Maybe after too many complaints for DT70 running extremely hot after just a minute they decided to make more advanced cooling that will always allow you to make as much lumens you can get and keep the temperature acceptably warm at 50 C… So 5000 lumens for 3,5 hours sounds more then enough for me.
One thing to note is accuracy of the temperature sensor. Just like how there often is reported of an off-set voltage readout with the DN and DT models, so could temperature sensor fluctuate a lot as well between actual and measured values.
They are indeed crazy specs. I also would have only expected 1 minutes of run time at 32k lumens unless it had an internally built-in fan for cooling. The run times on the lower modes are just glorious with the 8 x 18650 pack. What I am drooling over just as much as the lumen figures are the throw figures. 160 kcd out of a flood light is insane. I cannot wait to receive it. Apparently they will be releasing neutral white in the future. I was also told that the cool White Version will be probably more like 5000k.
Wow, I’m just now noticing this. The guy in the picture is not even holding the light. Plus it’s an aspheric lens beam pattern. That’s pretty shady right there.
About the heat and the runtime, I did some back of the envelope calculations. I collected some specs and data and wrote them down.
Now, from experience, I have noticed that these guys use the term “theoretical” when they state lumen vs actual.
It’s easy to figure out where their stated 32,000 lumen comes from, that’s just 8 LEDs times 4022 lumen each
Here is one way to calculate “theoretical” run time until 50 degree Celcius kick down. I will assume here that all heat is sinked into the body of the light and that there is no delta T from the star to the tailcap. Also I will assume there is no heat shed via convection or radiation during. (there would be some, and that would extend it somewhat)
Using the data I collected and something I call “dimensional analysis” I came up with a formula using relevant numbers from above and came up with this.
71 seconds until the body of the light reaches 50 Celsius, up from 25 room temperature.
If I count the wattage emitted out the front using 683 lumens = 1 watt, that would decrease the heat wattage from 256 to 209.15 Watts.
That would extend the runtime to 105 seconds. That is a lttile more, but not the stated 3 minutes, 180 seconds.
Even so, my calculations are very liberal and would not reflect real world results. All kinds of factors would reduce the runtime till kickdown.
Here are some that quickly come to mind.
Decrease in efficiency of the LED as temps go up.
Added heat from the batteries.
The Delta T from the LED and star to the body.
etc
As I mentioned, there are a number of other unaccounted factors that would increase the runtime, but not by that much IMHO.
Also I could be wrong in the use of my numbers and assumptions.
I think we all know that this light can not perform quite as they state, but WHO CARES, I want one!
I have to strongly disagree with you. The first part is definitely not what you call a back of the envelope calculation.
This is what is called a two ply toilet paper calculation.
Different source, but at the bottom there is a video of the DX80 seemingly not functioning too well, although it could be the capacitors discharging: 手电吧-百度贴吧--手电吧,手握光明,何惧黑暗!--手电吧——手电爱好者分享交流的场所,欢迎入坑!
There should be 8 x 30Qs in there. It’s 4S2P charging, so they probably don’t want people to stick their entire collection of 18650 cells collected throughout the years in this light.
Yup, me too. Even imalent on Facebook didn’t know when they told me they thought they were individual cells. I Think it would be quite easy to make your own pack with 8x 18650s, soldering and tape.