Paul C. Buff, Inc. Technical Forum

Paul/TS -

First off, Great job on the launch of the Einsteins as well as the new laser reflectors. The reviews looks great & can't wait to get them in hand along with one of those upcoming V2 PLM's. I'm sure you guys are all very busy, busy bees right about now but that's always a good thing for business! :D

I'm sure you guys have considered it (searched on these forums but couldn't find any posts), but how viable would it be to offer a high-output, low-wattage LED modeling light option for your lights. Surely they wouldn't be nearly as bright but may serve nicely for vagabond users on location who would benefit using the modeling lights for longer intervals without draining too much of the battery. Something akin to those 60-200 lumen 3-5V LED flashlights, but with a standard Edison mount and outer diffuser to spread the light somewhat evenly. As far as daylight correction and modeling light power tracking go, aren't LEDs easier & faster to set at any given color temp. or brightness?

Any thoughts?

Thanks for your time! :)

Good idea, but also a common suggestion. But there are many obstacles to this. First, LEDs aren't the magic people think they are. The are 2-3 times as efficiency as tungsten so it still takes a substantial amount of wattage . . . a one watt LED can produce the equivalent lumens of about a 3W incandescent at best. A 30-50 W LED array would provide reasonable modeling, but would cost a small fortune.

Next, it would have to be developed and produced by an expert LED company . . . probably a $100,000 effort if you could find the vendor, and the assembly would probably have to sell for well over $100. Finally, the dimming curve would not match that of incandescents so the WYSIWYG modeling would be lost.

In short, yes this technology is coming, but is premature for us to consider at this point in time.

I've seen standard edison-screwmount LED lightbulbs on the internet. Google it, you can buy bright lights albeit for a premium. They normally use several 2-5W ultrabright LEDs ganged into a compact reflective housing.

I think it's for rich people who want to be environmentally friendly or something...

The upside is that they won't shatter like normal bulbs if you bump, and they won't draw as much current, plus they run cooler than normal bulbs.

You can buy it and use it on your own just like other bulbs...

I don't believe there are any 2-5W individual LEDs on the commercial. Most Edison base LED lights I have seen are in the 2-5W total range (using multiple 1/4 to 1/2W LEDs) and produce the equivalent of 5-12W equivalent tungsten light, and cost a fortune and won't generally fit into monolight sockets.

Luxeon, Cree make some

5W of dissipation is far greater than a small LED package could possibly dissipate - a 5W resistor is about 3/8" around by 1 1/4 long and gets very hot. I haven't checked specifically but I suspect the LEDs you mentioned are arrays of many smaller LED which typically have a maximum dissipation around 1/2W . . . most are more like 1/10 to 1/4 W. Even with a 5W LED you;'re only going to get the illumination equivalent to about a 15W incandescent lamp.

http://www.cree.com/products/pdf/XLampXP-G.pdf is a direct link to the data sheet of one of the best generally available high power LEDs. It generates between 100 and 130 lumens at 350 mA@3V, depending on the chosen color temperature and the efficiency bin. That is about the maximum efficiency and it is a 1W LED there. That's about 5-6 times more light per watt than a typical incandescent bulb.

It can be driven without damage to at least 1000mA@3.3V and then will generate about 2.5x as many lumens at that drive current. All assuming you can keep it cooled to 25 C. At 100 C you lose about 15% efficiency. These are indeed not easy to cool. And an array of them with a substrate, the drivers for them and the optics would not exactly make for an inexpensive modeling lamp. The CRI is between 70 and 80. Another drawback compared to incandescent for some uses.

Ronald

Interesting info - I'll check it out.