[PART 1] Official BLF GT Group Buy thread. Group buy officially closed! Lights shipping.

Fritz,that`s it :+1:

Regards Xandre

Tonight at 23 local time (~9 hours from now) I am going to build the poll
Including two pics (or one with the design twice) of each design and a link to the post of designer where they can have more pics and a talk aboit it (so for Fritz the blpost above for example)
Of course an nonsense option and the possibility to choose “I am OK. With whatever design, just make the GT”

Very nice look'n design! I like the overall shape, just not a fan of all the smooth surfaces. Not sure bout the whole thermal analysis - the fins as designed do not gain all that much surface area - too few, too thick, too shallow, and from their location and size, I'd say they will do very little, not even worth having. Are the threaded connections taken into account, and if so, how does/would a good thermal conductive treatment on those threads effect it? I've seen actual thermal imagery of the effect and it's significant. Also for thermal imagery, all depends on time lapse - hot zones tend to spread over time. What you are showing there is a simple snap shot at one moment. You reach a critical max temp, then it spreads over time. In reality, it's pretty complicated, and with other factors you mentioned - holding it, ambient temperature/humidity, etc.

Hi, you are of course right, this is just a theoretical consideration. However, the conditions are worse in the analysis than they are in real life and it is just meant to see that it works. The heat distribution with the boundary that heat is added as well as the ambient temperature will converge to some equilibrium for a large Time t. Those thermal analysis consider the equilibrium state.

This has been one heck of a thread. Hard to follow as it seems to move at the speed of light :wink: but fascinating from the start.
Fritz, while I don’t necessarily want this light to be the BLF GT that is only because I was already completely sold on the 5AR design. Your unexpected entry here took me by surprise. I would definitely love to see your light in production not necessarily as the GT but how about the GTX? :student: I’m at the point where I refuse to choose. The unique look of your light with the handle is just awesome. I want them both. Well done. :+1:

I like the surface treatments of the HaikeLite MT07 for example, don't care for the smoothness of the TN42 design, though the overall shape/form factor of the MT07 is awful and the TN42 is better. The MT07 and TN42 are almost two opposite extremes of designs for big lights - TN42 is designed for light weight and a thin look, MT07 doesn't care how heavy or butt ugly it is...

Hi, thanks! Maybe if it won’t be the BLF GT design something with a 90mm reflector and an XHP70 would be interesting as well :smiley:

I’m IN! :+1:

Fritzs, I am really curious, what happens if you do the thermal analysis with ~240W+ of power from the LED as would be the case in a quad XHP70? Or even 120W from quad XHP35? Or what about ~80W from a single XHP70?

The idea with this light is to be overbuilt to the extreme to allow for anything we could possible dream up later to be easily done.

This is relevant, I am sure this will be a popular mod for some that need to maximize lumens, given the battery choice.

I can do that. To be honest, I am kind of hoping for a single die XHP70 which I could put in this light haha. But I actually feel like just taking the wattage of the LED is a little bit overkill. Do you know how much of the overall power gets actually converted into heat? Because surely not all of the 10 W of an XM-L2 are just heat right? Or am I wrong here?

I also have to say I feel quite insecure about the convection coefficient, as I don’t have any experience there. I also don’t find any real world values. Here they have a table with velocity and convection coefficient:

So according to this plot from the same page

I’d say a value around 27 sounds about right.

The majority of the wattage turns to heat, the only way energy is removed without heat in the light is the little that the light photons carry away.

I think it is around ~20% at the levels we drive LED’s at the most.

For ease of math I would just use the full number because someone will drive them harder and make up the difference lol.

The quad setup I know will be thermally limited even in a light this large but the longest possible runtime before a thermal stepdown would be ideal.

I think it is the opposite of luminous efficiency. A quick search suggests that about 75% of the input energy as heat would be a reasonable value.

Edit: Texas_Ace beat me to it. But it’s good that we came up with about the same numbers.

Tactical ring on the battery tube has a function as a place for a shoulder strap to attach, and of course that is removable.

cigar-carrying it LOL

fritz15 you're right, here compared to your new design.

fritz15 Congratulations on your design, I know it is not easy to make such a harmonious design. Well done.

and other designers want to thank you for a well done job.

Hi,

I found a lengthy topic on the cpf here: LEDs waste 75% as heat | Candle Power Flashlight Forum
The bottom line seemed to be that you take the maximal theoretical value of 300 Lumens/Watt and divide it by the efficiency of the led. For example for the XHP70:

3249 Lumens at 85 degrees Celcius and 4.8 Ampere and the Vf is 6.3 Volts. So 4.8 A x 6.3 V = 30.24 W
Next, 3249 Lumens/ 30.24 W = 107.4 Lumens/ Watt.

Now dividing the 107.4/300 = 0.358 which leaves us at an efficiency of around 35%.

Here is a thermal analysis with 120 Watts of heat, 20 degrees ambient temperature and a convection coefficient of 27:

Unfortunately I can’t perform an analysis x minutes after applying the heat.

Edit: Wow, awesome thanks 5ar! Well to be fair your design is amazing as well and I am looking forwards to seeing the others, too.

LED’s at their rated levels are indeed generally closer to 35% efficiency.

I have no idea where the 300 lumens per watt comes from or how it plays any role in this?

Although by some numbers for the XHP70 from my own test that is closer to how we would use it:

XHP70 @ ~10A has closer to ~66 lumens per watt, less if driven harder (rough math based on my L6 results, I have not had an XHP70 I wanted to kill for a proper test)

So that puts the final efficiency based on your numbers closer to 22%. Which is about right based on my other research. No idea why though.

You have to remember that we overdrive the heck out of emitters and they generally end up around 50-60 lumens per watt by the time we are done with them, that is unless they pop or “max out” first.

Thanks for the test, it gets nice and toasty for sure!

One thing that is shown is how the fins being below the shelf really reduces the amount of heat they can dissipate. The heat is all concentrated right around the shelf.

Luckily it would be enough for the stock setup, it would just limit the max brightness levels of modded setups.

Click on the image to view full resolution

5ar design v0.6 photorealistic renderings

This is my vision of the BLF GT flashlight I hope you like it.

I accepted your suggestion, and I think we have a pretty design and performing flashlight. Thank you all for your support and advices.

That reminds me, the other designs need to be sure to incorporate the threaded head section thus allowing the head to be made from a smaller hunk of metal saving money and allowing easier access to the LED.

Wow, that’s amazing!

I like it a lot. Which file formats can you export? Maybe I could perform a thermal analysis of your model. It would be interesting to know if it maybe performed better than mine.

The 300 lumens per watt is, as far as I understood it correctly, the maximal theoretical value a white led can reach (wikipedia). At this point the led would have a efficiency close to 1 or 1. Now if a led puts out around 150 lumens per watt it would have half the efficiency of a perfect led and consequently around 0.5.

But you are right, the way leds are driven here they are less efficient. But hey, our numbers agree here so it sounds about right. :+1: