XM-L vs. XP-G comparision - follow up to discussions about XM-L in ZL H51

Hello all, long time lurking - first time posting .
I plan to buy AA powered headlamp from Zebralight. XM-L introduction on the market made it much harder for me - the eternal problem : should I buy now or wait for ZL in H XM-L AA form?
Few times the possibility and gains or lack of them were discussed on cpf, for example here: http://www.candlepowerforums.com/vb/showthread.php?313074-Zebralight-AA-and-XM-L-am-I-dreaming

I decided to combine data from:
http://www.cree.com/products/pdf/XLampXM-L.pdf and
http://www.cree.com/products/pdf/XLampXP-G.pdf
into one graph. Here is the effect om my work:

http://min.us/mvoU5Nx


Because I already had all the data in one spreadsheet it was easy to calculate watts:


Maybe somebody will find it useful. My conclusions (keep in mind that my knowledge about electronic is minimal, I'm looking for confirmation from you):
-Comparing with 200 lm from XP-G it should be possible to get at last 230 lm from XM-L with the same run time or 200lm with 14% more runtime (1.4W vs. 1.6W). Maybe little more - because of lower voltage the losses in the driver should be smaller.
- It's interesting how much energy is wasted in the driver. In theory Eneloop AA has about 8640W (2.0 Wh *1.2V * 3600s) at 200 lumens XP-G uses 1.6 W. In perfect world with 0 loses ZL would lit at 200 lumens for 90 minutes. In real word it works for 54 min. 40% of power is wasted as heat and this is one of the most effective 1AA lights :sigh:

As for me I'm probably going to buy XP-G now. AFAIK there is no leaks about H51 XM-L (H60 AA?) coming soon. The difference in lumens is not that huge at 1AA power levels and I plan to use d-c-fix as a diffuser and have the option of little more throw after removing it. With XM-L it would be the other way and I would have to use fresnel lens or something similar for throw when needed.

What do you think guys?

Thanks for the great info! Also welcome to CPF

What an introductory post! Nice work and..

:welcome: - Jas.

I dont understand the first graph. Why does it show 700+ lumens at 100 milliamps? I guess thats why you translated into something useful.

I didnt realize that XM-L was more efficient even at lower wattage....I thought its main advantage was that it could be driven harder.



In any case, from what Ive seen, I'd recommend waiting for an XM-L light that strikes your fancy. They seem to be noticeably more efficient and brighter. Im waiting for an XM-L 1xCR123 twisty that can replace my itp A1. The Neutron 1C is close but its an inch longer and clickey.

joe1512 said:

Im waiting for an XM-L 1xCR123 twisty that can replace my itp A1.

Click to expand...

Have you looked at the Quark MiNiX 123?

- Jas.

stp said:

- It's interesting how much energy is wasted in the driver. In theory Eneloop AA has about 9504W (2.2 Wh *1.2V * 3600s) at 200 lumens XP-G uses 1.6 W. In perfect world with 0 loses ZL would lit at 200 lumens for 99 minutes. In real word it works for 54 min. 45% of power is wasted as heat and this is one of the most effective 1AA lights :sigh:

Click to expand...

An Eneloop is rated for a minimum 1900 mAh at 1.2V (1.9Ah x 1.2V = 2.28VAh = 2.28Wh). If the XP-G is drawing 1.6W, it has a theoretical minimum run time of 1h 26m (2.28Wh ÷ 1.6W = 1.425h = 85:30).

Thanks guys

joe: the first graph uses two y axes, second one is on the right. The two top graph curves with small circles represents forward voltage vs. milliampers so you need to read y values from right axis - no 700 lumens at 100 milliamps sadly ;-) (sorry for confusion).

Yes at higher power levels the XM-L is much better but I'm into 1AA lights. Anyway thanks for suggestion.

Its all not that important now because in the meantime Zebralight announced new lights. One of them will be H502 - 300 lumens from 1AA. It combines XM-L with new , more efficient driver. Sadly no info on runtimes so it's not possible to compare it to current ones. I will probably end with H51 for now anyway - H52 was not announced - I need something with at last little throw and the H502 will hit the market in about six months. I can't wait that long.

beamis:
For some reason I used 2200 mAh as a base for calculations. Probably I confused Eneloops with other standard "2500" NiMHs. Thanks for pointing it out. I will correct it in few minutes but I'm going to use 2000mAh because this value is used by ZebraLight. Is it ok for you?

stp said:

For some reason I used 2200 mAh as a base for calculations. Probably I confused Eneloops with other standard "2500" NiMHs. Thanks for pointing it out.

Click to expand...

But now there are 2500 mAh LSD NiMH: the Eneloop XX.

- Jas.

Yes, I have them too ;-) But because I'm calculating driver loses based on runtime taken from ZebraLight webpage I need to stick to what they use: " Runtime tests are done using Sanyo 2000mAh Eneloop AA batteries." Using XXX to calculate theoretical runtime and comparing it to real ZL with standard Eneloop would be pointless.

stp said:

- It's interesting how much energy is wasted in the driver. In theory Eneloop AA has about 9504W (2.2 Wh *1.2V * 3600s) at 200 lumens XP-G uses 1.6 W. In perfect world with 0 loses ZL would lit at 200 lumens for 99 minutes. In real word it works for 54 min. 45% of power is wasted as heat and this is one of the most effective 1AA lights :sigh:

Click to expand...

nothing surprising, boost drivers are very inefficient when supplied with 1-1,2v, especially cheap dx\kd drivers, run the driver off lithium aa cell (1,7v no load) and you'll see efficiancy of that driver jump to 80% or more.

stp said:

Yes, I have them too ;-) But because I'm calculating driver loses based on runtime taken from ZebraLight webpage I need to stick to what they use: " Runtime tests are done using Sanyo 2000mAh Eneloop AA batteries." Using XXX to calculate theoretical runtime and comparing it to real ZL with standard Eneloop would be pointless.

Click to expand...

Ah yes, good point...

- Jas.