I would like to see much thicker walls behind the fins.
The fins on the right side will be virtually useless as there is no path for the heat to travel.
Same for the upper fins on the left side.
You want the body of the light to hug the reflector outline as close as possible and make the fin depth such that you give the heat a good path to move up.
Note that my L6 with fully charged Liitokala’s typically does about 6000 lumen (P2 1A bin, 17A at tail). Since I use it at my work I have to be a bit more practical with it so I usually ramp it down just a little from max. This is like the 80% ramp we have in the GT. In my case it tends to be about 2500 lumen. I think that making it a bit higher would be okay. We just need to try it first hand.
If the prototype SP70 can do 9000 lumen, then that would be great, but I don’t think it will. It may end up around the 6k to 7k range. If it can handle 4000 lumen for 10 minutes then I would be fine with that. It’s a matter of finding that sweet spot where the output only seems to drop a little, but the amp draw drops a lot. I know what you mean. Plus I only know the L6’s mass and heat characteristics. The SP70 might be quite a bit different.
So maybe:
1-2 Moonlight
40-50 Low
250-300 Med 1
800-1000 Med 2
2500-4000 High (whatever is practical)
5000+ Turbo
Worse case scenario, set your own “High” level in ramping mode. 
I think beefing it up as the red area above suggests is important. It should be easy to do… so let’s try to do it. :+1:
I like the proposed levels as well, covers every practical use I would think, with sensible intuitive spacing. The high level should be determined by sustainability though, not predetermined. Turbo of course should be at or close to maximum achievable for a 1 to 3 minute burst.
I really like the 3 emitter design with deep heat sinking, would do for this form factor what the C8F did for the C8. If I recall correctly previous discussions on fins and spacing the hand drawn fins are maybe too close?
:+1: Anywhere in that ballpark would be incredible for a midrange thrower like this, and would really put it in another league.
Nice! Sounds good.
It’s not a cad drawing! Lol
I just wanted to show the shelf idea and I had to do it quickly. Feel free to tweak it or do your own drawing.
4-5000 lumens high is too much I think. MF02 has much more head mass. And the jump from that to 6-7000 is not significant. 3000-3500 lumens ramp max can run until battery dies with slight warming up.
OMG this is so important especially for a sustainable high output light like this! :+1: I don’t want a 7000lumen light that loses 2,000 lumens in 1 minutes
lol, I am aware but for some reason the china manufactures love to make the walls that thin unless we specifically tell them to make them thicker.
Basically they should leave as much material as possible inside the head, there is really no reason to remove it, the slight weight savings is not worth it IMHO.
Looking good. I like the levels as well. And agree with thicker walls. Thanks to all putting work into this. 
Sorry if this idea has been addressed.
we have “shorty” versions of small lights. How about a “shorty” (2 cell) version of this with a 1 tube if folks want to run 3 cells
Do you mean shorty 26350 2 Cell tube? I’d be in for a one piece 3 Cell 26650 tube for sure.
It already stated as a 2 cell version. a 2/3 cell driver will raise the driver cost. If somebody need more runtime carry 2 extra battery and problem solved with even more runtime than 3 battery.
Two things.
1. Earlier you suggested high could be up to 4k lumens. Why have you changed your mind and now think it’s too much?
Btw, I did make the high setting a more ambiguous 2500lm to 4000lm (whatever is practical).
2. This is not using NarsilM v1.2. There is no ramp max below 100%. Sofirns ramping is not that sophisticated.
It may be a bit early to talk about different battery tubes. That is something that can easily be machined later on.
It would only raise the driver cost to the person who is buying a 9v buck driver from Lexel or Texas_Ace, etc…
I would not expect Sofirn to build a special driver for 9v use.
I didn’t changed my mind. that 4000 lumens I mentioned about my MF02 setted up but it is not a huge jump from there to 9000. So anything between 2500 and 4000 for high mode I will be happy. If in ramping they cannot make separate turbo it is also OK but it need thermal regulation.
I know Narsil is not on option sorry for mentioning it again.
Here is an updated “design concept”.
The side walls are thicker.
Cool, cool.
I don’t really know what Sofirn is able to program for thermal control. Their newer SP32A v2.0 is a FET driver and is labeled as having Advanced Temperature Regulation (ATR). The C8F is also a pocket rocket. Can someone comment on how it’s thermal protection works? Does it keep the light at safe temperatures, etc…?
And I see now it has a working driver because the led glows