The BLF GT was a great project and produced a great flashlight, but the execution at the most critical times was unfortuneatly shall we say, a bit trying (not from the fault of any of those who stepped up in a storm of issues, hat’s off to all who made it happen, it was just not great to have things bouncing a thousand directions all at once)
To the point of I think a lot of members are at this time saying, the last thing we need now is a new project.
It was a project that became as they say “Became a life of it’s own” and it claimed victims of team members and BLF members at large, the forum has changed since this experiment for better or worse, it is what it is.
JMHO
Later,
A small note, I don’t think it claimed any team members. If you are thinking of the Miller leaving, that was due to his illness and personal life needing his attention. Plus other people stepped up to fill in his role.
Was there a different team member you were thinking of?
Maybe I said that wrong.
What I meant to convey is a portion of senior members who had projects in the works or who were regular contributors have seemed to have left the site either temporarily or permanent of just AWOL since the first of the year.
The large influx of youtube, facebook etc. additions has changed the vibe here for sure.
I know, you can’t stop progress.
Thanks.
Nitecore makes some compact lanterns, like the LA10 or LR10. In case that would be close enough.
There’s also the option of putting a diffuser cone on a regular AA/AAA/16340 light. Then it could serve as both flashlight and lantern. The Olight S1 Mini Ti works well for this, and IIRC it comes with a GitD cone.
I know, I know. Sometimes I’m acting a bit silly. But I can’t help myself. This light is so ridiculous that I just had to draw it. Or maybe it’s not?
This is expensive light. High budget. But there’s nothing cheaper in its niche.
Well, there’s nothing more expensive in that niche either.
Single 21700, CFT-90, FET+linear driver:
It seems like the driver designers on this forum have pretty much concluded that you need a high-powered buck driver to handle the CFT-90. I think this is why no one is really building a single cell flashlight around that emitter.
I suppose if you could get a driver to work well and then match it up with something like a Samsung 30T you would have a pretty interesting light.
I would like to make a donation for all the great work you guys have done, and for all I have learned on this great forum.
I have also bought some of the great BLF developed flashlights, and saved money with coupons etc.
Can I please confirm that this is the correct address to donate to on Paypal:?
The Miller has been pretty busy lately and hasn’t had much time for forum stuff, but I’m sure one of the other members like ToyKeeper or Texas_Ace can confirm things given a little time.
Given the situation with the person who started this thread, would it not be a very good idea to have sb assign this thread to a person active in BLF??
Or, at the very least; remove the donation information as it is now…. (copy & pasted below from post 1.)
wimmm777$gmail.com (replace the $ for @)
This outdated information should not be left active as far as I am concerned.
I’ve considered opening up a Patreon account or something in case people feel like donating to me personally, but I always found it a bit odd having a general BLF-project donation fund. Fair and agreeable distribution is difficult, and I don’t like handling other people’s money.
Anyway, The Miller has been mostly absent since late 2017. I don’t think he is really following current events on BLF or distributing donation funds any more. He has been a bit too busy dealing with illness and flooding and other issues.
I still do not see why a simple FET+1 driver would not drive the CFT90.
About the design: you emphasized heavily on the finning (I like that) but still what you have drawn I estimate can dissipate about 25W, not the 100W that you want the CFT90 ideally to run at. But since people do want that 100W (or whatever the 21700 will provide) for a while before the stepdown, you can buy some time with extra aluminium mass near the led, also good for distribution of heat to the fins. So I would suggest a bit less deep fins and a bit more aluminium.
I would love to know how do you estimate the performance of these fins?
Frankly, I think it would be good to make the reflector like 4 mm larger. Right now it’s quite small when considering the total size of the light. It was just easier to draw it like that because it’s just a modification of one unpublished drawing……
Making reflector larger would improve heat shedding. Though certainly not quadruple it. My intention with this light was to have fins remove at least 70% of heat. If that was the case, already quite large chunk of metal behind the LED should allow for a nice Turbo times - and superb sustained performance would be a nice bonus.
It would be wrong to treat those fins literally. I think if such light was to be done, someone should just design a light with lots of fins in this area, adjusting exact fin spacing and thickness to whatever thermal CAD shows to be best within structural limits.
But I was curious whether your 4x was about right, so I did just that: treated those fins literally and calculated area.
I got 5 times the area. So about right. Though in your S2+ analysis I see the light being run at 10W+, not about 6 like your calculation would show.
It seems to me that this light can shed very very roughly 50W with 90 C body temp (I don’t see ambient temp in your tests…can I assume 20C?). Or very very roughly 25W with more sensible 60C. In the winter, with –10C ambient very very roughly 50W again.
Though it doesn’t include cooling with hands, it adds some.
Minor increase of fin area by making reflector larger can help, but not enough.
One thing you have to watch for with heatsinking fins is that deep, narrow cuts between fins can result in stagnant, hot air at the bottom of each cut that doesn’t circulate because it’s buffered away from the cool air outside by the warm air in between.
This means that you can’t assume heat dissipation will scale purely with the surface area. The arrangement of that surface area in terms of getting as much free air circulation as possible also matters.
There is often an optimal point in width, depth and separation of fins, sitting somewhere on a spectrum of:
no fins and poor heat dissipation;
through several fins with optimal width, depth and separation for free air circulation and good heat dissipation;
to lots of thin, tightly spaced, deeply cut fins with stagnant air zones at the bottom of the cuts and mediocre heat dissipation.
The last option is better than no fins at all, but not as good as the optimal arrangement. In the absence of (expensive and complex) modelling software, it’s not unusual for manufacturers to build several prototypes with different fin arrangements and physically measure which does best.
OK, I hate to not have drawings. I still consider the questions above to be open, but I drew pretty much the smallest variant:
10 mm TIR optics, twisty with QTC pill and bypass to real direct drive (thanks CRX), tiny charging connector.
Pictured next to DQG Tiny AAA.
Adding features will make it larger than Tiny. So will increasing TIR size.
Though Tiny has 15-20 times lower output, so maybe that’s not the best light to compare to…
Is it worthwhile? I don’t know. But I’m partial to QTC twisties.
