Quark efficiency.

I am a bit confused.

Based on my calculations Quark AA (350mA) is 79% efficient (BTW my prediction is the R5 will be rated around 108 lumens.

Quark CR123 or 2AA (700mA) is most efficient at about 83%, and for some reason I calculate that Quark 2-123 (990mA) is only 67% efficient!

This is based on assumption that 700mA being 1.8 times brighter and 990mA being 2.5 times brighter then 350mA.


---------350mA----700mA---990mA---350.79%--700.83%--1000.67%
q5--------107------192.6----267.5----84.53----159.858---179.225
r2-------114------205.2----285------90.06---170.316---190.95
r3--------122------219.6----305------96.38----182.268---204.35
r4--------130------234------325------102.7----194.22----217.75
r5-------139------250.2----347.5----109.81--207.666---232.825

So if my calculations are correct then why such difference 83% vs 67%?

could it be due to heat?
in the end, the head's physical size remain the same across the board, so its understandable that 123^2's relative dissipating wouldn't be as good as the rest.

Well based on the Cree spec I am calculating that all losses due to reflector, heat, driver etc are 17% for single 123 and 33% for 2-123 flashlight.

My calculated numbers match the advertized out front lumens very closely, so either my 1.8 multiplier for 700mA and 2.5 multiplier for 990mA is incorrrect and I misread the spec, or the the drive levels are not exactly 350ma, 700mA and 990mA?

I do not know. I just did the calculation because I was interested how bright the new R5 Quarks will be and it looks to me that I was right on with the 230 and 206 prediction for 700mA and 990mA drive levels (my calculations were 232 and 207). But my calcs require the total efficiency to be 83% and 67% - and that is why I am confused...

What exactly are you calculating here?
Are you taking the rated lumen rating for the led at specified current and comparing to rated lumen for the light?

The actual led current in the Quark are very voltage depend when the voltage is low, I did some voltage/current/lux curves for the two types of heads: low voltage, high voltage (The article is in Danish, but you can still see the curve).

HKJ said:

What exactly are you calculating here?
Are you taking the rated lumen rating for the led at specified current and comparing to rated lumen for the light?

Click to expand...

That is exactly what I am doing. I derive the efficiency for the Quark R2 Cree based on rated lumens of R2 Cree vs the rated out front lumens of flashlight, and then use the same calculation to derive the out front lumens for Quark with R5 Cree.

Keep in mind that 47's under promises and over delivers, so your values are conservative lower bounds, and do not represent typical performance.

I believe you're calculating efficiency of the emitter by lumens output, rather than the efficiency of the driver based on current draw from the battery vs. current to the emitter(?), in which case there is always a point of (increasingly)diminishing returns in terms of brightness vs. drive level.