Well that was interesting to read… There is something that both “confronted” sides are missing. Look at this post Results: Direct bonded XM-L emitter to Copper mere difference can be caused by different mcpcb between fenix and btu. I’m not familiar with driver/wiring of both but I think we can all agree that fenix has much more experience in that field and it’s driver is probably more efficient. On top of that optical losses are also major factor… I’m quite sure that both flashlights are brutal but there are differences in a way they achieve this.
Ok guys, I decided to do some more extensive testing with both lights. This time I tested the lights, keeping them on turbo as best as I could, for almost 22 minutes. Both were tested simultaneously, alternating one and then the other on my lightbox. Also took temp reading after each lumen test. AND took a video of all of this.
Now some info before any data is listed. My meter was in the X10 mode, meaning every reading needs to multiplied by 10. And as I hope most know, every lightbox needs a baseline value to calculate that number by to get the true lumens. Or estimated lumens. That number is .50. And yes that sounds like an extra easy number, but I assure after testing about 60 lights, 2/3’s of them being ANSI rated, that number is VERY close to correct. If you don’t like or trust my meter or lightbox - go buy and make your own. I know my setup works well.
Also doing these like I did quickly putting the lights on and off, the numbers will be slightly off time to time. The foam gasket I use if not properly put in place can effect the numbers a percent or two. And if the lights are not centered and flat, the numbers can be off a percent or so. Especially with the BTU which is maxing out the glass shelf inside. When doing regular lumen tests I take more care putting these big lights in position. So basically keep this in mind if you see the numbers go up or down oddly from one reading to the other.
I did about 20 readings on each light, but I’m only listing here about every 5 mins or so. You are welcome to watch the clip if you want to observe every reading.
So here we go -
———————————————————————-
——————BTU—————————TK75
1 min - 2860L (82F) ——2965L (83F)
5 min - 2755L (90F) ——2900L (98F)
10 min- 2695L (100F) —–2830L (118F)
15 min- 2605L (109F) —–2790L (128F)
20 min- 2465L (120F) —–2740L (132F)
————————————————————————
So that’s the basic results. The BTU has a soft drop from turbo to high about every 3 min. Dropping roughly to 1900-2000L. The TK75 make a quick drop from turbo to high right at about 21 min. Went to roughly 1200 with the drop. You can easily see it when it drops.
Also did minor testing with the batteries, meaning before and after voltage.
Both sets of batteries were charged with i4 interllichargers.
———————————————-–———————————
BTU 3ea Keeppower 3400mah protected 18650
Voltage at start - 4.20V
Voltage at finish - 3.84V
———————————————————————————
TK75 4ea Panasonic 3100mah unprotected 18650
Voltage at start - 4.20V
Voltage at finish - 3.91V
———————————————————————————-
Should have done the same batteries for both lights, but we didn’t think about it until afterwards.
Now for the extra BTU testing. I decided to see what difference the plain glass lens made vs no lens at all. I was surprised with this one. The test was done after the big test, perhaps 10-15 min later. The light was still a little warm and the batteries were down at the slightly weak 3.84V range. Here is the result -
————————————————————-
BTU with lens at 30sec - 2855otf
BTU with no lens at 30sec - 2970otf
—————————————————————
So this light would certainly benefit from a UCL or AR lens. The stock one measures about 3.6”, but inside bezel measurement is 3.86”, so 3.75-3.80” would be about perfect. With a good UCL lens, cooled down, and fresh batteries, the BTU likely would put out better numbers than the TK75 at start. Though the TK holds its lumens better so would still likely catch it and pass it.
Well that’s all folks. I give the victory on this lumen and runtime test to the TK75. If you intend to run your light for 15 min. or longer at a time and want to keep the full 2500-3000 lumens, then the TK75 is likely the smarter choice. And it’s lighter weight will help prevent hand fatigue. However running the BTU on turbo for 5-10 mins and then a 2000L high until the batteries run out will still give you tons of light and more run time (then it kept on turbo).
With all the testing now and previously, I still prefer my BTU’s throw, UI, and heft. I fully respect the TK75 power and regulation. It’s a heck of a light.
Well guys here is the video clip if you feel like sitting through 22min of pretty boring clip. And remember we’re amateurs on this clip and VERY little talking is done.
Here we go -
Thanks for the testing, you did a very nice job.
The TK75 looks like an awesome light.
Thanks fronty :) for your precious time and patient in doing the tests.
. so the lens is holding back a 100 lumens .its full of impurities. Better AR lens needed from Ric free of charge to all customers . Im changing mine to U3 and it should go above 3000 lumens at turbo
With just a quality lens, fresh batteries, and a cool light, I’m confident ithe BTU will post a reading over 3000otf at 30sec… Add U3’s or XM-L2’s will be gravy on top.
No question there. The TK75 is an awesome light. Fenix did their homework on this one.
Thanks for the wicked detailed trial run on these lights 
Did you take any temp readings of the emitters? That + external temps reveal a lot about overall heatsinking effectiveness.
What a great testing pal.
I believe both lights are winners. I think we won’t go wrong with both
Yup im Adding U3’s asap .
I didn’t do any at the emitters. Didn’t know if I could do that. Will these infrared guns work through the glass?
I’ll likely leave my U2’s in there for now. If I upgrade, it will be a ways down the road and likely XM-L2’s.
I would like to do a UCL or AR lens swap. It was suppose to have that from the factory anyway.
IR temp guns do not work through glass.
Thanks for the testing.
It looks like the BTU could use a little better thermal coupling to get the heat away from the emitters. A 400 lm drop is significant IMHO.
I’m thinking of doing 3x XM-L2 U2 on copper SinkPads in the future.
The IR temp gun thing is new for me. Never used them before and just got this one Saturday. Wanted to play it safe just in case one of the lights started getting very hot. Both did ok I think on temps. But yeah, looks like the TK does a better job of getting the heat away from the emitters. A 395L drop for the BTU vs 225L drop for the TK75. I still wonder if the drop was partially due to the driver trying to bring it down slowly to high. I think the drop figures about a 13.7% drop on the BTU and about 7.6% drop for the TK. The TK did very well for sure.
But I’m not too worried about it. I never intended to use the light for 20-30 min. anyway. I got it for fun. And if I want to use it for extended periods, I’ll just let it go naturally into a 2000L high and leave there. But I know now I can run it on turbo longer if needed.
Swap in XM-L2’s on copper sinkpads, along with a UCL - the BTU would be even more beastly then.
I had XM-L2 U2’s installed… Had some resistor work done. It’s just under 4a on turbo… Really happy with it. ( btu )
Thank you rdrfronty for testing them out.
Your TK75 result match with runtime chart I have seen that output drop about 5% in 20 minutes Turbo mode. That is very good performance I would say.
No comment from me for BTU.
I confirmed from preliminary testing that the driver current does drop during the first minute or so. It started out at 3.7A and dropped to 3.5A rather quickly (emitter current). Nothing huge, but it should contribute a bit to the fallout. I’ll be tweaking my output a bit to compensate, shooting for 3.8-4A on turbo.
Once I move to XM-L2 U2 on copper, I might bump it up closer to 4.5A (depends on output curves from XM-L2 when they are done). This driver seems to be comfortable with high current output.
Thank you for the performance analysis and the temperature / lumens statistics.
Was really looking forward to your report.
Now, for a ton of light and extended use, the BTU is good with a conservative 10 to 15 mins on turbo and the rest on high. This is a 2000 + lumens light all the way until battery failure. Now, with a little tinkering, the BTU is set up for an even greater performance and greatness.
In the flashlight universe, rarely does a torch delivers both - massive throw and flood of over 2000 lumens all the way. The BTU is a truly living testament to all discerning owners and lovers of led flashlights.
Great job RDR. I am certainly impressed.
That slight current drop is good to know. Thanks for that little detail. So perhaps the light loses maybe 200 lumens or so due to heat and 200 lumens due to driver output dropping slightly.
I think a 200mA reduction across three LEDs would be quite close to 150-200 OTF lumens of loss (around 50 or so per emitter). The rest is likely heat sag. Since the driver sage is quite quick, one could compensate by jumping up to 4A drive, which would fall back to 3.8A within a minute.
so we all agree Ric Awes us a nice coated lens ??