SEE POSTS FOR UPDATE:
I compared the SR32 to my other brightest lights like MS18 Acebeam X75, and my GT4.
This light is so much better than the MS18. The only thing I can think of I like better about the 18 is the way the carry strap hooks on at the front.
Neal sent me this light for review. Along with this 25% off coupon code: Lumencraft32
I could use some math help if you guys don’t mind. I don’t need it to be perfect, I just want to get a realistic estimate. It’s hard to get an accurate measurement on the Imalent SR32 drive current, but I wanted to know the wattage.
I read that it uses 8x 21700 cells in the battery pack. I’ve not seen a test of the XHP50.3 HI that overdrives, but Texas ace did one on the 50.2 HD
Taking Imalents claimed 120,000 lumens divided by the 32 LEDs is 3750 lumens per LED. Looking at TAs chart that would indicate 34.2 watts used
at that output. So 34.2x32= 1,094 watts. However, the 50.2 HD is 4.85% more efficient drive at Crees stated max. 2160 lumens vs the 2060 produced
the 50.3 HI. In addition TAs chart 11.5% higher output than what cree claims for the 50.2 HD emitters at spec. So like 16.35% higher in total (theoretically) than what a 50.3HI would put out. So 3750 lumen x 1.1635= 4363 lumens would be the actual target value to look up on the chart.
This correlates to right about 46.8 watts. 46.8 x 32= 1497.6 watts. Is this right or am I way off somehow?
No idea on this specific light of the 50.3 since I have not used either.
But I am not surprised in the slightest if they were buffing the numbers for clout.
The bigger question on this would be the power it is pulling from the cells IMHO. If it is 1000w, that is 120w per cell or around 30A depending on the driver. If it is a buck/boost drive it could be more. That is about the safe limit for 21700’s unless they limit it to specific models.
If it was 1500w, that is 190w per cell or around 45A, that just seems way too much for a production light unless they are somehow controlling what cells are used and everything is top notch. Not to mention that building a driver that could handle that much power and is not FET based would be difficult to say the least.
Overall, yeah I would think they are exaggerating the claims to some extent, how much is the question.
Thanks Matt…Great video as usual!
Without incorporating any of the numbers that you guys did, just based on most of these manufacturers and their blown up and exaggerated numbers, my guess it’s 90,000 to 100,000 lumens.
It’s still an amazing amount of light.
Unless you had one light that was absolutely 120,000 Lumens and another one absolutely 100,000 lumens and you put them side by side , I’m not sure you could tell the difference.
I built this heavy test rig so I could find out how many WATTS the SR32 is putting out/consuming. Put your GUESS in the comments.
Video comes out tomorrow…
Looking forward to that video as I can’ t make heads or tails about how the light and batteries are going to be attached to that rig! 
+1. I also want to know what is inside that thing…
A glance at the XHP50.3 data sheet, making some absolutely bonkers naïve assumptions:
120,000lm / 32 LED = 3750 lm/LED3750 lm/LED / 1200lm ref = 3.125 reference current3.125 • 0.7A = 2.1875A
2.1875A • 12V = 26.25W/LED26.5 • 32 = 840W
EDIT: Typo
Back of the napkin is fantasy. Reality will be markedly worse. Lumens claim is suspect. LEDs won’t be top bin. Driver will have inefficiencies. Fans introduce loss. And heat will surely impact LED performance to the point that the output curve has started to flatten.
While I’ve been peripherally aware of the likes of Imalent and Acebeam producing >10k lm lights for some time now with active thermals, do any of them really push anything close to 1kW of power? That’s more than half of what you can net from a typical wall outlet in North America.
This has to be one of the best flashlight videos I have seen. Love the comparison with the older technology.
Well done.
Just shy of 950W per the video I found on YouTube. Impressive.
Beautiful work!
Thanks for the video
Wow. That looks like it can light up a football stadium.
Critical Update: My Imalent SR32 120,000 lumen has LED dies in it that have BURNED OUT!
This is an issue they have had before. I waited to get my light while Imalent addressed the problem (why my video came out 2 months later than other reviewers). During the initial two weeks of testing I had no problems at all. I CHECKED THE DIES and reported in the comments section to viewers that I was having no issue.
After this I ran a series out output tests where the light runs approximately 10-15 straight minutes (total of about another 45 minutes use on turbo across 3 tests), during which time the thermal protection did kick in and should have prevented this. I then sent the light to Ethan and he ran a few more similar tests and that is when he discovered the dies had failed. This failure occurred during what I consider to be “Normal” use. No “torture tests” the light made.
I intend to go through the process of getting the light warrantied. I will post updates to the situation as it unfolds.
Interesting, this is another kind of defect. Previously the phosphor was burned, now one half of the emitter is dead, like it desoldered itself.
It may not be a manufacturing defect, they may have exceeded the capabilities of the materials.
I recall that the peak power measurement was very nearly 950 watts which calculates out to 29.7W/ LED, gross (driver overhead will reduce the net, obv). With that many LEDs spread over such a large area being markedly overdriven at turbo it’s entirely possible that there’s sufficient variation in heat dissipation LED to LED that some cooked without tripping thermal throttling.
It’s Imalent as usual. Some are fine, some aren’t, but t9 be fair I had this happen to a Fenix LR80R. On a Turbo test thermal limit kicked in, but too late (at 120 C). The reflector started melting and a couple LEDs (sst70) started to fail. They warrantied it no problem! Brand new flashlight.
I can only speculate, but I’ve seen this particular thing happen many times. It’s been my experience that the reflow soldering can have a lot to do with it. Mass production requires the use of solder stencils and machines that do tend produce very consistent results. However there are two slight imperfections you can expect from it. Number one is that the LEDs will be sitting up just slightly (probly less that a thousandth of an inch), on the puddle of solder. Or if the solder paste was reduced slightly it can sit all the way down, but there can be very thin or slightly inconsistent spots under the LED.
Believe it or not I’ve actually found that the most reliable thermal path comes from reflowing the emitters by hand. Puddling up the solder and then pressing down on the dome of the LED, then sweeping away the excess solder. Hard to do on a large scale though.
I agree with what Zoulas said about exceeding the emitters capability. It could be that Crees QC is just not tight enough to support this kind of power. A variation from emitter to emitter could easily cause this. We all accept that driving the LEDs this hard is fine, but Cree is only targeting about 18 watts max for the chip, not 29+. I would bet that this light exceeds the recommended thermal junction temp as well, but I have no data to back that up.
It could be both. Hard to say though. Those are robust leds and you can run them pretty hard and they survive. cooling and thermal path are very important at that limit. When you have that much heat concentrated over a small-ish area, it can overwhelm a so-so thermal path. If you’ve got multiple LEDs failing in the same way, to me that’s less quality control and more of a design limit failure. That’s what happens when you push the envelope.
Based on Cree’s “XLAMP portable LEDs” feature sheet document, Cree evidently thinks it is reasonable to overdrive the 6V XHP50.3 Hi to 5A, with good heatsinking.
As for the MS18, I don’t know how hard the LEDs are being driven or how they are wired.
