it’s output capability, I have a harder time seeing even close in things with it at lower levels. I’ll have a 6.5k IF25a here in a month or so and really be able to compare the two. Several times I’ve though these four LEDs in a slightly larger head (a little more mass and deeper heat sinks) would be killer. The C8F would be too heavy for what I am imagining, but something in between would let you run these LEDs full out without the instant heat. But hey… maybe there are five lights out there already that fit the bill and it’s an afterthought on a HOT July night. Come winter, I’ll probably be GLAD it’s a little hand heater 
After further efforts, I realized the range of clicks is well interpolated. I had to click DOWN 40 clicks to get my ramp MAX to a level where it’s plenty hot still, but I can hold it in the bare hand. And YES… turbo is working now when I double click from ramping!
I am VERY impressed by the light’s output on turbo. But it’s not a realistic option for more than about ONE MINUTE without a glove. Even where I have it running now at about 65-70% of turbo (estimated, and this is 40 clicks DOWN from the max ceiling setting), I hope it’s bright enough for outdoor woods out to 50 yards or so.
I would love MORE heatsinks and would give up a few ounces and a slightly longer (or wider) head to tame the heat just a tad bit. I’m still sure it will make a great winter light for hunting and such as it is, but in July in Kansas it’s really too hot to use much above 60% output due to the heat in the hand factor.
Anyway… Anduril and this IF25a ARE 100% working as expected, I just didn’t expect the heat which is manageable, but a factor if you aren’t comfortable with holding very warm things. 
I received my 4000K version today. The box didn’t contain a manual. Mine shipped May 19 from sofirnlight so it was probably one of the first batch or so. I set ramp floor to 10/150 and ramp ceiling to 100/150, and it still gets warm fast at ramp ceiling but not hot. After running the flashlight under cold water the temp sensor read 36C, so I recalibrated it to 25C. I don’t have any charged batteries that can really drive this thing on Turbo, so I haven’t tried running it really hot yet. The light won’t work with flat top laptop cells but works fine with Sofirn button top cells, due to the flat contact in the head.
The head has good machining but the body and tailcap have a lot of sharp edges. I found the body tube was impossible to remove from the head, anyone else’s light like that?
Tint at similar brightness is generally more yellow than my SC31B with 25deg TIR, which uses the same emitter. At higher brightness though, the tint approaches and surpasses the SC31B in rosiness. I’ve heard that the SST-20 gets less green at higher currents. As far as I know the SC31B uses PWM+FET so current stays pretty constant throughout its modes, so the IF25A uses something different. I decided to measure output at different voltages with my phone, so big margins of error:
3000mAh Sofirn cell @ 4.1V:
Some ramping level: 85 lux
Turbo: 1056 lux
2200mAh Sofirn cell @ 3.4V:
Same ramping level: 90 lux
Turbo: 315 lux
As my first Anduril light the UI takes some getting used to, but basic operation isn’t too complicated. I wish there was a way to access moon discretely from ramp floor, and I wish single click from turbo went into the previous state instead of off, but otherwise no complaints.
Uhm… I have used flat-tops in mine fine. The front contact IS wide, but I thought it a design choice to make the light smaller (shorter). Both a Sofirn brand 4000mAh and Molicel 4200mAh flat tops fit and run fine in my IF25a.
Also, given this light runs so hot, and running it on turbo for long isn’t a proposition 99% of the time… I now wonder if a higher capacity battery might be a better fit. The Molicel is a VERY high drain battery and it does a great job, BUT the Sofirn 4000mAh 21700 “heats it up” just as fast 
The difference in output on turbo probably isn’t 5% or so (if that… I want to “feel like” its brighter because the Molicel is a 35A drain battery!).
Anywho… the point is that without a glove, I’m not running this light high enough to benefit from a high drain battery like the Molicel. The Sofirn 4000mAh battery runs it just as good with the (possible) exception of a touch more output on turbo with the Molicel. But after that 30 second show… maybe a 5000mAh medium drain battery like the Samsung 50e (5000mAh capacity with 10A drain) 21700 would be a better fit as it wouldn’t heat up as fast and would increase run time by about 20% or so.
On that note; I HAVE a Sofirn 4800mAh 21700 I’ll probably try next. I doubt it would heat up any faster (passing it’s thermal/current rating) than the light itself. So that battery paired with the IF25a MIGHT just make a great “reciprocally regulating” pairing together 
Playing with this IF25a WAY too much lately
I can’t get over how super cool Anduril is. I’ve been SLOWLY memorizing the IU the last three days and have now pulled out the LT1 and SF36 I didn’t get around to messing with much before.
After setting the temp sensor yesterday I tried an experiment. But first a few things to explain before the experiment:
I am running smooth ramping with a ceiling of 40 clicks down from the high of 150 possible levels. So the ramp starts at moon and goes to a set ceiling of 110/150 * which is the same thing as saying 150 (max/turbo) LESS my 40 clicks = 110*. So this is often shown in a ratio of “110/150” which I didn’t figure out the first time I read it- had to get into Anduril first— THEN it all “clicked” for me. Looking at this ratio, one might say the light is on at 73% output… and to my eye that would be about right.
I ran the light at the top of the set smooth ramp set ceiling for 5 minutes today. I then immediately turned it off, and went into the thermal check function. The light “blinked” out 5 flashes, then 3 flashes = so it reported a temp of 53 Celsius about 10 seconds after I turned it off. THEN… I did NOTHING. I let the temp check keep running. After about 15 minutes the light got to room temp and was reporting 28 Celsius. THEN I took it and placed it on my AC vent for about 3 minutes until it read 21 Celsius. Finally, after another 15 minutes on my desk (where I am NOT getting any work done today) it steadied to reporting 24 Celsius (75 degrees) WHICH IS EXACTLY the temperature in the room!
So, despite the factory temp sensor being off from the factory, MY IF25a (after calibration) is VERY accurately reporting the ambient temp in my house (within 1 degree accurate).
So in summary: I have a 60 Celsius thermal limit set. My light running at 73% of Turbo (at 110/150) hit 53 degrees after 5 minutes running (in the same room tested). And the final “resting” temperature is “dead balls on” accurate!
So I finally got the 4000mAh battery charged, and tried running turbo for about 3 minutes. When I went to turn it off, the flashlight was far too hot to hold, and I just barely was able to click to temp check. It read 67C! I swear I set the temperature limit to 55C with 25 clicks in temp config, do you know why it’s getting this hot?
Because it’s a little spitfire light
Did you set the temperature sensor up yet? You can try resetting the thermal limit in that mode too while you are at it. Probably did, but a check box for sure.
Then in the ramping setup, I clicked down 40 clicks to finally get a bare hand hold temperature of 54 Celsius. I think I could have gone a BIT less on the upper limit attenuation (maybe 36 or 37 clicks) to settle into a high 50s Celsius ceiling. But 30 was still TOO HOT for me!
I am beginning the think this particular 4K model might end up being an indoor light except in winter when the heat will WORK when it’s 10 degrees out 
I have a 6500k version on the way and given it’s summer now, that model may be better for outdoor use right now. It seems that outdoors a few times now… I can’t QUITE get the brightness out of this 4K IF25a that I am wanting- without it getting JUST a little too hot in the hand today.
Hi guys, does anyone know which optics are installed in the Sofirn IF25A, and where can buy narrower optics for it ?
I had set it to 55C. Since then I lowered it to 50C and it’s more manageable.
I don’t know, but there is a teardown of the IF25 here on fonarevka: Показать сообщение отдельно - Sofirn IF25
Says the TIR is 23.4mm wide and 7.8mm tall.
Note that this is the IF25 not IF25A so electronics and leds will be different, but I think optics are the same.
Thank you, I know about this topic, but unfortunately there is no mention of who is the manufacturer of optics or where you can buy an alternative.
Decided to try my Sofirn 4800mAh battery out in my IF25a last night. Went out on a charge of about 4.15V which is about where the lights stops charging.
With the heat factor in mind, I figured running a lower drain, higher capacity battery would be a better way to run this light right now in summer. I suspect there is some voltage drop and all occurring. But I wasn’t going for turbo, just wanted to run it with my ramping set at about 75% at the top of the ramp where I have been using it since setting my initial ceiling there.
It ran great! When I double clicked up to turbo a few times, it still reached a great brightness and started heating up just like I’d does on the Molicel. But after the 10 minute foray, the IF25a was warm, but not as hot as it usually gets. I ran only 20 second bursts in turbo, maybe three times. So the light didn’t get quite as hot with the 4800mAh Sofirn battery given the same test as I’d run before.
I’m going to guess the Sofirn 4800 mAh battery is similar to the LG M50 which is a 7.3A lower drain battery with about the same capacity. And as batteries are rated at a certain current for a given length of time, they often can/do produce more current for a short time before hitting the wall and going over 60 Celsius.
As I won’t be using this light in turbo for more than a few minutes, I figured the battery is getting me ALMOST as much output for that short time I’m running it at turbo levels (I can’t tell the difference). All the while otherwise, it’s giving me the same output at my ramp ceiling where it is set to.
So the idea is that the battery and light are paired nicely given they BOTH are not running at full speed 99% of the time. I figure I can live with the slight loss in output on turbo (which I didn’t perceive as the light threw to the same distance down my fenceline as it always has when I did go to turbo).
The trade-off is a cooler running light that also has another 800 mAhs of run time
A few weeks later…
Tonight I went out for THREE hours hunting squash bugs. I have about 80 ft of cucumber rows in the garden and for several weeks I’ve spent the nights out there picking bugs (I am a super-organic gardener, so can’t use ANY pesticides). I had Sofirn’s 4800mAh battery at about 4.1V loaded when I started.
I kept it right about where I’m guessing I was hitting approx 45 Celsius. Or in other words, about 60% of it’s output capability. The light ran great without stepping down at all for a few hours. At about 2 1/2 hours it started stepping down a tad, but kept on running and I started double clicking it to push it back to turbo as I was wanting to push it to it’s limit. I checked voltage when it started stepping down more than one step and I was probably getting about 30% output at that point. That was at 2.9V.
I then put it in a low output setting- not moonlight, but probably 10-15 lumens. I left it on for over an hour before it finally shut down at 2.7V.
I have to say… this light WILL get you home if you need it too. It starts getting “frugal” with it’s output management around 3V and if I was packing it out on a late night, I’d appreciate it’s ability to manage it’s output on a LOW battery- conserving energy with decent output to light a trail on the way home.
I also did push it into turbo (about 20 times) below 3V and it put out a LOT of light for 10-15 seconds before stepping down slowly over a few minutes. It never blinked, it never shut down (until 2.7V). And after resting it for 10 minutes it popped back up to maybe 200 lumens for 30 seconds before stepping down and finally shutting itself off in about a minute. It did this several times BTW.
Yes… I am torturing the battery here and I know it. One thing I expected was the switch light to go red. It is now done and won’t turn on and sitting on my desk as I write this. The switch light is still green… but barely I just turned it on again and it gets bright for 10 seconds then steps down and turns off in about 45 seconds now. I’d guess it would stay on moon for a while if I really wanted to push it The Fluke says the battery is at 2.73V as I just put it on charge.
Anyway… a testament to a great UI here with Anduril managing things very well throughout this test. And the IF25a did hold great bright output for hours while I “worked the bugs” this evening. But I was surprised and particularly impressed with the low voltage management giving usable light when one MIGHT need it for sure.
A really SUPERB UI TK 
I’m tempted to get this light. Not a whole ton of reviews on it. The lens is listed as “TIR”. It doesn’t look TIR to me, at least not in the sense of the TIR lens in headlamps like ArmyTek. I don’t have any thrower flashlights, all of my lights are very floody and I thought it would be good to have at least one somewhat throwy light. I like the size and 4000k tint option on this though.
How’s the moonlight level? Is it sub-lumen?
Can it take a pair of CR123A’s?
IF25A runs Anduril, the lowest brightness can be configured. I would think sub-lumen is possible.
By default, the lowest brightness is probably around 1 lumen (just judging by observation — I can definitely look at the light without it being “glaring” to my eyes).
The IF25A doesn’t take CR123A’s…
I’m also not too sure how “TIR” would be defined. There are the “honeycomb” / diffused TIR optic, and there are TIR optic that captures light and shapes them in a different shape (compared to say an “reflector”-type beam).
My IF25A (4000K) came 2 days ago and so far I love it 
It was an unpleasant surprise for me that the IF-25A does not have a battery status indication in the button, it just glows green until the flashlight is completely discharged. Despite the fact that when charging, a two-color indication is used. Why not do the same for discharge? I think this is a very big drawback of a modern flashlight.
Thanks for noticing that and pointing it out. It’s still so new that the typical detailed reviews aren’t out yet. The Sofirn site’s info on their product is very sparse. Or, I guess we just expect to know a lot about a flashlight.
The IF25A runs Anduril. The side button works similar to an “aux-LED” (on other Anduril lights) and can be programmed to 4 states: Off - Low - High - Blinking (Breathing), however, it’s not linked to battery voltage… Battery check mode can be done via triple-click from Off (will count the voltage). Granted, this is a slightly more “accurate”, but uses more time to check…
Three clicks from OFF. The first flashes are the Volt units, the flashes after the pause are decimal points. So, 4 flashes followed by 1 more after the slight pause means 4.1V. Click once to cancel battery check.