looking to build first flashlight

[burro breath]

Hi there. I am from laserpointerforums.com and have built a high power laser in the Cree c6 flashlight. I am wanting to build a very bright flashlight using the Cree c6 as the host. Now I do have access to aluminum nitride as a heatsink. I can custom machine the heatsink at my work. (We make heatsinks for middle defense out of the same material so its very thermally conductive) since this material is SO thermally conductive I hope to put a very high power led in the Cree c6 host. I was hoping to use the sst-90 or Cree XML. But was wondering if either of these are really the best options.

 

[moderator007]

If its the same DIY c6 from DX, I don't think you'll ever be able to control the amount of heat that a sst-90 will put out under full power. The xm-l might be controllable at 3 amps. A maglite would be a more suitable host for either led if you wish to drive them at full power. And the drivers for the sst-90 putting out 9 amps seem to be a bit on the large side. They will never fit the c6. 3 amp drivers will fit.
The xml would be my guess. Might as well get a U2 bin xml while your at it. What I would try would be the a U2 xml with the 8-amc7135 2800ma driver board. Good efficient driver for single cell 18650 lights. Shiningbeam has these in stock with 3 modes low,med and high. The main reason I would use this board is that the 7135 ic's used on this board each give 350ma of current. They are all wired in parallel. The current adds up 8-7135's 8x350ma = 2800ma. If its to much current and the light tries to overheat to quick. You can simply unsolder one 7135 to give you 2450ma output. If its still to much you can unsolder another 7135 to give you 2100ma output. If the c6 is able to cope with the heat from the 2800ma drive current then its possible to add more 7135's to the board piggy back style. Soldered on in parallel with a existing 7135. Ever 7135 you add will give you a additional 350ma. You will also need a donor board to strip the 7135's off of. Here's a thread with the c6 using a p7 and the 7135's piggy backed.
I have used a xpg r5 in that host before. And it got warm within minutes of being on. Not to hot to hold onto but definitly warm. Drive current was at 2 amps.
You will need a good battery able to give that kind of current. I'll suggest some of the rebranded panasonic batteries. Callie's Kustoms, Redilast or AW's.
One more thing, I stuck the c6 reflector over one of my xml maglites one time. And best I remeber the xml had a large hot spot compared to the xpg that was in the c6 or the maglite with the xml using a rebel reflector. So probably want make a good thrower. More of a flood with a large hotspot.

 

[yazovyet]

just wondering, but do you know what the thermal conductivity of alumina silacate is? i'm mostly wondering since everything i read about alumina silacate is that it is used as an insulation and has a very low thermal conductivity, unless i am reading about a very similarly named product that is the opposite

 

[burro breath]

Sorry I ment aluminum nitride.

 

[burro breath]

Also...is it even going to help to use an exotic material like aluminum nitride for a headsink or am I better off just using copper or aluminum for the heatsink?

 

[moderator007]

Copper will absorb the heat better but you have to be able to give off the heat. Aluminum gives of heat very well. The c6 doesn't have a lot of body to give off the heat. The heatsink will more or less just transfer the heat from the led to the body of the flashlight. It's up to the body of the light to get rid of it.

Aluminum nitride. From what I read the copper is better and only slightly better than aluminum. This how well it absorbs the heat, not give it off. Surface area is what your after. Mass does help also. The more mass the longer it takes to heat up.

Thermal Conductivity

Aluminum nitride 285 W/(m-k)

Aluminum 250 W/(m-k)

Copper 401 W/(m-k)

 

[burro breath]

i dont mean to be a pain in the butt, but how about Pyrolytic Graphite. i can easily get that as well and it has a conductivity of alot more. or am i just over complicating something that should remain simple. also the 8-7135's yo mentioned above....from what i have read they are pwm. is this correct? why are they so cheap compaired to say the shark drivers that i see on cpf that deliver the same amount of current? if im just asking noob questions please call me out on it and ill read even more then i have been on these forums. thanks,
burro

 

[moderator007]

why are they so cheap compaired to say the shark drivers that i see on cpf that deliver the same amount of current?

There are to many difference to list you can read up on the shark here. The shark board can do so many different things compared to a 7135 driver. The amc7135 board is very easy and cheap to produce. Not alot of R&D goes into these boards. The most expensive part is probably the attiny13a microcontroller. AMC7135 Datasheet here. One amc7135 gives you 350ma each. They are all wired in parallel. The current adds with every one you add. 8 x 350ma gives 2800ma out to the led. They are linear current regulators. They take the voltage over the led drive voltage and give of the excess voltage as heat. Sort of like a varible resistor. As the voltage decreases in the battery to the led drive voltage. Then it has to burn off less and less of the voltage. Until it becomes direct drive. The efficiency starts out good but as the battery voltage gets closer to the led drive voltage it becomes very efficient. Good thread here to read.
Yes if you want modes they use pwm for the modes. Lightmalls has one without modes if you prefer that Sku #m0123.
These boards can also have 4 different sets modes if you choose. Info here a 1/4 of the way down the page click on the driver picture. I have never purchased a 1 mode driver. So I dont know If it's modes can be changed. It may not have the attiny13a microcontroller.

I think you have missed the point I was trying to make. No matter how thermal conductive your heatsink is the light body still has to get rid of the heat. The heatsink is only the transferer. The better the thermal conductive it is the more it will transfer the heat to the flashlight body. Aluminum should be fine but copper is better. Pyrolytic graphite being superior at 600 to 800 W/(m-k). And don't forget to use some kind of thermal paste between heatsink and flashlight body. Artic silver being the most prefered.

 

[yazovyet]

i had read (from wiki) that aluminum's thermal conductivity was 237 (and from a different source, thermal of 180 for the 6061 T6 grade alloy).

another thing to think about is volumetric heat capacity (so much heat energy it takes to heat up a fixed volume). if you are planning ot run the light in relativly short bursts then you might not have to worry too much about radiating the heat out of the light if the heat sink will absorbe most of it (and the heat can slowly radiate out while the light is off)

the volumetric heat capacity of copper is 3.45, of Al is 2.42 (both from wiki), and of Al-n is (if i calculated it correctly) 2.41 (all in J/(C * cm^3))

`

 

[burro breath]

Basically I am looking to built this flashlight so I can have a hi Med and low setting so I can leave it on low and not worry about heat. And then when I need a short one or two minute burst of bright light I can have it. I'm looking for quality. Something that will last me and if a 50$ shark driver is what I need then I'm happy to but it. But if its way overkill then there's no.point in spending extra money. I will definately be doing some reading tonight and the following days.:)

 

[moderator007]

Basically I am looking to built this flashlight so I can have a hi Med and low setting so I can leave it on low and not worry about heat. And then when I need a short one or two minute burst of bright light I can have it. I'm looking for quality.:)

The light can probably handle short one to two minute burst. The amc7135 drivers are very dependable and efficient using a single li-ion 4.2v. The amc7135 driver can be modified to give you more current if you wish. Just piggyback more of them. Each gives 350ma. Run the current up until it gets to hot to hold in your short burst. Back the current off a little or leave it there. That will give all she can do.

i had read (from wiki) that aluminum's thermal conductivity was 237 (and from a different source, thermal of 180 for the 6061 T6 grade alloy).

I got that figure from here. Temperature does also play apart in thermal conductivity of materials.