Written by tandem
on 01-09-2011 11:03 AM GMT
Originally Posted by pae77
You don't need to speculate, you can look at Cree's data. Here is a snapshot I took of Cree's product comparison application
showing output at different drive current rates for the XP-G R5 compared with the XM-L T6 and XM-L U2 output bins. The T6 is readily available right now.
Of course short of asking the maker or doing surgery on our lights we've no way of knowing exactly what the drive current at the emitter is in the SC51 so we'll have to estimate by working backwards, sideways and maybe upside down. Bear with me while I puzzle this out and realize that I may be way off base on this:
How about we use 75% as the efficiency factor to convert emitter to out the front lumens so to go backward we'll divide by 0.75. 200/.75 = 266 emitter lumens. From the data sheet we can see the XP-G can generate ~266 lumens with roughly 0.75A but with Vf at 3.2V. A fresh AA NiMH is no where near 3.2V so that voltage has to be stepped up and since you get no useful work without power this requires more current... roughly 2.2A of current in fact, just over 1C for a 2000 mAh cell. I wonder if we can estimate current at the LED, even if roughly so? 2.2A / 0.75A = a factor of 3.
Does that hold up? Let's see... the light is said to put out 100 lumen out the front on High 2. 100/.75 = 133 emitter lumens. The Cree table doesn't go that low so we'll have to extrapolate. Assuming Vf of 3.2 and lumens per watt at 130 (close enough) drive current P=IV so I = P/V = 1W/3.2V = 0.3A or thereabouts for the XP-G.
Using the same cell for measurement of the light on High 1, I get 0.8A for High 2, so the conversion factor isn't exactly linear (no surprise) but works out to 0.8/0.3 = 2.66. That sounds about right as these LEDs are all supposed to be more efficient at lower drive levels. I suppose I could have used lumens per watt and nominal NiMH voltage to get at this too.
Anyway... would a light designer want to draw more than 2.2A from a NiMH cell? To do so you start dropping the light's runtime on max. Let's say they were willing to introduce a burst or "High 3" feature and go up to 3A draw from the cell. 3 divided by our conversion factor of 3 = 1A.
1A at the emitter produces 347 lumens on a XP-G R5 and 388 on an XM-L for a 12% improvement. 388*.75 = 291 OTF lumens for the XM-L vs 260 for the XP-G. I guess 11% is worthwhile enough a difference, **if one were to decide pulling 3A from a 1xNiMH cell was a design objective**. But as the XP-G can already be driven by 1A (at the emitter, 3A at the tailcap for a 1xNiMH light) and since no production 1xAA does this today, one has to assume the introduction of the XM-L won't change that.
I guess that is the real bottom line - for 1x NiMH or alkaline powered lights the XM-L doesn't really introduce anything new by way of potential except for perhaps a different beam profile much as the XP-G is different than the XR-E. There is a downside to going XM-L early on as initially at least you lose access to all manner of tint bins.
Where the XM-L really shines for flashlight makers is lithium ion powered lights, particularly those taking larger capacity cells that can deliver 1.5 - 3A without being driven (much) over a 1C rate. Li-ion powered lights have an efficiency advantage over NiMH/Alkaline given their drivers can be more efficient since they operate much closer to Vf for the LEDs. Plus they are physically larger lights that will handle heat better.
Edit: To bring this all directly back to the Zebralight -- Maybe a 14500 powered SC51/H51 would auto-enable a somewhat higher high simply by virtue of detecting the higher input voltage. But the lack of official support for 14500's seems to suggest that Zebralight isn't as keen on supporting lithium ion cells in AA format lights, so why would they enable such a thing? Lithium ion cells are truly an enthusiast's domain, not the mainstream. As this industry matures the volume producers surely are looking more at the mainstream than a few of us here on CPF.