Maelstrom S12 OTF Lumen Measurements

First of all I'd like to say thanks to CPF member Got Lumens? for lending me his S12 for testing. Now for my results:


This light has thermal protection, which keeps it from being damaged by too much heat. I wanted to test this light in the same manner that most users will actually use this light, but I also wanted to test it with additional cooling methods, just for giggles..... So I used a freezer pack as one method, and my hand for heat sinking as another method, and also without any additional cooling at all, which means holding the light by the tailcap and letting the thermal protection do its job.

Obviously the freezer pack method gets warmer the longer the test is in progress, so this is not exact or controlled by any means. Also, I'm not certain if anyone actually lives in an environment that could keep this light as cold as my freezer pack for this amount of time, so please take this as it is. Also, when testing with my hand as an additional heat sink, I didn't see any need to run the light any longer than 20 minutes because the light had already stabilized, plus I wasn't doing a runtime test. Obviously the runtime will vary like the output at different temperatures, but I don't have time to do those tests. Here are my results:







I also documented the temperature throughout all three above tests. I used an infrared thermometer to record degrees fahrenheit, and while I'm aware of the fact that infrared thermometers have some degree of error, especially if you're trying to take readings from a surface that's reflective (not the S12), I feel like it works well enough for what we need to know.

Here are my results:

120 degrees isn't that hot. (and of course not even that hot in your hand) I'm surprised it doesn't get much hotter considering how much output it loses.

Thanks for all the work. Not the light for me.

Seeing the OTF numbers makes me want to purchase one even more. To prove how much better it would be in direct drive mode off IMR 26650 cell.

Of course I should have gotten the copper model.

Wow, I would have never thought that "hand heatsinking" made that much of a difference... you must have icewater running through your veins! :naughty:
Great comparison, now I can't wait for my S12 to get here. Were you holding onto the body like you normally would or the bezel?

Thanks for doing this work ti-force! And thanks to GotLumens for donating his light for a bit. That's crazy how much holding it in the hand helps. Just doing that keeps 200 extra lumens coming out the front end....impressive.

Thank you ti-force for the work you put in for all of us.

This "freezer pack" curve is much like what I experienced with the S12 on my bike for night rides.
It was about 55-60F degrees with air passing with ride speed. The light felt like 70-75 degrees when I felt the head during the ride. Thermal management did not come into play. I saw a gradual decline in lumens over the hour ride due to battery drain.

It is nice to see that the S12 will maintain 600 L if hand held for heat sink. I was wondering about that.

I would like to say Your Welcome for the loaner. My part in this was the easy part. I would have said this much earlier, but the url that You sent me PM didn't come thru .
I just want everyone to know that Ti-Force has worked long and hard testing this S12 with built-in Active Thermal Management, and deserves a :thanks:, as many have said already.
With all Ti-Forces hard work, we all can now have a base to play from. Like say doing the tests that Ti-Force doesn't have time for. JermsMalibu brings a good point about the results regarding hand sinking. It will be curious to see if handsinking would reduce the amount of runtime because your driving the LED harder. And, What about that S12G's copper heat sink? Agh, I think I'm afflicted with Flashism. :sick2:

So how can we override thermal regulation? There should be a sensor of some sort that I could remove...

Allex said:

So how can we override thermal regulation? There should be a sensor of some sort that I could remove...

Click to expand...

Do you want to fry the LED? I have a Fenix L2D Q5 that appears to output ~100 lumens on turbo, but probably a lot less, compared to the manufacturer spec of ~180 lumens and the CPF measured value of ~140 lumens. I think I fried the emitter by repeatedly running on turbo for 1 hour.

Thanks for the kind words, everyone :thumbsup:.

brandocommando said:

Wow, I would have never thought that "hand heatsinking" made that much of a difference... you must have icewater running through your veins! :naughty:
Great comparison, now I can't wait for my S12 to get here. Were you holding onto the body like you normally would or the bezel?

Click to expand...

Hehe.... flashaholic blood is powerful stuff . I was actually holding the light by wrapping my hand tightly around the head for maximum heat transfer at the hottest part of the light. Also, I actually alternated the light between both hands every 15 seconds (roughly) throughout the entire test. This allowed some air cooling for my flashaholic blood, while the other hand wicked heat away from the S12 .

Got Lumens? said:

I would like to say Your Welcome for the loaner. My part in this was the easy part. I would have said this much earlier, but the url that You sent me PM didn't come thru .
I just want everyone to know that Ti-Force has worked long and hard testing this S12 with built-in Active Thermal Management, and deserves a :thanks:, as many have said already.
With all Ti-Forces hard work, we all can now have a base to play from. Like say doing the tests that Ti-Force doesn't have time for. JermsMalibu brings a good point about the results regarding hand sinking. It will be curious to see if handsinking would reduce the amount of runtime because your driving the LED harder. And, What about that S12G's copper heat sink? Agh, I think I'm afflicted with Flashism. :sick2:

Click to expand...

Sorry about the missing link in the pm. I guess I simply forgot to insert it...... I've also thought about runtime differences, and from looking at the data, I would say that without any additional cooling the light would have the longest runtime. The next longest would be the hand heat sinking method, and the shortest runtime would be the freezer pack method. Looking at the lumen comparison chart we can see that the hand heat sinking test started to overtake the freezer bag test at around 18.5 minutes. This tells us that the freezer bag test was decreasing in lumen output because of battery depletion. After putting all the data into a chart, I really wish I would have taken the hand heat sinking test out to 30 minutes........... but I didn't. I feel like the light would have stayed pretty flat up to at least 30 minutes, but that's just a guess.

I read a comment from 4sevens in the past, stating that the copper heat sink didn't make much difference at all for this light, and that's why they abandoned the copper model. It's a fact that copper has much better thermal transfer than aluminum, but perhaps it just doesn't make much difference with this light because it's thermally controlled. The only problem with directly testing a copper S12 against an aluminum S12 is the fact that no two lights are exactly the same, and there are quite a few variables that can't be removed from these lights for a direct comparison between the two. That's my opinion, what do you think?

There is obviously more than heatsinking and thermal management here at play in those graphs. At 18-20 minutes there is @42° difference between freezer pack and hand hold with @0 Lumens difference and @10° difference between hand cooling and no cooling and a @200 Lumen difference. Then at 90 minutes the freezer pack curve is @5° cooler than hand cooling, but is down by 200 Lumens. I'm not downing the charts at all; I think the effort here is applaudable and shows real world usage. I just don't think you can draw any finite conclusions as to the scientific operation of the light and it's thermal management solely. What I would be interested in seeing is two (well actually four) things: (1) a battery discharge curve and (2) the tests above performed with the light powered by a constant voltage, a constant current and then a constant power source. I would suspect this light is affected more by source voltage and current capacity than it is its thermal management. Purely out of scientific curiosity on my part, however. I own a S12 and have no valid complaints for my usage. I just like knowing how stuff works.

recDNA said:

120 degrees isn't that hot. (and of course not even that hot in your hand) I'm surprised it doesn't get much hotter considering how much output it loses.

Thanks for all the work. Not the light for me.

Click to expand...

It doesn't get much hotter not because its "losing" output but because the thermal regulator is dialing back the current to the LED and therefore cutting back the drive Power intentionally to keep it from getting hotter. To go from over 800 lumens to 400 lumens the power is roughly cut back by at least a third and something less than half of what it was. Not the same as the phosphors over heating and the radiant light coming out of them goes down due to loss of efficiency.

Overdriving this LED would not make it brighter it would most likely kill it or ruin the phosphors on the die surface. They knew what they were doing when they put thermal regulation in to protect the LED and also the driver chip from heat.

Really great work Ti-force.

It appears the freezer pack temp starts at 50 plus F. Not what I would call frozen unless you didn't actually freeze it intentionally? I would expect that the IR sensor is not reading it correctly at the surface. It would be interesting if you shot your IR sensor "gun" straight into the freezer or into the ice tray and if it would read correctly? You could always get a magic marker and make the area where you are reading the ice pack black and see if that helps to get a better reading. I have seen guys make the metal cans for their RC car motors black to get more accurate IR readings. In general the more reflective the surface is the less accurate the reading from what I understand. I have also seen people wrap things with 1 layer of black electrical tape to get a better reading off of what was a shiny metal surface that was reading incorrectly.

Nice lumen measurement, it is good to know that the 800 lumen (or maybe 850) is with fresh batteries and a cold light.

The reduction in output when measuring with the ice-pack is mostly because the light is unstabilized, the output will drop with the battery voltage.

I have done a review of the light with my usual voltage curve and also a runtime curve for both hot and cooled light.

MrGman said:

Really great work Ti-force.

It appears the freezer pack temp starts at 50 plus F. Not what I would call frozen unless you didn't actually freeze it intentionally? I would expect that the IR sensor is not reading it correctly at the surface. It would be interesting if you shot your IR sensor "gun" straight into the freezer or into the ice tray and if it would read correctly? You could always get a magic marker and make the area where you are reading the ice pack black and see if that helps to get a better reading. I have seen guys make the metal cans for their RC car motors black to get more accurate IR readings. In general the more reflective the surface is the less accurate the reading from what I understand. I have also seen people wrap things with 1 layer of black electrical tape to get a better reading off of what was a shiny metal surface that was reading incorrectly.

Click to expand...

Thanks G

I actually recorded temperature measurements from the light itself. During the freezer pack test I simply wrapped the light inside the gel pack with a couple rags on the exterior to keep my hands from freezing haha.... Then I let the light sit for a few minutes to drop the light temperature down from 74F to in the 50'sF. Fresh out of the freezer, the gel pack measured 3 degrees F. I guess I should have stated that earlier. Also, room temperature was 74 degrees F for each test. Good idea's for taking readings from reflective surfaces :thumbsup:, I'll have to remember those .

Thanks for the data and charts.

Again, Ti-Force, :thanks:.

I agree. I think the battery discharge curve had alot to do with how the ATM functioned. By using a controlled power supply we could see and measure the affects that the battery condition has on how ATM functions by regulating it's power output.

I get more from Ti-Force's test results, what is typical, than from testbench results, what is possible.:candle:

As far as directly comparing the S12 to the S12 Copper, I think the results would vary between the two across all the different test scenerios. Although the differences may be measurably reportable, I Agree with Ti-Force, in the end, ATM in both the lights wins over external factors with its designed function of thermal regulation. Including the differences between the thermal behavior of the copper and alluminum heatsinks to keep the LED comfortable:sweat:

I will be testing runtime between copper and alluminum very soon, and will share my results.

Where's the Micro-OBDII plug to check temp on a S12 PCB when you need one

Anynone knows how hard it is to open up the head? Then I'd buy myself one and get it modded to DD.