Originally Posted by
HerrKaLeun
Oh, I would totally be OK to buy one single headlight at higher price instead of fiddling with multiple ones. and I don't doubt the light from your light will be great. not sure if $175 great, but sure better than my $30 solution
I perform lighting design and know the actual outcome (an average lux/fc, uniformity etc.) is more important than just a single lumen number. with my flashlight I get a lot (a lot!) of light in one spot and little light over a wider area. So i see the limitation of normal flashlights for biking.
Would it be possible to have just 1.5 hours runtime with the batteries integrated? The issue I see with the external battery is that first more clutter. Second, when you have multiple bikes you take the light from bike to bike. Or when we commute and take off the lights to take in. Seems unnecessary hassle to deal with two "boxes" and cable. I assume no one wants to leave the light on the bike when it is locked up somewhere.
I see the limitations of lm/W and Wh per battery you have to deal with. Standard li-io batteries probably are where they are now, and LED may add 10 lm/w per year. so if you have that much lumen output, you will need a physically large battery for a while.
May I ask what lm/w your LED have? I assume manufacturers bin their LED and the lower efficiencies (100-140 lm/W or even less) go to light fixtures, and the higher ones to battery device? I found this 2014 anouncement of Cree breaking the but none of their fixtures goes anywhere near that efficiency. Actually Cree is trailing a bit behind Philips these days regarding efficiency.
We are running a 1x5 Lumiled Altilon. You can check out the datasheet here: (apparently still can't post URLs), but google "Lumiled Altilon" and it'll be one of the first results.
The older Altilon PnP has been out for a bit, but that was a full PCB assembly that was being pushed to try and drive lower cost alternatives. Most OEM's have opted for a custom PCB board since the overall actual cost doesn't really decrease once you start making hundreds of thousands of modules. The Altilon SMD's are very new. OEM's were given access to them in the last year or two, so you'll start seeing it pop up on automotive OEM headlights in about a year or two (typical design lifecycle for a new headlight is 2-3 years). We actually had to get approved by Lumileds to use the chip, and have to buy through approved US suppliers. It's a pain sometimes, but the chip performance is worth it.
The actual Lm/W is probably around 110-125 depending on what bin and a lot of other factors. The larger the chip the better Lm/W you'll get, but the optical performance will decrease quite a lot. CREE rates their chips at standard room temperature, aka the second the chip turns on. As I am sure you know most chips actually run at 85-90*C, so trying to rate a chip at 20*C I find is just stupid. OSRAM does this as well, very annoying.
Lumileds thankfully rates their chips at actual operating temperatures, so their published numbers *seem* lower, but they are actually more in line with real world applications.
As for the internal battery, it's next on the list to design. I've got plenty of ideas for a bar or helmet mount specific internal battery lamp. Also in the future I do want to do a whole low/high beam lamp that's a little bigger than the current one, but that'll also be fairly pricey just because now have to double up on the chips and tooling sizes. The chips are one of the biggest cost driving factors (along with the battery). But if the kickstarter goes well, and specific optic type lamps like the Focal series can help us bring in revenue, can buy chips in larger quantities to help bring price down.