Fireflies E07 preview

Yes true, but there is also a lot more space for cooling in a flashlight… vapor chamber seem to be the latest innovation right now in cpu cooling for mobile devices… it might be just a gimmick I don’t know… not seen A/B tests.

There are two ways to keep the light from getting hot so fast:

  • Add mass
  • Reduce power

So if you want something which heats up slower or can stay on the highest level indefinitely, you should look for something much bigger or much less bright.

There’s also the option of trying to increase efficiency of the LED and/or driver, but that has limited room for improvement. Basically instead of wasting 85% of the power as heat, an ideal light could get as low as maybe 56% wasted as heat. Still not great. So it’ll still need more mass or less power.

Some bigger lights have also resorted to the use of active cooling, building a cooling fan into the light itself… but that’s not a great option on small items, and even on the bigger ones it’s kind of obnoxious.

I think some kind of liquid cooling is the way to go…

Unless there’s a way to change physics, the E07 thermal performance is pretty much as good as it gets. Liquid cooling or vapor chamber will do nothing for a light of this size and output. If you don’t want it to heat up that fast, very simple, just run it at 1,000 to 2,000 lumens which is equivalent to the turbo modes on other flashlights of this size. Even a flashlight with 10x the mass will not be able to sustain the max output of the E07. The max output is meant for small bursts.

Yeah but this is one thing that surprised me, it heats up very fast also on High mode… not just turbo mode, with turbo mode it heats up very very fast…we are talking under a minute here…

It actually works really well with liquid cooling. I’ve done runtime tests this way before. Simply run cold water over the heat sink during use, and it should be able to stay at turbo until the battery is empty.

But if you mean a fully-sealed liquid cooling system, that doesn’t actually make much difference. Liquid heat pipes are good at getting heat from one place to another, but they have literally no effect on the overall heat-shedding capabilities. That depends on the surface area and the surrounding environment. So if it can disperse 20W of heat in the current design, it’d still only be able to disperse 20W of heat after having liquid cooling pipes installed. The main difference is it would theoretically heat up the host more evenly, so the back half would get hot faster and it would be more uncomfortable to hold.

Yep.

A vapor chamber works really well if you have the surface area+cooling to be able to dissipate the heat.

The only thing that would be left to do to aid in cooling would be to go the Amutorch way, but with some modification.

See, the Amutorch VG10 21700 has a very thick body tube onto which the head sits on.

It’s one 21700 light that can actually sustain around 1000 lumens constantly, but gets a bit hot.

Why? The thick body tube transfers heat to the battery tube, meaning more surface area can be used to cool off the light.
The disadvantage to this is that since it isn’t finned, it doesn’t dissipate a lot of heat, but still helps.

So, here are the theoretical improvements that could be done:

  1. Make a thick battery tube with a large connection to the head.
  2. Make some fins on that battery tube. Even small ones can help.
  3. Put some high melting point paraffin wax inside of the light.

If you think a minute is fast, you should try some other “hot rod” lights. Some get too hot to hold in like … 10 seconds.

Mostly though, it sounds like you probably should avoid hot rod lights entirely. They clearly aren’t what you’re looking for. They’re designed to burn through power as fast as possible to make the brightest beam possible, even if only for a short time. So if you want the highest mode to last a while, look elsewhere. Or pretend “medium” is the highest mode, and it’ll behave more like other lights. The higher levels are entirely optional, and only intended for brief bursts.

As I said before, Personally I dont mind the heat… I know where the hold the flashlight to not burn… I hold my hand on the battery tube which is somewhat unaffected of the heat… the problem I have is how the heat throttle the performance…sustained performance is important… or atleast improve the heating build up time…

Then this is probably the wrong light for you.

Well… sustained performance is what I wish for and I think every one wish for…that does not mean that very low lumen options are a better option… cause I can get sustained performance using low lumens with this light…

We all want our flashlights to sustain as high of an output as possible for its size. The E07 is actually one of the best out there in this regards for the size and mass. I use my E07 at about 1,000 to 1,500 lumens for long periods, which is as bright as most flashlights of its size running on max output.

The highest sustained output in a passively cooled light is 6,400 lumens, which is achieved by the massive 1,804g Olight X9R with one of the highest surface area (heat fins) I’ve seen in a flashlight. In comparison, the E07 weighs a measely 125g.

You have to understand physics cannot be changed. You can only do so much with a given size/mass and wattage. What the E07 does differently from other flashlights is that it is able to output several times higher than the typical flashlight of its size. But keep in mind, you don’t have to run it at its max. It is an added bonus. For a flashlight of this size, sustaining even 1,000 lumens is pretty much the max allowed by LED tech and law of physics.

Yes but PC industry have the same problems with the physics… yet some manufacture get a lot better cooling with the same mass… so… yes to think all is impossible will not make any creative solutions… to think that it can be better will…

But yes better emitters would be the best solution…
Does any one have any curve that show us how the lumen and watt usage have changed the last years ?
So to know what to expect the coming years in improvement.

In the case of E07 I have already find one tweak that would help a little bit, it would be to change this “stylish” silver ring with a extended heatsink this would this flashlight to look different… but this unnecessary big ring does not fill any function as I can see… And Also the big ring might look stylish but make it also look less “professional” and more for “home use” a´sane with this AUX lights… not sure what trade offs where made to include these AUX lights… related to cooling.

The steel bezel serves an important purpose. It greatly reduces the damage caused by dropping the light. Converting it to aluminum might help a little with heat, but it would be too small of an effect to even really notice.

I gave some numbers a few posts ago, and they’re not made-up… But I’ll explain in more detail:

The theoretical highest possible efficiency of a lighting device is 683 lm/W, but that can only be achieved with a narrow-band green light source. Regardless, we’ll use this as the baseline for perfection.

An old incandescent bulb gets about 5 lm/W. That’s a 5/683 efficiency ratio. So it wastes 99% of its power as heat.

Many of the LED lights around here get about 100 lm/W. This is dramatically better, but it still wastes 85% of its power as heat.

The highest possible efficiency for a white light source is about 300 lm/W. So even a literally-perfect LED with a 100% efficient driver would still waste 56% of its power as heat.

So… let’s say there’s a light which uses 10W of power. Here’s how the numbers work out:

  • Incandescent light: 0.07W of light, 9.93W of heat.
  • Fireflies E07: 1.5W of light, 8.5W of heat.
  • Literally perfect ideal white light source: 4.4W of light, 5.6W of heat.

But more realistically, taking into account losses from drivers and optics and LEDs, we’d be lucky to make 200 lm/W. And that would produce 3W of light plus 7W of heat. So in the future, a light like the E07 could maybe sustain 2400 lm instead of 1000 lm.

Even at the extreme limits of what is theoretically possible, the E07 would still not be capable of sustaining 8000 lm. It just doesn’t have enough mass and surface area to shed that much heat.

The computer CPUs are MUCH bigger and heavier than the E07 and even then it won’t be able to sustain the heat wattage generated by the E07 on max and still be able to hold it comfortably. For a flashlight to use the same cooling tech as cpu cooling systems, the host needs to be much bigger and will no longer be pocketable like the E07. The recent Imalent MS18 is a great example.

The problem right now is that the higher efficiency LEDs are CW, bad tint, and low CRI. The HI CRI emitters like the SST-20 95CRI and Nichia 219B 9080 are much less than 100 lumens/watt on the higher outputs.

We could use PC tech all we want but PC’s always have a constant supply of air (fans) moving across their heat sources that flashlights will never have.

Here’s what you can do:

Buy this XHP50 x 3 PCB. Get a copper spacer. Make your E07 a 3x XHP50 (3V) mule. It will probably sustain the same lumens for twice (at medium outputs) as long but it will be a hideous looking beam and even heavier up front. Did I mention it will be 10K lumens turbo?

some other complaints I have about the flashlight is… its way too big step from high to turbo mode… I would also want to be able to ramp up to turbo mode… or atleast half turbo…

But I think the most stupid thing with thish flashlight is that for some reason you have to double press twice to enter turbo ? what is this, some kind of “child/security” protection ?
This is very stupid and not necessary, I am pretty sure all agree with me. I am not the only one who complained high to turbo is too big step…
or the high setting is too low… compared to turbo…

It really get very very hot… I wonder… what temperature in celcious should I set I am going to try to reconfigure the sensor… next week… sigh
is 90 degrees safe for the electronics ?

Anduril UI can be programmed to ramp all the way to turbo. You get the flexibility to program the top of the ramp to whatever you want it to be.

This is good news!!! I will read up how to do it…

Hi all,

What’s the mcpcb size for the eo7? I’ve been wanting to convert it to an dedicated mule, the mcpcb is a bit too deep, thinking of adding spacers to bring it up closer to the glass