how to determine suitable LEDs?

[aethyr]

I'm a newb, trying to learn as much info as I can before I embark on building my own saber.

One thing I noticed in my reading of this forum is that often times a particular LED will be deemed unsuitable for sabers. I can understand if their efficiencies are not very good, ie lumens/watt. You clearly want the most efficient, right?

So then why wouldn't this led (http://ledsupply.com/creemce.php) be suitable? Granted its only white, but 430 lumens at 350mA is pretty good, no? I've seen posts about a P7 that produces 900 lumens at 3000mA. But that cree led I linked to is way more efficient. Is light output linearly additive? Can I put 3 430 lumen leds together to get over 1200 lumens at 1000mA? Or even just 2 together to get 860 lumens at 700mA is darn good, no?

The other things I would imagine are factors are heat, but as a corollary to efficiency, the more efficient a led is, the less heat it should produce right?

So what other considerations are there for determine whether a particular led is suitable?

[Sunrider]

Hey why lose all those lumins through a color filter. At the same size is the all powerful ssr-90. 2000 colorful lumins. Boyaca;)

http://www.luminus.com/content1517

[cardcollector]

well, do you want it for a filter?

or for a white blade? The seoul p4 white is my personal favorite with a white blade.

If you are planning to use a filter, I would suggest you find a RGB alternative...

[aethyr]

Oh, I just picked a led that seemed to have really good efficiency. I wasn't attached to the white. But it seems to me, if you could have a very efficient white led, then why not use a filter?

That SST-90 seems like it would need all 9000mA to get 2000 lumens. Thats a lot of juice. I believe with the 900 lumen P7, people were struggling to come up with "just" 3000mA.

My quick research shows that lumens are linearly additive. So two 430 lumen leds is equivalent to a single 860 lumen led. So using my cree led I found, you only need 700mA to hit 860. And at merely 1400mA (less than 1/2 the juice to power the P7), you get 1720 lumens, which is almost twice the light output of the P7. So that cree led is 4x more efficient than the P7 and over 5x more efficient than the SST-90.

What am I missing?

[Sunrider]

The cree is efficient but if you put a filter on it besides yellow you will lose 2/3rds of the lumens. The ssr-90 is bad at the upper end but at 3 or 4 amps way more lumens than the rest.;)

[FenderBender]

It isn't just the LED, its what power it needs and can you fit that power solution in a hilt. Its also about heat management, the single greatest killer of LEDs. Also efficiency is determined by its optics solution. Can you get all those lumens up the blade tube? We have this problem with the tri-rebels, lots of power/lumens, but can't find a great optics solution to get it all up the blade tube.

[Rhyen Skytracker]

The main issue with the LED you specified is the voltage requirements. You have to keep in mind you don't have much room for batteries. With the batteries you not only want to look at the voltage, but you also want to look at the mAH. That will determine the run time. The Li-Ions will give you higher voltage and the NiMH will give you higher mAh for the same sized batteries. We like to keep the voltages around 4.8 V - 7.4 V. Anything higher than that and you will have problems fitting them in a normal sized saber hilt.

[aethyr]

It isn't just the LED, its what power it needs and can you fit that power solution in a hilt. Its also about heat management, the single greatest killer of LEDs. Also efficiency is determined by its optics solution. Can you get all those lumens up the blade tube? We have this problem with the tri-rebels, lots of power/lumens, but can't find a great optics solution to get it all up the blade tube.
Ah optics. Can you elaborate on this? Is optics merely the case of focusing the beam in a straight line up the tube, sort of like the parabolic reflector on flashlights? Or is there more than this? Are the leds so different than you need different optics for each one? I mean, if I recall my math correctly, for a parabola, its a matter of putting the led at the focii, right?



The main issue with the LED you specified is the voltage requirements. You have to keep in mind you don't have much room for batteries. With the batteries you not only want to look at the voltage, but you also want to look at the mAH. That will determine the run time. The Li-Ions will give you higher voltage and the NiMH will give you higher mAh for the same sized batteries. We like to keep the voltages around 4.8 V - 7.4 V. Anything higher than that and you will have problems fitting them in a normal sized saber hilt.
Ok, so that cree led I posted requires 12 volts, so that is too high? Is it really that hard to find a battery solution that provides 12V with decent amp*hours that fits in the hilt?

So I found these "gumstick" batteries: linky (http://www.megabatteries.com/item_details2.asp?id=15521&cat_id=217&uid=1417)
They are 0.22 inches thin, 0.6 inches wide, and 2.6 inches long. 10 of them together produce 12V with a total of 14000mAh of capacity and take up less volume than 4 AA batteries. They are extremely expensive and 10 of them will set you back $100, but the solution to voltage does exist.

Now back to the filter issue. At 430 lumens each, lets say we install 5 of them, for a total of 2150 lumens. Lets say we throw a filter and lose 2/3 of the lumens. That still gives us 716 filtered lumens at only 1750mA of consumption. The gumstick batteries can provide 12V and plenty of amp*hours.

The other issue is focusing the light beam...I don't have a solution to that yet, but what if I just yanked a suitable flashlight reflector?

[Rhyen Skytracker]

You won't find too many people who will spend $100 for batteries for 1 saber. I know I wouldn't. Price is a concern for these too. Parts and electronics for sabers can add up very quick.

[Sunrider]

Ok three of those batteries would be 16 x 15mm square. To fit even 9 in a mhs hilt you would need 165mm or 8 inches of hilt for only batteries.

The deal with the lens is it focuses the light through a 6 to 10mm aperture into a narrow beam so the led has to fit directly under the small hole. The led must also be centered with the 10mm hole so only one led can be used. Even multi die leds like the mce suffer from not being directly centered with the lens. Reflectors don't make a narrow enough beam.

Check out the Rebles you can get 200 or 300 lumens from them.

If you wan't a higher power challenge the nice thing about the ssr-90s is the light is well distributed across the led with more junctions to produce more light.:cool:

The most power dense batteries right now are the 18650s. ;)

[aethyr]

Ok three of those batteries would be 16 x 15mm square. To fit even 9 in a mhs hilt you would need 165mm or 8 inches of hilt for only batteries. Not sure how about your math here. AA batteries are 50mm long and 14.5mm in diameter. The gumsticks are 66mm long and 16x15mm wide. So a stack of 3 gumsticks are a tad wider(1.5mm) than 1 AA and a bit(1.6cm) longer. The 9 gumsticks are thus only 7.5% more volume than 4 AAs. Plus, due to the flat shape of the gumsticks, you could, in theory actually mount them outside of the hilt, as both cosmetic and functional additions.



The deal with the lens is it focuses the light through a 6 to 10mm aperture into a narrow beam so the led has to fit directly under the small hole. The led must also be centered with the 10mm hole so only one led can be used. Even multi die leds like the mce suffer from not being directly centered with the lens. Reflectors don't make a narrow enough beam.
So is this just an issue of nobody making an optic able to handle a larger aperture for a multi die led? There are far larger light sources, like search lights that are able to focus a beam of light, so it seems hard for me to understand why a tiny multi die led can't focus its beam? I mean, yes, a single led is small, but in the context of light sources out there, a multi die led is still pretty small, no?

I'm not sure I understand why a reflector can't make a narrow enough beam? Is it just an issue of mismatched reflectors or imperfections of reflectors in general? I mean, telescopes use reflectors and are able to precisely focus light...

I'm not trying to rock the boat, I just want to understand why certain conventions have been decided upon in the saber building world :-)

[FenderBender]

Because a lot of us have spent lots of time and money trying the available options for optics and we have settled on a paltry half dozen or so combinations of things that DO work for our specific application.

On that batteries, unless those gumsticks are protected, you will also have to solder protection PCB's to them and thus make for more volume. Though, the idea DOES have an attraction to it:D

NEVER mount LI-Ion batteries exposed on the outside of the hilt. That would be foolhardy. One missed spin or stunt and it falls on the ground damaging a cell and it could go boom. Literally.

I would like to re-visit the gumstick idea, I haven't seen them in Li-Ions just NiMh's, I'll have to look at this some more.

[Rhyen Skytracker]

You have to keep in mind that all MHS blade holders are made to accept the optic holder in the store. Anything larger or smaller would have to have a custom machined blade holder. Believe me, many of us have researched all the new LED technology and optics that have come out and we use the ones that work best for lightsabers. The problem with most reflectors that will fit in the MHS blade holder is it produces too wide a beam of light which will make the blade super bright at the bottom and dim at the top. If you find any that you think will work please post them and we will see if anyone else has tried it or if it will work or not. We are not trying to discourage you at all. We have just tried alot of different LEDs, batteries, drivers and optics and that is how we ended up with what we use now. As new technology comes out, several of us will test it to see if it will meet our needs. We try to progress as new technology progresses.

[Onli-Won Kanomi]

Aethyr in principle it should be possible to make effective collimating lenses for our purposes that work with that LED...however the issue is mainly a commercial/marketing issue more than a technical one [getting 4 emitter dice to work WELL with a single collimator lens IS more difficult than one emitter die...but Lux Vs can do it so it IS possible].

You have to understand that our lightsaber replica prop application is an almost invisibly TINY 'niche' market to the LED manufacturers and optical companies that make lenses...a single big LED SIGN or 'Jumbotron' can use more LEDs than ALL of us saberfans together do in a whole year...so they just don't consider us worth the financial bother to do the optical engineering necessary to make optic solutions specific to our needs..not enough 'Return On Investment' from the few of us.

So we are forced to look for optical solutions among the PREexisting optics they have designed for other purposes.

Unfortunately those other purposes are usually the OPPOSITE optically from what we want to do - the aforementioned jumbotron/signage usage for example where probably the most LEDs are used nowadays wants it's millions of lenses yearly designed to produce the WIDEST possible viewing angle - we want our few thousand optics a year designed to produce the NARROWEST...guess which market gets served and which gets ignored by manufacturers producing thousands of LEDs and optics hourly?

So even when a potentially amazing new LED comes out we have to wait and see if anyone will 'just happen' to make the VERY narrow [5-10 degree] collimation optics we need...and we normally keep waiting...and waiting...and waiting...

When the Luxeon "Endor Star" came out it excited some of us...afterall the name ENDOR is a Star Wars reference eh...maybe someone was thinking of us?...and 540 lumens white was [and still is] pretty good...and the announcement said optics would be "forthcoming"...

...we waited, and waited...eventually optics came...25 degree optics...good for signs maybe, USELESS for us...despite the name Endors just never became useful for sabers...XWingband found an alternative Tri-Rebel configuration that worked better [AceRocket had also made a Tri-Lux even before Endors]...but even with his solution which as good as possible the COMBINED efficiency of the tri-rebel AND best optics solution he could find for them was less than ideal and some of their potential was wasted.

The 'saber microniche market' just isn't big enough for them to design optics SPECIFICALLY for us so as Fender says we are 'stuck' with the few combinations we can 'find' that 'happen' to work together...and smiths and saber suppliers like him, Strydur, Xwing, Ace, Goodman and MANY others [apologies to all I've forgotten] have done a lot of experimenting to find those combinations, for which we should all be grateful.

Someone might make a useful very-narrow optic for our purposes for that LED someday...by all means tell us if you see it...keep rocking the boat...some of us like to rock! Rock On! ;-)

[aethyr]

On that batteries, unless those gumsticks are protected, you will also have to solder protection PCB's to them and thus make for more volume. Though, the idea DOES have an attraction to it:D

NEVER mount LI-Ion batteries exposed on the outside of the hilt. That would be foolhardy. One missed spin or stunt and it falls on the ground damaging a cell and it could go boom. Literally.

I would like to re-visit the gumstick idea, I haven't seen them in Li-Ions just NiMh's, I'll have to look at this some more.
The gumsticks I linked to are NiMH, so they should be fine on the outside of the hilt and should be ok without protection electronics, no? I think they would be cool, functional additions to the hilt. Afterall, a lightsaber's physical appearance is supposed to be indicative of its fictional functionality. For once, we can add things to the hilt that serve a functional purpose! Not to mention, we would then have additional, freed up space in the internal hilt.




Aethyr in principle it should be possible to make effective collimating lenses for our purposes that work with that LED...however the issue is mainly a commercial/marketing issue more than a technical one [getting 4 emitter dice to work WELL with a single collimator lens IS more difficult than one emitter die...but Lux Vs can do it so it IS possible].

You have to understand that our lightsaber replica prop application is an almost invisibly TINY 'niche' market to the LED manufacturers and optical companies that make lenses...a single big LED SIGN or 'Jumbotron' can use more LEDs than ALL of us saberfans together do in a whole year...so they just don't consider us worth the financial bother to do the optical engineering necessary to make optic solutions specific to our needs..not enough 'Return On Investment' from the few of us.

So we are forced to look for optical solutions among the PREexisting optics they have designed for other purposes.

Unfortunately those other purposes are usually the OPPOSITE optically from what we want to do - the aforementioned jumbotron/signage usage for example where probably the most LEDs are used nowadays wants it's millions of lenses yearly designed to produce the WIDEST possible viewing angle - we want our few thousand optics a year designed to produce the NARROWEST...guess which market gets served and which gets ignored by manufacturers producing thousands of LEDs and optics hourly?

So even when a potentially amazing new LED comes out we have to wait and see if anyone will 'just happen' to make the VERY narrow [5-10 degree] collimation optics we need...and we normally keep waiting...and waiting...and waiting...

When the Luxeon "Endor Star" came out it excited some of us...afterall the name ENDOR is a Star Wars reference eh...maybe someone was thinking of us?...and 540 lumens white was [and still is] pretty good...and the announcement said optics would be "forthcoming"...

...we waited, and waited...eventually optics came...25 degree optics...good for signs maybe, USELESS for us...despite the name Endors just never became useful for sabers...XWingband found an alternative Tri-Rebel configuration that worked better [AceRocket had also made a Tri-Lux even before Endors]...but even with his solution which as good as possible the COMBINED efficiency of the tri-rebel AND best optics solution he could find for them was less than ideal and some of their potential was wasted.

The 'saber microniche market' just isn't big enough for them to design optics SPECIFICALLY for us so as Fender says we are 'stuck' with the few combinations we can 'find' that 'happen' to work together...and smiths and saber suppliers like him, Strydur, Xwing, Ace, Goodman and MANY others [apologies to all I've forgotten] have done a lot of experimenting to find those combinations, for which we should all be grateful.

Someone might make a useful very-narrow optic for our purposes for that LED someday...by all means tell us if you see it...keep rocking the boat...some of us like to rock! Rock On! ;-)
Ah! Thank you. This is exactly the response I was looking for. So basically, us niche saber enthusiasts have been forced to "scrounge" around for components that suit our needs.

I found on candlepower forums (http://www.candlepowerforums.com/vb/showthread.php?t=226934) a post about a 25 die led that produces 4000 lumens at 3.5A. Is there such a thing as too bright? Cooling definitely an issue here, along with probably power. And finally, the optics issue...

What is the optimal or acceptable angle for optics?

Has anyone looked into custom optics? I know our niche is small, but lets assume we could find our dream led, perfect power and light output, but no suitable optic. Would it be worth it to custom manufacture the optic and sell it to the community at cost?

[Sunrider]

Ok you got me on the three battery estimate. You could cram 4 in the mhs hilt side by side, but I would be careful subjecting the corner of a square battery to impacts. This usually shorts out the battery & could result in fire.:(

The down side to NiMh is they are a high maintenance battery with a high discharge rate & tend to heat up under heavy draw.

I have heard the ledengin 10w is a slightly tighter multi die powerhouse than the mce. The 10w is a good led & pretty easy to drive with a parallel pcb.;)

Battery comparison; 5 sticks mentioned earlier = 8.4wh. 1 li-ion 18650 = 9.62wh

[Rhyen Skytracker]

The degree lens we look for are between 5 degree and 10 degree. I have use some 8 degree that were perfect.

[aethyr]

I found this optic (http://www.polymer-optics.co.uk/Cree%20MC-E%20Optics%20Range.pdf) for the cree mc-e. There is a tri-led narrow optic (part 158 ) that seems like it could fit the bill.

3 x MC-E = 1290 lumens at a mere 1A of power. Granted the forward voltage is pretty high at over 12V, but would boostpuck work to supply the necessary forward voltage?

[Sunrider]

I think I read on the fx sabers forum that someone tried all of those. Diffusive lenses are bad and the tri optics pattern is not good.:-(

[aethyr]

Well, maybe the custom optic is the approach to take?

What is the best LED out there, but that lacks appropriate optics?