DIY Cree Helmetlight- Mtbr.com
Results 1 to 29 of 29
  1. #1
    mtbr member
    Reputation: Gilbo's Avatar
    Join Date
    Nov 2006
    Posts
    60

    ... and if we just ... DIY Cree Helmetlight

    Today I made my first ride with my own build helmetlight "Dinottestyle" Using two cree XRE leds combined with a 4 and 9 degree lens and driven by a 700mAh powerpuck (because I'm still waiting for my buckpuck) The batterypack is also selfmade with 8 4000Mah NiMh babycells, all glued in a shrinktube. My first impression is This thing really kicks ass, in fact I think the rabbits will need some glasses
    Here are a few pictures of the end product;

    the very beginning, building a housing for 1 cree XRE led and Ledil Lens


    the light placed on a helmet.

    Full blast

    Beamshot in my garden, the tree is 10 meters away, the fence 15 meter, I also have a DIY 12v 20 watts halogen lamp and compared with that I think the output is about the same.
    Next week........more pictures to come

  2. #2
    mtbr member
    Reputation:
    Join Date
    Jan 2005
    Posts
    586

    nice

    Very professional looking.

    I have a question on thermal control. How are you dissipating the heat from the MCPCB (junction) to the outer case?

    I'm building a dual head XR-E body but I haven't settled on a thermal dissiaption method as yet.

    Has anyone ever taken the Dinotte apart to see how they are doing it?

    Bob

  3. #3
    Gone riding
    Reputation:
    Join Date
    Jan 2004
    Posts
    3,351
    Very nice work Gilbo, it must be handy having access to a lathe and milling machine like that. I can tell you’ve put some serious planning into this baby!

    You’re going to get some sweet run times with nice light output too judging by your setup. Do you think you will try making a cut out / rubber pad DiNotte style so that you can mount them on your bars?

    Do you have a shot from the rear to show how you have things mounted? It looks like you have a removable heat sink to mount the Cree to, and I assume the battery will be external and there will be enough room left in the rear to mount your smaller Buck Puck. Have you decided on a switch for it yet?

    More photos to come hey… they’d better be posted sooner then next week!

    Keep up the good work!



    Bob, just to clarify a few things, the “junction” you may read about in various datasheets is actually internal to the emitter and is on the bottom of the LED die, which sits between the clear lens and the ceramic substrate. The rectangular substrate is then reflow soldered on to the Star MCPCB.

    For a good thermal connection between the Star MCPCB and your heat sink you will need some form of thermal compound (thermal grease), as used for computer CPU heat sinks. Personally I use Arctic Silver thermal compound, but if you need to glue your Star, then Arctic Alumina will work wonders.

    Most of us who build LED based systems use the whole housing as a heat sink. You need to have a read of the Cree XR-E datasheet to ascertain the exact thermal requirements, but the Cree based projects posted here on MTBR so far should be a good indication to what size you will need. Just remember physical mass isn’t as important as surface area and the heat sinks ability to dissipate heat.

    I hope that helps.

    Dave.

  4. #4
    mtbr member
    Reputation:
    Join Date
    Jan 2005
    Posts
    586

    low_rider

    I understand.

    The heat is max at the junction and must be transported away if the temp is to remain within spec and allow the LED to operate without degradation. I'm going to use a rectangular aluminum tube (2"x1.25" with a wall thickness of .125") for the housing. This is a stock shape (no machining) and almost perfectly fits 2 LEDs with optics.

    My current concern is moving the heat from the back of the MCPCP to the outer case. The two surfaces are at 90 degrees to each other. I see dropping a right angle piece of aluminum down from the inside of the case housing and bolting it to the back of the MCPCB. My calculation shows at least 1 square inch in contact with the MCPCB and then a square inch contacting the inside surface of the housing. With a .125 thick housing I can mill some heat dissipating channels as well on the outside. The .125" thick housing wall also allow screw holes to be installed which will greatly simplify the end cap design.

    Thanks for letting me bounce this off you. Any input is appreciated.

    Maybe a thread on housing design would be worth while.

    Bob

  5. #5
    mtbr member
    Reputation: Gilbo's Avatar
    Join Date
    Nov 2006
    Posts
    60
    Quote Originally Posted by Low_Rider
    Very nice work Gilbo, Do you think you will try making a cut out / rubber pad DiNotte style so that you can mount them on your bars?
    .
    Thx for the reply, I already have made a groove in the bottom of the light to fit on a handlebar
    I'll try to make some more pictures today, but since whe have organised a MTB trialride this sunday I have lots off work at the moment.

  6. #6
    Gone riding
    Reputation:
    Join Date
    Jan 2004
    Posts
    3,351
    No problem, I couldn’t quite make out the cut out in the photos. I’m looking forward to your follow up.

    Cheers.

  7. #7
    mtbr member
    Reputation: Gilbo's Avatar
    Join Date
    Nov 2006
    Posts
    60
    Took some more pictures today, here is the image gallery http://gilbertosphotos.fotopic.net/c1156732.html
    I'm waiting for your comments!
    and by the way thanks to all the folk at this forum who gave me the inspiration off building this light.

  8. #8
    mtbr member
    Reputation:
    Join Date
    Jan 2005
    Posts
    586

    Question

    Looking at the picture with the heatsink it appears to be about 30 - 35 mm in dia. Is it touching the housing inside wall anywhere?

    How are you getting the heat from this assembly to the housing? Convection through the air?

    Bob

  9. #9
    mtbr member
    Reputation: Gilbo's Avatar
    Join Date
    Nov 2006
    Posts
    60
    The heatsinck/ledholder has the diameter of the inside of the housing minus 0.10mm so the heattransfer is garanteed. the light does not warm up at all, last wednesday Took it out for a testride and the light does not warm up at all (outside temp was 7°c)

  10. #10
    mtbr member
    Reputation:
    Join Date
    Jan 2005
    Posts
    586

    Gilbo

    Just so I'm clear on this. There is a gap - a small one - between the heat sink and the inside of the housing?

    Even a small gap will act as a relative thermal insulator between the housing and the heatsink. I say relative because the heat sink and the housing, being metal, are far better conductors of heat than air. However, you may not need to actually make contact to the housing as long as the junction temp remains in spec. They only way to tell would be to run the light for more than 20-30 minutes and measure any intensity drop. If none is detectable you're good to go. Your heatsink does appear to be robust. The design I'm working on would place two XR-Es in the same housing so I'll probably have to design a means to conduct the heat to the case efficiently while keeping the weight low.

    Bob

  11. #11
    mtbr member
    Reputation: Gilbo's Avatar
    Join Date
    Nov 2006
    Posts
    60
    Bob, I think when the heatsink gets hotter it also becomes larger in diameter so the very tiny gap will be as good as gone. If there should be a problem I can always put some thermal paste inside the housing, butt again, I don't think there will be a problem.

  12. #12
    mtbr member
    Reputation:
    Join Date
    Jan 2005
    Posts
    586

    Gilbo

    I agree. I don't think you'll have a problem either. There is a lot of metal there. Also, I sent my last post before looking at all of your pics. The one which shows the entire interior design answers my question. Your heatsink is connected to the case via the two rods that screw to the rear cap. The rear cap is touching the case - with significant surface area. I doubt if there would be a problem. You have considerable metal inside.

    What does a single head weigh?

    Bob

  13. #13
    mtbr member
    Reputation: Gilbo's Avatar
    Join Date
    Nov 2006
    Posts
    60

    Bob

    The two housings together weigh just 250 grams ( there is a difference between the two because one light holds the switch and connect cable for the other lamp) I did not use a very accurate scale so i'll measure the exact weight next week at my work.

  14. #14
    some kind of hero...
    Reputation: airman's Avatar
    Join Date
    Jan 2004
    Posts
    543
    Really nice handywork!

    Thanks for the photos and explanation...

    Keep up the good work.

    Cheers
    GEVELTERSCHMIDT RACING

  15. #15
    mtbr member
    Reputation:
    Join Date
    Nov 2005
    Posts
    500
    Nice job man! Good to see the final product.
    The Crees are certainly ruling the LED scene.... huh!

    Are you happy with your lens choices?

  16. #16
    mtbr member
    Reputation: Gilbo's Avatar
    Join Date
    Nov 2006
    Posts
    60
    Quote Originally Posted by achesalot
    Nice job man! Good to see the final product.
    The Crees are certainly ruling the LED scene.... huh!

    Are you happy with your lens choices?
    Thx Allen,
    Yes the 4 and 9 degree lens works together very well, I supposed that I would have a good amount of lightoutput, but this amount I had not expected, I can only say

    Now I'm planning to change the halogen bulb in my other light also in a tripple cree setup ( have to do some measuring before to see if my housing is big enough, otherwise i'll have to make a new one
    It's a petty that there is such a big distance between the forummembers, otherwise we could have a forumnightride. then you could see two things from space, the chinese wall and the MTBR nightriders

  17. #17
    mtbr member
    Reputation: Wombat's Avatar
    Join Date
    Jan 2004
    Posts
    2,165
    Quote Originally Posted by Gilbo
    Then you could see two things from space, the chinese wall and the MTBR nightriders
    Given the number of astonshingly bright lights that have been demonstrated here lately I think you might be correct.

    I think this is a great forum. I have enjoyed the inventiveness that posters show, and the enthusiasm that is demonstrated when talking about the latests project's potential to deliver something brighter, longer or cheaper is heart warming.

    Thanks to all the posters.

    Wombat

  18. #18
    mtbr member
    Reputation: ICanDigIt's Avatar
    Join Date
    Jun 2004
    Posts
    579
    Nice one Gilbo! Very solid looking...

    Shot through the heart
    And you're to blame
    You give love a bad name...

  19. #19
    Spanish biker
    Reputation: msxtr's Avatar
    Join Date
    Dec 2006
    Posts
    732
    Woooowwwww, is not seem home made, I like, excellent work

    Greetings - Saludos

    msxtr
    Warning!!! my english is very very bad, sorry.

    Easy DIY led light1
    Easy DIY led light2

    The Beast!!!

  20. #20
    Gone riding
    Reputation:
    Join Date
    Jan 2004
    Posts
    3,351
    Bob, I think it’s already been covered pretty well by Gilbo, but there is plenty of material there considering the drive current and the fact there’s only one emitter per housing. I can’t see there being any issues.

    Quote Originally Posted by Bob
    They only way to tell would be to run the light for more than 20-30 minutes and measure any intensity drop. If none is detectable you're good to go.
    While definitely measurable with quality equipment, any intensity drop caused by temperature would be barely noticeable with your eyes (if noticeable at all) with most emitters, even with a relatively huge temperature shift. Most emitter manufacturers agree that Vf shift as an indication of temperature change is a much superior method.

    For most DIY guys, the “touch” test will probably suffice if a little consideration has gone into the design to start with. There are a few other relatively easy methods for those a little sillier.

    Keep up the great work all!

    Dave.

  21. #21
    mtbr member
    Reputation: grawbass's Avatar
    Join Date
    Aug 2004
    Posts
    2,044
    There seems to be plenty of metal there, but if you do happen to have overheating problems, the color of the exterior of a heatsink has a huge effect on how much heat it can radiate.

    Silver is the second worse color behind white.

    Flat or brushed finishes are better than polished.

    The best is FLAT BLACK.

    You would think that the layer of paint would create some insulation, but the effect of the color far outweighs the layer of paint. A heatsink painted flat black will radiate heat much better than a non painted silver one. The best paint is automobile radiator paint. It goes on thin to lessen the insulation while still providing the black color.

    Also, cast aluminum is better than extruded or CNCd and copper is better than cast aluminum.

  22. #22
    Gone riding
    Reputation:
    Join Date
    Jan 2004
    Posts
    3,351
    While an extremely light coat of high temperature black paint will see some gains, a few turned grooves in the housing would make a far greater difference. As continuously mentioned, I don’t see that there will be any thermal issues whatsoever. The setup that Gilbo has designed and built is more then sufficient.

    Dave.

  23. #23
    mtbr member
    Reputation:
    Join Date
    Jan 2004
    Posts
    301
    The colour only makes a difference for radiated heat transfer (ie IR). At the temperatures LEDs run at (or at least the temperatures you should be running at with decent heatsinking), radiation is a very minor form of heat transfer - it is largely conduction/convection. Therefore whilst a coat of black paint improves the radiation, it will actually inhibit the conduction/convection (by providing an extra insulating layer between the heatsink and the air), so is likely to make the overall heat transfer worse.

  24. #24
    Homer
    Reputation: brum's Avatar
    Join Date
    Dec 2004
    Posts
    215
    Quote Originally Posted by chrism
    The colour only makes a difference for radiated heat transfer (ie IR). At the temperatures LEDs run at (or at least the temperatures you should be running at with decent heatsinking), radiation is a very minor form of heat transfer - it is largely conduction/convection. Therefore whilst a coat of black paint improves the radiation, it will actually inhibit the conduction/convection (by providing an extra insulating layer between the heatsink and the air), so is likely to make the overall heat transfer worse.
    Wrong. Check out CPF, theres a thread by member 'newbie' explaining the theory and conducting some experiments, very interesting.

  25. #25
    mtbr member
    Reputation:
    Join Date
    Jan 2004
    Posts
    301
    I apologise - was actually basing my observations on other stuff I've seen on CPF, but hadn't spotted that one by Newbie before. What I was neglecting of course is that at the moment these lights are polished, which is very bad for heat transfer.

    Having reread the thread, what is actually a bigger issue, and quite a big one IMHO is that gap between the heatsink the LED is attached to and the body of the lamp. Gilbo asserts that when the heatsink heats up it will expand and so close this gap - however what he's missing is that when the body heats up, the hole in the middle will also get bigger thus maintaining the gap. The only way for this gap to close due to heating is if there is a temperature difference between his heatsink plug and the body of the light, which is very definitely a bad thing, and implies poor heat transfer of itself. In fact I can't see any way in which he can get good heat transfer with it as it is - either there is a gap as machined, which implies poor heat transfer, or there is no gap due to differential expansion, which also implies poor heat transfer! Finally I note him commenting on the light not heating up being a good thing - wrong, that's actually a bad thing as it implies the heat is locked up inside the lamp and not getting to the surface. Certainly there is not enough surface area there for it to transfer all the heat to the air without noticeably heating up if all the heat was making it from the LED to the body (I have a Exposure Joystick, which is a commercial light of similar size with a single Lux 3, and that definitely gets warm in use).

  26. #26
    mtbr member
    Reputation: Gilbo's Avatar
    Join Date
    Nov 2006
    Posts
    60
    Quote Originally Posted by chrism
    Finally I note him commenting on the light not heating up being a good thing - wrong, that's actually a bad thing as it implies the heat is locked up inside the lamp and not getting to the surface.
    Since Aluminum has a very good heat transfer of it's own i'm almost certain that there is no problem whatsoever with my construction. I fact i'm just back from a ride and with an outside temperature of 0°c the housing is icecold ( there was a lot of fog today, I didn't see sh#t)

    Also the heatsink is also used to keep my light together as seen on the earlier pictures, so there is a physical contact between the sink and the housing!
    today I tested my light in a different setup, i've put it on my bike, works fine!


    here a picture of the light on my Bike

  27. #27
    mtbr member
    Reputation:
    Join Date
    Jan 2004
    Posts
    301
    Quote Originally Posted by Gilbo
    Since Aluminum has a very good heat transfer of it's own i'm almost certain that there is no problem whatsoever with my construction. I fact i'm just back from a ride and with an outside temperature of 0°c the housing is icecold ( there was a lot of fog today, I didn't see sh#t)

    Also the heatsink is also used to keep my light together as seen on the earlier pictures, so there is a physical contact between the sink and the housing!
    today I tested my light in a different setup, i've put it on my bike, works fine!


    here a picture of the light on my Bike
    Yes aluminium has good heat transfer, so the heat gets spread nicely round that body. However in order to get rid of the heat you have to transfer it from the alu to the air, and that mechanism is far less efficient. With a housing that size you really don't have enough surface area to transfer the heat well enough to keep it cool in normal use, unless that is your internal heat transfer is so poor that you have a nice big temperature gradient inside your casing. If your housing is ice cold you really should be worried!

    There may well be physical contact, but from what I can see it is only light physical contact, which means the heat transfer will be poor, since given the microscopic surface roughness actual physical contact area is proportional to the contact force, of which you have very little - conventionally heat conducting interfaces are either forced together with heatsink compound between or bonded with thermal adhesive.

  28. #28
    Gone riding
    Reputation:
    Join Date
    Jan 2004
    Posts
    3,351
    While I have been known to preach about thermal requirements, heat transfer and dissipation here on the boards in the past (as anyone who has been around here for a while will have painfully noticed); let’s face reality here for a minute. We are talking a difference of .1mm, which is roughly the thickness of an average piece of paper. I have seen “Star” bases with worse tolerances then that.

    On top of that it’s only a single emitter, and even though it is being driven reasonably hard between 700 and 1000mA, it’s not going to be producing a great deal of heat anyway. I don’t dispute that thermally it may not be an ideal solution, but it’s certainly not going to be detrimental to the emitter in the short term. Perhaps Gilbo could turn a new heat sink with a slightly tighter tolerance for a better fit with the housing, or run a ring of thermal tape or paste around the outside to improve things a little if he really wanted. But by the time lumen maintenance becomes an issue with this design the Cree XR-E will be ancient history anyway.

    I have seen many “drop in” style Cree XR-E replacements on CPF that are far worse in design then Gilbo’s neat solution, and some of these have been in operation for months, and driven harder then 1000mA without noticeable degradation so far.

    Being picky, the Cree XR-E emitter is a terribly poor thermal design to start with. The 7090 package is a ceramic substrate with the die placed on top, and a lot of distributors are reflow soldering the substrate to a “Star”, which depending on the distributor is either a piece of fibreglass with a couple of bits of copper layered on the top of it, or in Gilbo’s case a piece of aluminium with an insulating pad placed on top. All up you’ve got quite a large amount of thermal resistance before you even think about attaching them to a heat sink to dissipate the heat.

    I’m actually quite interested to see Seoul Semiconductors implementation of the Cree XR-E die in the Z-Led P4 series emitter. Being the same shape and package design as the early Lumiled Luxeon range it should be a lot more thermally efficient, and easier to use for those without reflow soldering capabilities. Hopefully it will be available early next year.

    Dave.

  29. #29
    Spanish biker
    Reputation: msxtr's Avatar
    Join Date
    Dec 2006
    Posts
    732
    Quote Originally Posted by Gilbo
    Since Aluminum has a very good heat transfer of it's own i'm almost certain that there is no problem whatsoever with my construction. I fact i'm just back from a ride and with an outside temperature of 0°c the housing is icecold ( there was a lot of fog today, I didn't see sh#t)

    Also the heatsink is also used to keep my light together as seen on the earlier pictures, so there is a physical contact between the sink and the housing!
    today I tested my light in a different setup, i've put it on my bike, works fine!


    here a picture of the light on my Bike
    Have you a Orbea?? incredible It's Spanish bike.

    Greetings - Saludos

    msxtr
    Warning!!! my english is very very bad, sorry.

    Easy DIY led light1
    Easy DIY led light2

    The Beast!!!

Members who have read this thread: 0

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

THE SITE

ABOUT MTBR

VISIT US AT

© Copyright 2020 VerticalScope Inc. All rights reserved.