Maximum Lumens Over Time Via Liquid Cooling of Hand-Held Lights (Flashlights)

Rusty_Joe · 2018-03-06T05:54:28.000Z

My friend, this is why I have super-low expectations with all new lights. I rarely buy anything new. If it doesn’t have a way to disable a step-down, I don’t buy it.

It’s not like the old days where an XM-L U2 or XML2 could give 1,100 for an all-time max and not dim at all. Somehow, people started excepting that high lumens was not needed except in a limited-time, toy-like capacity. Crowd consciousness took over, and nearly everyone just went along, like a mob at a political protest or soccer game. Five-minute or less step-downs are now the norm.

That’s why my best and most used lights are Solarforce my M9 (in shorty mode) and the S1100. They aren’t the latest and greatest, but they will throw with the best of them and not have to be nursed back to full capacity. All this hype about dome-less emitters and extra lumens doesn’t do anything for heat sag or the eye’s ability to perceive illumination over great distances.

If I can’t run the mile in 5 minutes, but I say: “Oh I can run fast enough to theoretically make the 5-minute mile, but I have to stop and walk and cool off after three minutes, and then go again,” that’s NOT a 5-minute mile. And yet, here we are buying lights with “6,000 lumens” claimed that can’t maintain their lumens (and it even says so right on the packages most of the time, and we still buy them — some of us).

I’d rather take less for longer, but people have been bamboozled and don’t care.

In short, we been had.

Enderman · 2018-03-06T06:30:48.000Z

Same, I design all my flashlights to run at 100% with no step down until the batteries are completely drained.
This requires very good cooling, as well as a buck driver to avoid lack of battery voltage for the LED.

Agro · 2018-03-06T06:35:12.000Z

Enderman:

Watercooling in a flashlight is pretty useless because

water just transfers heat, it does not cool, so you would still need a radiator

a pump not only takes up room but also adds an additional point of failure

watercooling also increases the risk of failure due to a leak

the size of a flashlight is fairly limited so there really isn’t a need to carry heat far away, which is the main point of watercooling

another main point of watercooling is being able to use much larger radiators than a heatsink that can be stuck on a CPU. Again the size of the flashlight is a limitation unless you want to carry around an external radiator box with tubing running to your flashlight

With size being the limitation you can get equal or better performance than watercooling by using a large heatsink or heatsink+fan.
There are lights that can do several thousand lumens without dropping down due to a larger heatsink.

It is definitely possible to get 30k+ lumens using a COB LED and a large CPU heatsink with fan.
The only problem is that few people would want to spend $500+ on a large flashlight that only gets 5 minutes of total runtime due to power draw.

Said a man in the process of building a water cooled flashlight.
Yes, WC makes sense for recoil as you can hardly find another way to cool the LED (well, OK, there’s another option). Otherwise, not really.

danallen · 2018-03-06T21:09:15.000Z

mrheosuper:

bad news for you: you lost the bet
for example, look at acebeam x80, a soda can light, now how can you add pump, fan, water to such a small light, even if you can, where is the rad to cooldown water another

My favorite posts always are the posts that come across as if the person writing feels it is safe to assume that they are responding to an uninformed, unintelligent person. Thank you contributing a post like that.

I was thinking the light would have to larger, to accomodate the coolant, pump and fan. Here is the logic sequence that led me to that expectation:

the x80 and any other high output light has no space available for adding significant new components. Exception c omponents might fit in the driver board if the are small electronic components

cooling a high powered light enough to extend max output runtime cannot be done with electronic components small enough to fit on the driver board

cooling is going to require added components

so, basically, the light is going to be bigger

#I agree with you 100%, neither Lumentop nor any other manufacturer can make a bigger light the same size as a smaller one.

Since the light will be bigger, I was thinking of one way it might work. My understanding from my friends who are engineers is the first idea for a design nearly always is perfect and can be used as-is, with no changes. Oh yea, I just remembered, the first idea always has to be changed or abandoned, so the first idea always is a waste of time, and should be ignored, while we wait for the valid ideas that will show up later.

Since I enjoy insulting responses, I am posting my useless design idea, looking forward to the condescending responses to pass time while I wait for the good ideas coming later.

If you start with the x80, then add a section to the battery tube, similar to the BLF GT, except instead of adding four more batteries, the 2ND section of the battery tube could accomodate the cooling system. About 1/4th extra section would be for the pump, 1/4th for the fan, the rest for the radiator. It might be possible to skip the pump, if a heat pipe using capillary action for moving the water from the radiator to the hot surface where the leds are mounted. At least two ducts carrying water would fit in the battery holder or on the outside the battery tube.

I don’t know the cooling power or cooling requirement. This thread provides data from a working prototype. It shows the cooling capacity of a system like I just described is not a universe away from the cooling required to keep the x80 running on max Turbo as long as sufficient current at the operating voltage is available.

kiriba-ru · 2018-03-06T06:57:03.000Z

ZozzV6:

Big head, big tailcap with fins and heatpipes integrated to the tube to transfer heat from head to tail. On big lights like L6 the tailcap side always stay cool. Let it do the heat shredding too.

I dont own L6 but almost sure that you are not right. With direct drive mod, turn it on and put at the table for 10-15 minutes. If tail stays cool when you hold it in hands, this is because your hands can absorb 5-7W of heat.
Heatpipes wont work well in such conditions. 300W is 3-5 times more than modern CPU TDP. Those CPU heatsinks with 6-8 heatpipes have 8000-12000 sq.cm heatsink area. This is 30-40 times more that Q8 have.

mrheosuper · 2018-03-06T06:57:33.000Z

the problem is not about how you transfer heat, but how fast you dissipate heat
using water can help you transfer heat to other place faster, but if you dont cool down water, then the hot water keep coming to the led, and that’s not good
if you want to transfer heat fast and using small room, use Heatpipe instead, it’s also more solid

about the acebeam x80, what acebeam was trying to do is making a little monster in your palm, adding another battery tube will kill its point ai firsst place
people will be more impressed with the small one

EasyB · 2018-03-06T12:27:29.000Z

Enderman:

EasyB:

I think that’s a neat idea. Based just on the enthalpy of evaporation, 100mL of water has about 3800 W*min of cooling capacity. So it could remove, for example, 300W of heat for 12.5 min, or 50W for 76 min. Of course you would need forced air to effectively evaporate the water.

To boil water the LED would need to be much higher than 100*C which is horrible.
It is possible to do without boiling the water, by spraying it into mist and cooling it with air.

Ha, I wasn’t thinking of just boiling the water. Something as simple as spraying water droplets on the heatsink and blowing air over it would help significantly I think. It wouldn’t take anything big like the video you posted.

dchomak · 2018-03-06T13:56:53.000Z

That’s true.
Evaporation of 1 gram of water will carry away 5 times the energy than heating that same gram of water from freezing all the way up to boiling!

danallen · 2018-03-06T21:42:48.000Z

SKV89:

For active cooling, I’m really looking forward to the Astrolux MF05 and Acebeam X70.

SKV89, I did not know about these models until following through on your post. Thank you for letting me know about these. If I had known about them, I would have used a different title for this thread.

kiriba-ru:

Water is not cooler. It just transfers heat to external heatsink. Which usually have to be equipped with fan.
Both new elements (water pump and fan) were widely tested with big stationary lights and both were not able to work long in real life. Flashlight working conditions are ways more hard. Shockproof waterproof fan can cost more than Q8.
30000+ lumens=300+W. Four 30q cells=38Wh. (38/300)×60=7.6 minutes.

kiriba-ru, For a given size cooling system (lengthxwidthxheight) liquid cooling can provide more cooling than air-cooled only. The higher performance is not automatic and it comes with a cost.

jch10400:

Personally, I use flashlights every day in my occupation and I could care less if there is a 30,000 plus lumen water cooled light and don’t consider this to be in the top ten flashlight problems . If you really feel there is a demand for this type of light, I’d suggest you design one and put it on the market.

jch10400, I have no use for a flashlight in my profession. I think bright hand-held lights are awesome to the point of producing a spiritual experience. I like mine as bright as possible. It is a hobby and nothing else to me.

Regarding design, formulating a conceptual design has become one of the main purposes of this thread. For the sake of making a point, imagine a practical, complete, detailed design existed and was in my hands to do what I wanted with it. I would not be able to develop a product like this or put it on the market, because those activities are beyond me. My objective is find out if enough interest might exist among enough people to justify developing a liquid cooled hand-held led light with as much brightness as any handheld light available, but with the ability to run at that brightness level continuously, as long as power is available (i.e. batteries have not drained). If enough people would buy it, there are people here who can develop the product and bring it to market. My hope is that is what will happen.

dchomak:

The number one problem is runtime?

Why is it called flashlight?

“The first trustworthy lighting device was the flashlight, invented about 1896. Portable electric lights were called ”flash lights” since they would not give a long steady stream of light.”

dchomak, Interesting background. Today, I think the key feature is not the flash effect. I think the defining feature of a modern flashlight (or torch) is the whole unit is small enough to hold in one hand, and that it is not a lantern.

Newlumen:

Not a problem for me… we use the highest mode for like 2 minutes to check for the large surrounding area… then we step down to medium or 70% mode. Runtime ? No problem. 8x 18650 will do the job.

Newlumen, What is the maximum size of the large surrounding area searched with highest mode? Is there ever a need to check an area larger than that? In other words, how is it that the maximum surrounding area happens to correlate exactly with the the space that can be checked in two minutes using a light of… how many lumens in your search light? Could it be your search parameters are being dictated by the capability of your lights?

Have any of your searches succeeding in showing the target was not in the area you searched? If so, do you have proof the target was not in the area that would have been searched had the max mode of your lights been available for 3 minutes instead of two?

Unless your searches do not result in damage or loss to any one, or your searches have a 100% rate of finding your targets, I suspect your organization’s attorney will advise you not to answer these questions.

JeffroGymnast:

A water-cooled light seems way too impractical. It would add a ton of weight and many (comparably) fragile parts that wouldn’t survive a good drop (pump and fan especially).

I think a good solution to the heat problem is to use heat pipes, as Zozz already mentioned. You can get rid of a lot of heat if you transfer it in a spot that you aren’t touching. A well sized heat pipe going straight down the center of a soda-can light (in between the batteries) could lead to a big fin array inside or on the tailcap where your hands won’t be holding it. This could also reduce the outer temperature of the light at a given power level, as you could isolate the shell from the driver/LEDs instead of thermally linking them. If you really want to shed even more heat, a fan inside the tailcap blowing on the fins would help and I guess it would be easy enough to replace there if it broken during a drop.

JeffroGymnast, I think this is a great design that might be adequate and certainly cheaper than liquid cooling. Are any data or working prototypes available to show whether this design would provide enough cooling to keep a 50,000 lumen light cool enough to run continuously? The tradeoffs between liquid cooling and air-cooling with a heat pipe mostly are around physical size and cost. When size constraints are an issue, liquid cooling can perform better, but it cost more. If air cooling can get it done, then liquid cooling is waste of time and money.

Enderman:

Firelight2:

You could make a water-cooled light where water actually IS the cooler. :sunglasses:

Think of those battery powered fans with water bottles with pump attached. You could build a flashlight with one of those installed. Fill the bottle up with water, then as you run your flashlight, give the pump a squeeze to spray water aerosol all over your heatsink. As the water evaporates it should pull heat out of the heatsink. Repeat as necessary (or until you run out of water).

Air is still the fluid which cools the water.
Also that’s called evaporative cooling not water cooling for that exact reason.

Ironically, air cooled systems often use heat pipes that employ evaporative cooling inside a closed system.
Liquid cooled systems with a pump for the liquid can remove more heat than air-cooled heat pipes, if both systems have to be the same size.

Enderman:

Watercooling in a flashlight is pretty useless because

water just transfers heat, it does not cool, so you would still need a radiator

a pump not only takes up room but also adds an additional point of failure

watercooling also increases the risk of failure due to a leak

the size of a flashlight is fairly limited so there really isn’t a need to carry heat far away, which is the main point of watercooling

another main point of watercooling is being able to use much larger radiators than a heatsink that can be stuck on a CPU. Again the size of the flashlight is a limitation unless you want to carry around an external radiator box with tubing running to your flashlight

With size being the limitation you can get equal or better performance than watercooling by using a large heatsink or heatsink+fan.
There are lights that can do several thousand lumens without dropping down due to a larger heatsink.

It is definitely possible to get 30k+ lumens using a COB LED and a large CPU heatsink with fan.
The only problem is that few people would want to spend $500+ on a large flashlight that only gets 5 minutes of total runtime due to power draw.

Enderman, The reason liquid cooling systems are used for cooling CPUs instead of air-cooled is they are more powerful when space is constrained. The liquid cooling is more expensive than air cooling, but in cases where air cooling cannot do the job, liquid cooling is used. In flashlights, space is constrained. Why would that lead to liquid cooling being impractical in a flashlight but not in a computer?

pommie:

Even if you could design a flashlight that can keep cool at that kind of power, how about runtime?
For 30,000+ lumen you could use 4 XHP-70-2 at 12.5A each (8000lmn x 4) ~6.85v.
Using IMR35000mah cells in 2s5p config you would get 15 minute runtime, probably less depends on voltage drop.
Sure you could use more cells, but I would call 4s5p a portable light with a backpack power supply, not a flashlight.

Cheers David

Edit :- for those that want to work out the cooling required that is 342.5w
. . . . vvvv

pommie, I agree. Flashlights are handheld, without a backpack or any other supplemental components. That requirement sets limit on how much light can be produced and it sets a limit on how much light can be produced with cooled LEDs. A huge tradeoff is lumens for cooling. The big plus for cooling is you get more lumens longer, that is until you run out of battery power. I don’t have a perfect picture of the domain of where these tradeoffs land. Regarding costs, I figure a liquid cooled light put together by BLF is going to be the most cost effective implementation possible. That is all I am aiming for. Regarding run time, I guess it will be necessary to have extra sets of batteries to use while some are charging.

kramer5150:

I find that once my eyes dilate for night darkness (even just a little) I never use anything over ~350 Lumens in a constant-on fashion.

I like walking around my neighborhood lighting up people’s houses and other objects of interest. 350 lumens is pretty boring for that game.

Rusty Joe:

Crowd consciousness took over, and nearly everyone just went along, like a mob at a political protest or soccer game. Five-minute or less step-downs are now the norm.

That’s why my best and most used lights are Solarforce my M9 (in shorty mode) and the S1100. They aren’t the latest and greatest, but they will throw with the best of them and not have to be nursed back to full capacity.

kramer5150, I think you are onto something. It seems clear that intelligence, which is present in abundance, is no protection from mob mentality. Seems no matter how smart, people incline to circle jerk instead of thinking original thoughts, driven by one’s own observations, not just reports of others’ observations. There are exceptions, but not nearly as many I expect from a group like this one.

Enderman:

Same, I design all my flashlights to run at 100% with no step down until the batteries are completely drained.
This requires very good cooling, as well as a buck driver to avoid lack of battery voltage for the LED.

Enderman, Very sensible and solid, but then you miss out on the brightest possible light. My goal is bright and long run time.

Yokiamy · 2018-03-06T21:43:50.000Z

Rusty Joe:

My friend, this is why I have super-low expectations with all new lights. I rarely buy anything new. If it doesn’t have a way to disable a step-down, I don’t buy it.

It’s not like the old days where an XM-L U2 or XML2 could give 1,100 for an all-time max and not dim at all. Somehow, people started excepting that high lumens was not needed except in a limited-time, toy-like capacity. Crowd consciousness took over, and nearly everyone just went along, like a mob at a political protest or soccer game. Five-minute or less step-downs are now the norm.

That’s why my best and most used lights are Solarforce my M9 (in shorty mode) and the S1100. They aren’t the latest and greatest, but they will throw with the best of them and not have to be nursed back to full capacity. All this hype about dome-less emitters and extra lumens doesn’t do anything for heat sag or the eye’s ability to perceive illumination over great distances.

If I can’t run the mile in 5 minutes, but I say: “Oh I can run fast enough to theoretically make the 5-minute mile, but I have to stop and walk and cool off after three minutes, and then go again,” that’s NOT a 5-minute mile. And yet, here we are buying lights with “6,000 lumens” claimed that can’t maintain their lumens (and it even says so right on the packages most of the time, and we still buy them — some of us).

I’d rather take less for longer, but people have been bamboozled and don’t care.

In short, we been had.

+1

I really like your thoughts, it looks a bit like the megapixel race on small camera’s and phones from a couple of years ago.
I saw an ad on Fb from the Acebeam X70 which claims 40.000 Lumens on 8x18650 cells.
Everyone just instantly says, I WANT ONE, are they stupid?

I don’t get it, how long do they think it is able to achieve that? It probably needs new cells every 10 minutes :person_facepalming:

EasyB · 2018-03-06T21:42:08.000Z

danallen, are you interested in active cooled lights in general, or just water cooled lights? As others have said, water cooling is generally used to move heat, not to dissipate it. A fan blowing through a radiator or heatsink is the real mechanism for dissipating a lot of heat in a smaller area. In a flashlight you can (usually) design it so the heatsink is close to the heat source, so a water cooling system is not needed.

The_Driver1 · 2018-03-06T21:57:48.000Z

danallen:

Enderman:

Watercooling in a flashlight is pretty useless because

water just transfers heat, it does not cool, so you would still need a radiator

a pump not only takes up room but also adds an additional point of failure

watercooling also increases the risk of failure due to a leak

the size of a flashlight is fairly limited so there really isn’t a need to carry heat far away, which is the main point of watercooling

another main point of watercooling is being able to use much larger radiators than a heatsink that can be stuck on a CPU. Again the size of the flashlight is a limitation unless you want to carry around an external radiator box with tubing running to your flashlight

With size being the limitation you can get equal or better performance than watercooling by using a large heatsink or heatsink+fan.
There are lights that can do several thousand lumens without dropping down due to a larger heatsink.

It is definitely possible to get 30k+ lumens using a COB LED and a large CPU heatsink with fan.
The only problem is that few people would want to spend $500+ on a large flashlight that only gets 5 minutes of total runtime due to power draw.

Enderman, The reason liquid cooling systems are used for cooling CPUs instead of air-cooled is they are more powerful when space is constrained. The liquid cooling is more expensive than air cooling, but in cases where air cooling cannot do the job, liquid cooling is used. In flashlights, space is constrained. Why would that lead to liquid cooling being impractical in a flashlight but not in a computer?

A PC is stationary and can consume any amount of power that you wish for. Sensible laptops don’t use watercooling. So why should a flashlight which runs off of batteries use this? PCs don’t have to be robust, but flashlights do. Also you have a cooling loop which just transfers the heat from point A to point B. You don’t need this in a flashlight, it would just add unnessacary wheight. Also, water cooling adds additional points of failure, not good in a flashlight.

danallen · 2018-03-06T22:11:28.000Z

Throw me in to darkness:

danallen:

I like walking around my neighborhood lighting up people’s houses

So basically intentionally annoying people? I would set my 80 pound Shepherd/Husky cross on anyone who did that. Well, not literally, but she sure as hell would taken outside to show her huge teeth in anger.

I have not got a response from anyone except the people walking with me. When I light up a house, it is for three seconds or less. I don’t light up the same house more than once every few weeks. With all the cars driving our city streets, I doubt what I am doing is noticeable from inside the houses or if it is, it is just a momentary bright light, nothing much to notice in this busy city.

If someone inside a house responds with, “what was that” and looks out the window, I doubt it is any big deal. Is there anyone who is annoyed by this? Let’s say there is. I find sirens 10 blocks away as they ramble to the hospital 150 times a day annoying. Why does my otherwise quiet street need to have that sound coming here? We are not on the route ambulances use to get to the hospital. We are half a mile off the route. Why aren’t the sirens directional enough to prevent this spill? No one seems to care about that.

What I am trying to say is the upset you are asking about it too minor to worry about. You disagree? That’s fine, start another thread, it is off topic from this one. There are plenty of things I light up around here besides houses. Things like office buildings (empty at night), highway overpasses, billboards, dark sidewalks, allies between streets, ant trails and cockroaches. I like using a bright light for this stuff. I mentioned my use of brighter lights in response to someone suggesting 350 lumens is enough to do anything a soul might have reason to do with a flashlight. If that were true, there would be no need for a liquid cooled light. All of that is on topic. Someone’s feeling about my use of flashlights is not on topic.

danallen · 2018-03-18T01:27:49.000Z

ThrowMeInToDarkness:

I didn’t initiate the house lighting angle.
It’s clearly done with some kind of weird intent to get a reaction from the house occupier.

That is not true. I don’t want a reaction from anyone inside. I just think it is amazing seeing what can be lit up with this amazing light is all. There are more houses than anything around here so that is what I light up.

Enderman · 2018-03-18T02:35:37.000Z

Agro:

Said a man in the process of building a water cooled flashlight.
Yes, WC makes sense for recoil as you can hardly find another way to cool the LED (well, OK, there’s another option). Otherwise, not really.

Haha yeah it’s a bit ironic.
Of course in this thread we are talking about regular tube or compact flashlights, my large searchlight is a bit of an exception

danallen:

Enderman, The reason liquid cooling systems are used for cooling CPUs instead of air-cooled is they are more powerful when space is constrained. The liquid cooling is more expensive than air cooling, but in cases where air cooling cannot do the job, liquid cooling is used. In flashlights, space is constrained. Why would that lead to liquid cooling being impractical in a flashlight but not in a computer?

Enderman, Very sensible and solid, but then you miss out on the brightest possible light. My goal is bright and long run time.

No.
Liquid cooling is used for CPUs because there is only so much mass that you can attach to the CPU mounting bracket before you risk damaging the motherboard.
Also the area for a heatsink to fit is limited due to other components around the CPU socket such as ram and GPUs.
By using liquid cooling you can move the heat dissipation area elsewhere in the case, where there is a lot more room allowing you to use multiple large radiators.
Of course this depends on the case’s radiator support.
If the case is compact, such as mitx with only one or two fan spots, there is literally 0 improvement by using liquid cooling instead of a heatsink.
The heat dissipation depends on the area only, not where the area is in relation to the heat-producing component.
Liquid cooling just allows you to move the area elsewhere, where there may be more space, allowing for better temps.

So no, liquid cooling is only better than heatsinks when you can actually have more space to place radiators.
Unless you want a giant radiator stuck to a flashlight, so you need to carry a backpack around to move the flashlight, liquid cooling isn’t really useful for flashlights.
.
.
If you want the brightest possible light and long runtime you will have to find a compromise in the middle, because LEDs have limited efficiency, so you cannot get both long runtime and tens of thousands of lumens at the same time unless you want to carry around a backpack full of batteries.

One of the lights I’m modding will be ~10k lumens and have a few hours of runtime on max, the batteries cost hundreds of dollars and weight like 15 lbs.
If you want stuff like this you have to build it yourself because no manufacturer is really building this kind of stuff yet.

EasyB · 2018-03-18T13:46:45.000Z

Enderman:

One of the lights I’m modding will be ~10k lumens and have a few hours of runtime on max, the batteries cost hundreds of dollars and weight like 15 lbs.
If you want stuff like this you have to build it yourself because no manufacturer is really building this kind of stuff yet.

Well, you could use cheaper and more energy-dense batteries. If you have a long runtime, you don’t need 70C high discharge turnigy graphene batteries.

Aleister · 2018-03-18T22:58:55.000Z

What about oil, couldn’t oil be used inside a - hollow - pill for cooling?

Enderman · 2018-03-19T00:35:57.000Z

EasyB:

Well, you could use cheaper and more energy-dense batteries. If you have a long runtime, you don’t need 70C high discharge turnigy graphene batteries.

These batteries need high current capacity for my other projects.

Aleister:

What about oil, couldn’t oil be used inside a - hollow - pill for cooling?

I don’t think you understand, simply adding stuff inside a flashlight makes 0 difference. The heat is dissipated by the outside surface area of the flashlight which contacts ambient air.
Oil will have worse conductivity than aluminum or copper making it even harder for the heat to move from the LED to the exterior of the flashlight.

arficus · 2018-03-19T00:59:56.000Z

>> I saw an ad on Fb from the Acebeam X70 which claims 40.000 Lumens on 8×18650 cells.
>> Everyone just instantly says, I WANT ONE, are they stupid?
>>
>> I don’t get it, how long do they think it is able to achieve that? It probably needs new cells every 10 minutes

Umm, look how popular D4 is…

arficus · 2018-03-19T01:22:48.000Z

Where do we typically have everyday experience with cooling issues similar to what we are discussing? Engines.
Where size and weight are an issue water cooling is out. There are virtually no water cooled airplanes.
Water cooling would work great for non-portable installations.
If you want handheld I think air is the only practical option.
I imagine something similar to a VW engine.
Imagine the outermost tube of flashlight enshrouds a “core tube” having numerous cooling fins (with effective heat paths) running lengthwise, a fan at the tail cap, and exhaust around the head.