Just started a dental flashlight project

[constantine_a_f]

After reading several threads

http://www.candlepowerforums.com/vb/...dental-student
http://www.candlepowerforums.com/vb/...for-DENTAL-use
http://www.candlepowerforums.com/vb/...ental-headlamp

I decided to give it a try and manufacture my own dental headlamp. I made a rough sketch using acad of what parts I plan on using/manufacturing.

I will use a Rebel White or Warm White LED mounted on a pcb star, 8 degree angle carclo optics and an 1 inch aluminum tube body + possibly a 19mm round X 20 mm high heatsink (the parallel lines, you will see as soon as I upload the drawing). Do you believe I might need a heatsink that big, or I could use something more compact. I was aiming at a 70-80 sqmm aluminium 6060 surface (12sqin). I will power the led with 8 AA NiMH batteries through a 700mA constant current regulator, connected with a dimmer. I am a dental student so please excuse any mistakes I make with the electronics or acad. I will post the acad sketch as soon as I am allowed by the forum administrator.

[hellokitty[hk]]

I would use a high CRI XP-G.

[constantine_a_f]

Why would you use a XP-G? Could you please inform about the benefits? Thanks

You can download the sketch from here

http://www.2shared.com/document/H_zM...ng1-Model.html

[MikeAusC]

Unlike others here, who have received extensive help and advice with building specialised lamps, you will hopefully give feedback on your final result, rather than just disappearing.

If you promise to give someone a sample of your lamp in return for providing detailed help, you will hopefully do so.

[Harold_B]

Is this for personal use or are looking to market the design as a product? Design parameters would be very different for the two I would think.

[constantine_a_f]

I am a dental student at Aristotle University of Thessaloniki, Greece. I do plan on using the lamp for my clinical and lab work. I will also make it sure is relatively water safe so that I can use surface dissinfectors. I wanted to use a Rebel because the angle when it is combined with a Carclo lens is 8-9 degrees. I will post images of the whole project online (I hope I will be able to post shots by then, or else I will use 2shared.com). As I live in Greece and I will buy the parts from US it will take some initial time to start posting pictures but I will do so. I also plan on giving detailed feedback (with beamshots and reports on the led efficiency and usability) from time to time. I cannot give someone my lamp as I am on tight budget and cannot make 2 of them (though I will buy twice some of the parts). If I make more of them I plan to sent them to anyone who asks for them (classmates, forum members) for the parts cost (around 70 USD). I will use a CNC lathe to make the aluminum body out of a 1" al.6060 or 6061 tube with 2mm thick wall. Could you please share your ideas with me and I will not dissapear. I would also like to say that providing me detailed help could make this project easier for me, but I will make it happen easy or not, with help or not, because it is something I need for my job.

Thanks for all the replies. I will be checking the thread many times daily, answering questions and reading all the valuable info you send me,

Constantine

[Harold_B]

Sounds like a great project Constantine. I have been involved with the design of a few dental/medical illuminators and each has its own requirements. The suggestion to use a high CRI LED is a good one no matter which one you choose. Judging the health of tissue under LED lighting is difficult if the CRI is low. You have stated the Carlco Lens is 8 - 9 degrees. Is that half angle or full and is that at half max? What color temperature are you looking for? The color separation on the Rebel is excellent and would work well with a TIR optic but there might be other choices for higher CRI.

Let me know the part number of the Carlco optic you are looking at using. I might have it in the component library within my software and would be able to run a quick simulation with a Rebel in a day or two (if you are interested).

Harold

[hellokitty[hk]]

I'd think that the high CRI would be beneficial, that way things won't look washed out blue. Also a warmish tint too.
Also, AFAIK, XP-G's are going to be in a smaller package with higher efficiency than a rebel. Illumination supply has high CRI XP-G's available for $7.00.
XR-E's have a bit of a tighter beam pattern and come in warm, but not high CRI's.

I'd love to hear results (preferably with a pictorial review/build log) if possible .

[ledpwr]

Have you looked at the lxm3-pw51 rebel LED? It has a relatively high CRI of 85 and has a very nice neutral white tint of 4000k, it will work with the optic to produce an 8 degree beam and can be used on the same pcb board, its only slight downside is it is slightly less efficiant than the best current LED but you will still get over 180 lumens at 1 amp drive current.

[constantine_a_f]

First of all I want to thank you all for the support. As I try to keep the budget down and the quality up, I will buy all the parts from one place (shipping is around 10 USD for every time I don't combine). I want a close to daylight cool white 6500K led. The reason I chose this one is because I currently work with this temperature and I find it very good as a mean of distinguishing pathological conditions. I will not do any color matching for restorative procedures under any type of artificial illumination. I use broad daylight for that (sometimes I even call my patients twice or three times in one day for that). I came down to this list from this shop because I find their drivers small, neat and uncomplicated. I think that 120-200 lumens is very high but again the standard of care is that I protect the patients eyes with protective glasses and I don't plan on powering the led that high. This is the list of the parts:

http://www.luxeonstar.com/Cool-White...s.htm?CartID=1
http://www.luxeonstar.com/Carclo-8-7...3.htm?CartID=2
http://www.luxeonstar.com/Harness-fo...p.htm?CartID=3
http://www.luxeonstar.com/19mm-Squar...b.htm?CartID=5
http://www.luxeonstar.com/Pre-Cut-Th...2.htm?CartID=4
http://www.luxeonstar.com/Carclo-20m...5.htm?CartID=9
http://www.luxeonstar.com/700mA-Ext-....htm?CartID=10

I calculated that I need a heat sink with around 20 C/W thermal resistance if I power the led at 700 mA (which is not the target as I said before). I am thinking of either incorporating a heat sink as the above or actually mounting a solid copper base at the end of the tubing. Harold, I would greatly appreciate if you run the simulation. Again thank you all.

PS I will post pictures in a DIY build log way. If you have some counter suggestions about the shop, led, driver,etc they are all welcome

[srfreddy]

6500k will wash all the colors out. You want a high CRI light.

[constantine_a_f]

6500k will wash all the colors out. You want a high CRI light.

http://www.heine.com/eng_US/PRODUCTS...D-LoupeLight-R

this is the current market leader at least here in Greece (6250 K)

and some others

http://store.orascoptic.com/apollo-l...able-p159.aspx

http://www.designsforvision.com/DentHtml/D-LEDDay.htm

(5500-6000K)

I thought that temperature describes the color appearance of the light source and the light emitted from it while color rendering describes how well the light renders colors in various objects. I understand they are connected but the rebel light has a cri of over 70 (I thing it is good, not excellent but good).

Again I will not use it for color matching but rather for illumination of the surgical field.

[MikeAusC]

Constantine, have you considered a Heatpipe - you may be the first to use this technology for a Dental Lamp !

It seems the ideal solution where you need to keep the heatsink both light and small, yet the metal parts need to be kept cool enough to avoid injury. A hollow copper tube can provide the conductivity of solid copper with many times the weight of a heatpipe.

You want to keep the lighthead small, but that makes it poor as a heat dissipator. You can bend heat pipes easily, as they are annealed copper, but the minimum radius is limited.

I'm suggesting using the Heatpipe itself as the main heatsink because of your low power - conventionally the Heatpipe takes the power to a bigger heatsink with minimal temperature drop. You could run the tube along the harness, or you could form it into a circle near the Lamp.

You must select a Heatpipe that has a braided wick inside, as this is the only type that will work well at any angle.

Heatpipes are available from RS Components or Element14 (Farnell).

I'm currently building a 25 watt Headlamp that uses a Heatpipe to provide lightweight fanless cooling.

[MikeAusC]

6500k will wash all the colors out. You want a high CRI light.

Colour Temp and CRI are separate measures.

6500 is sunlight and no-one complains that sunlight has a low CRI.

LEDs don't have the smooth spectral response of sunlight and generally the cooler LEDs have a worse spectrum, so people associate a high CT with poor CRI.

High-CRI LEDs generally have a mid-range CT

[constantine_a_f]

Mike I did consider heat pipe as a way to cool the led, but I didn't find anything small enough. From what I understand you cannot overbend or cut a heat pipe. I cannot fit the minimum 100mm long heatpipe anywhere, nor can I bend it so much. If you have a solution for that I will give it a try. The design I posted above has enough heatsinking abilities even without a heatsink, just with the aluminum body (not that it would be cold or something).

[hellokitty[hk]]

Here are some beamshots of a high CRI vs cool white.
http://www.candlepowerforums.com/vb/...h-Cri-and-170T
I would chose ≈5400Kish high CRI, but if you think cool tints will do well...

[constantine_a_f]

Here are some beamshots of a high CRI vs cool white.
http://www.candlepowerforums.com/vb/...h-Cri-and-170T
I would chose ≈5400Kish high CRI, but if you think cool tints will do well...

I will as I said buy two leds, in order to have some error margin. I will buy one 70 CRI 6500 K and one 85 CRI 4200 K. I will post some intraoral photographs as soon as I assemble the leds (before soldering). Hopefully I will not damage one of them, so that we can compare.

Meanwhile,I just received this email

Dear Constantine,
Thank you for your email.
We have more heat pipes that are not displayed on our website.

Pls let us know your detailed requirements,then our engineers can study and give some advice.

We can make 2pcs samples for you.We will ask for the sample free based on your requirements.


Best Regards,
Cary Kang

I will probably make the hole thing out of a heat pipe.

I will post the new draft as soon as I have time to draw it at acad.

[MikeAusC]

I would just run the Heatpipe horizontally, across the top of the lens frames, with the LED housing clamped on at the centre.

[constantine_a_f]

Another draft incorporating a heat pipe

http://www.2shared.com/document/SrAP31ND/drawing.html

[Harold_B]

I have a study complete Constantine. Please send me an email with an email address where you'd like me to send the pdf.

[constantine_a_f]

Over the last days I communicated with several companies capable of making a heat pipe on demand. Two of them told me that it is impossible to make it that slim and light and one said that it could do it for 700usd but thought the heat pipe wouldn't work very well as a heat dissipater. Therefore I am back to my aluminum tube+heat sink design which was meticulously tested by forum member Harold B and worked fine. I ordered the parts mentioned above with only one difference. I got a 6500 K and a 4100 K rebel. I expect them in 6-10 days (probably 10 because of the orthodox easter vacation, here in Greece). I will know how to solder by then and I will also find the aluminum tube or rod (I have not yet decided whether to make it out of a rod or a pipe/tube).

Thank you all for supporting my project.

[kosPap]

why don;t you combine both?

make the light out of a cylindrical auminum section then take a CPU heatsink, drill a hole the size of the cilidrical section and the mill the heatsink round around this...in effect you will make a cooling sleeve..it will not be the same but press fitted and tehrmally epoxy bonder will offer considerable gain....

now in Greek....
kane ena kylidriko maniki apo mia psiktra epe3ergasti

[constantine_a_f]

I found this

http://www.allerc.com/hyperion-20-mm...nk-p-1887.html
Do you think it is possible to use the carclo lens without a holder, or cut the base of the holder and use only the base? Will this provide the needed heat dissipation? The diameter of the holder is 21.5 mm. I could mount the lens without the holder, use epoxy glue to fix it (or make a fastener, then glue the star on a 20mmX1cm high aluminum rod and insert it from the other side of the heatsink. This way the heat transfer would be

led->star base->aluminum base->heatsink body.

Can this work??

[srfreddy]

Colour Temp and CRI are separate measures.

6500 is sunlight and no-one complains that sunlight has a low CRI.

LEDs don't have the smooth spectral response of sunlight and generally the cooler LEDs have a worse spectrum, so people associate a high CT with poor CRI.

High-CRI LEDs generally have a mid-range CT

What I'm saying is that cool white leds don't have good CRI-and plus, CRI is based off a blackbody radiator under a certain color temperature.

[constantine_a_f]

I have a final design. I decided to make the heat sink and the body out of aluminum. I will use a solid rod for the heat sink (see draft) and a tube for the body. I will glue them with thermal epoxy. My gf is a civil engineering student and helped me calculate how much heat dissipation is needed. The final design made by aluminium 6060/1/3 will be more than enough for the 2.1 watt rebel (at 700mA).

http://www.2shared.com/document/gn2Q...2-Layout3.html

[ledpwr]

I have just noticed that luxeon has brought out a new rebel ES (LXW8-PW50), it has a colour temperature of 5000k, it has a cri of 85 and an efficiency 110 lumens/watt.
But since it has a larger die (same size as an xp-g) it will have produce slightly wider beam (1.4 times wider). You will have to wait a few weeks for shops to get it in stock though as it is new.

[constantine_a_f]

I have just noticed that luxeon has brought out a new rebel ES (LXW8-PW50), it has a colour temperature of 5000k, it has a cri of 85 and an efficiency 110 lumens/watt.
But since it has a larger die (same size as an xp-g) it will have produce slightly wider beam (1.4 times wider). You will have to wait a few weeks for shops to get it in stock though as it is new.

That's the good with this design. I will be able to upgrade it. I already ordered some leds. I will start with them and then upgrade from there. In order for the design to be upgradable, I will attach the lens and the holder mechanically and I will not glue them with epoxy.

Thanks Ledpwr

[srfreddy]

The LXW8-PW40 are what Zebralight is going to use for their High-CRI line of lights-warmer than the 5000k, but I think I would prefer the 50 over 40, but I have no experience with anything other than incans and cool LED's.

[MikeAusC]

. . . . I decided to make the heat sink and the body out of aluminum. I will use a solid rod for the heat sink (see draft) and a tube for the body. I will glue them with thermal epoxy.

Keep in mind that the best Thermal Epoxy has a Thermal Resistance 30 TIMES higher than aluminium, so make sure there is a large metal-to-metal contact.

Epoxy should only be used in the thinnest possible layer to fill in surface irregularities.

[kosPap]

I found this
Do you think it is possible to use the carclo lens without a holder, or cut the base of the holder and use only the base?

probably only the latter..usually the holders alingn the LEd to teh optic and take care of the proper focusing distance

also, did you see thsi?

http://www.cutter.com.au/proddetail.php?prod=cut955 - 9.9mm

http://www.cutter.com.au/proddetail.php?prod=cut962 - 16.1mm

since you will mounting them on the glasses?