ergotelis
Enlightened
damn i think i will get dissappointed! :huh:
@CKOD, excellent work, this shows exactly what i believe too. Also, to validate your results, rerun you test @1amp to see the difference there!
Why a good thermal path really matters
- Thread starter saabluster
- Start date
Help Support Candle Power Flashlight Forum
damn i think i will get dissappointed! :huh:
@CKOD, excellent work, this shows exactly what i believe too. Also, to validate your results, rerun you test @1amp to see the difference there!
MikeAusC
Enlightened
The best thermal compound has less than one-tenth the conductivity of aluminium - it works under a star only because it's microscopically thin, to fill in air gaps caused by surface irregularities.
Putting a thick bead of it around the edge means the path is very long, rendering it pretty useless.
I'll be doing a bit more testing, but for that I'm waiting for someone whos been posting a lot in this thread *cough cough* to fix something, and start making some parts *cough cough cough* ( 
)so I can compare that to the DX board, and maybe the cutter board if I feel like putting a LED back down on it.
)so I can compare that to the DX board, and maybe the cutter board if I feel like putting a LED back down on it.
I'll be doing a bit more testing, but for that I'm waiting for someone whos been posting a lot in this thread *cough cough* to fix something, and start making some parts *cough cough cough* (
:nana:
Hey sabluster,
I have bee dong some digging on this but I was wandering what your thoughts on it are.
The data sheet for XML says that max junction temp is 302F (150C). So what I am wandering is if you have determined a relative max temp for the surrounding heat sink. For example I put some XMLs on a copper plate and put a temp sensor in the center of the leds and powered it up. Do you have an idea what point I shut it down should be based on the temp of the heat sink at close proximity to the LEDs?
I have bee dong some digging on this but I was wandering what your thoughts on it are.
The data sheet for XML says that max junction temp is 302F (150C). So what I am wandering is if you have determined a relative max temp for the surrounding heat sink. For example I put some XMLs on a copper plate and put a temp sensor in the center of the leds and powered it up. Do you have an idea what point I shut it down should be based on the temp of the heat sink at close proximity to the LEDs?
It's not as easy as just specifying a set temp for the heatsink. It can be higher or lower depending on the specific properties of the heatsink and LED used. This data sheet should give you an understanding of the math involved.
Hey saab,
I am guessing you have also done similar tests with XRE. Could you give me an idea how far it really pays to drive an XRE? I have my LED soldered to a thick copper slug on a heat sink that is mag sized. I need a rough idea of how hard I can drive it and still see gains. I would like the output to be sustainable to at least 30 seconds. Thanks
I am guessing you have also done similar tests with XRE. Could you give me an idea how far it really pays to drive an XRE? I have my LED soldered to a thick copper slug on a heat sink that is mag sized. I need a rough idea of how hard I can drive it and still see gains. I would like the output to be sustainable to at least 30 seconds. Thanks
bigchelis
Flashlight Enthusiast
My Surefire L1 XR-E R2 by Milky is direct drive and pulls 2A off fresh AW 17670 and did 390 OTF lumens. Copper heatsink in L1 bezel and thermal glue paste of some type is what he used. I am sure LED to copper bonding would be better and Sott felt 2A was about perfect for XR-E R2 as long as it had copper. It gets warm quick too. My XR-E R2 was purchased at cutterelctronics and EZ100 type.
Do you guys happen to know if the R2s cutter is selling now are the EZ900s?
bigchelis
Flashlight Enthusiast
I called them and just about every reputable LED supplier and they all said no.
bigC
My Franken Mag Jr. with Cutter XRE R2 EZ1000 bare led, arctic silver glue-down under pressure to copper heat sink in mag C did 221,000Lux @ 1 meter using Dx 66mm Aspheric lens. it's 2.2Amp under direct drive using Panasonic 2900mah 18650.
Last edited:
Wow dude, 221K is right about where I need to be. I upped the current on my light last night from 1.4A to about 2.18A, it did give a bit of a gain I think. I dont have any equipment to test but I think even with the led soldered to the copper slung the Q5 I have on there starts to drop off a bit after 1-2 min. Nothing drastic, but a slight fall. I have also ordered a larger lens to.
FYI I started a group buy in the MP if you guys need anything from cutter right now. I will probly place the order this weekend.
FYI I started a group buy in the MP if you guys need anything from cutter right now. I will probly place the order this weekend.
AaronM
Enlightened
Is this ceramic stuff any good? Looks pretty neat....
http://www.ceramtec.com/index/products/ceramcool/heat-sink-design/
http://www.aavidthermalloy.com/product-group/interface/insulators/beryllium
http://www.americanberyllia.com/
http://www.ceramtec.com/index/products/ceramcool/heat-sink-design/
http://www.aavidthermalloy.com/product-group/interface/insulators/beryllium
http://www.americanberyllia.com/
Yeah that is awesome stuff. Just nothing that I know of made of it that would really be of any use to us.
AaronM
Enlightened
It's been about a year since I called them, but I think American Beryllia said they would make stars for around a buck a piece if a guy ordered two hundred.
EDIT: I don't think there is an insulating material between the metalized portion and the main body of the part as the main body is an electrical insulator already.
http://www.americanberyllia.com/manuf_capa.html
http://www.americanberyllia.com/services.html
Another EDIT: Heck, I don't believe my own memory on this one: A dollar each sounds too good.
Anyway, whatever the price was, it seemed fair considering it would be the best star ever.
EDIT: I don't think there is an insulating material between the metalized portion and the main body of the part as the main body is an electrical insulator already.
http://www.americanberyllia.com/manuf_capa.html
http://www.americanberyllia.com/services.html
Another EDIT: Heck, I don't believe my own memory on this one: A dollar each sounds too good.
Anyway, whatever the price was, it seemed fair considering it would be the best star ever.
Last edited:
bshanahan14rulz
Flashlight Enthusiast
BeO scares me.
AaronM
Enlightened
My glasses are BeCu...and I'm not dead...yet. :/
MikeAusC
Enlightened
Berrylium Oxide has the same thermal properties as Aluminium, so all you'd be gaining would be eliminating the electrical insulator between the copper track and the aluminium star - plus you'd have very brittle star !
For significant improvement, you'd use a copper star with the LED soldered directly to it.
For significant improvement, you'd use a copper star with the LED soldered directly to it.
AaronM
Enlightened
I think the BeO star WOULD be a significant improvement over a typical board.
...Of course you're right, solid copper should be even better, but how? Even the unit in the link below has a gap between the emitter and copper chunk that would need to be filled with solder.
http://www.led-tech.de/de/High-Powe...U2-auf-Kupferkernplatine-LT-1790_120_170.html
The ideal would be to photo-etch the copper so the thermal pad would stand proud a bit and be at the same level as the top of the other layers.
Anyway, this is all speculation and ideas I'd love to see put to the test on my part.
Sometimes an experiment will show unexpected results and I'm cool with that.
...Of course you're right, solid copper should be even better, but how? Even the unit in the link below has a gap between the emitter and copper chunk that would need to be filled with solder.
http://www.led-tech.de/de/High-Powe...U2-auf-Kupferkernplatine-LT-1790_120_170.html
The ideal would be to photo-etch the copper so the thermal pad would stand proud a bit and be at the same level as the top of the other layers.
Anyway, this is all speculation and ideas I'd love to see put to the test on my part.
Sometimes an experiment will show unexpected results and I'm cool with that.
