Is this a defect in my light?

diyau: “You should measure the potential difference between the battery cells. The lithium-ion 16340 cells are rated at 3.6V whereas if your using CR123A (lithium), it is only rated at 3V. A 0.6V difference can make a huge difference in terms of light output with the Cree XML emitters because from the 3V, the emitter will dim alot, and using the lithium ion 16340 (3.6V), it will put out a bit more light.
So if your running 3V Primary Lithium cr123a batteries, you are lacking the extra voltage needed. However, if your using 3.6V Lithium-Ion CR123A, they should work like the 16340 (which is a replacement for the cr123a).

Correct me if I’m wrong here”.

Since ‘almost’ all the lithium-Ion (3.6v-3.7v rated) batteries finish charging at 4.2v - what happens when the voltage drops to 3.2v which is the voltage of a fresh cr123 primary? I’m thinking that batteries are not the problem here.

Well yesterday after Scaru suggested I do the paper clip test I just wanted to try them all again real quick since I had the tailcap off before doing the clip test and every cell worked like it was supposed to. I tried aw 16340, uf 16340 3.7vt, uf 16340 3.0vt and energizer 3.0 primary cr123a. Everything worked perfectly. Each cell allowed me to cycle through every mode without so much as a flicker. I did notice a difference in brightness between them. The imr and the cheapy 16340 were noticably brighter than the primary and the cheapy rcr.

Nomial voltage of the 18650 battery cell is rated at 3.6/3.7V. An protected 18650 battery will stop working at about 2.75 volts for protection against over-discharge. When the voltage drops to 3.2V, in theory it should have similar brightness output at the emitter because LEDs do not have a constant resistance, rather as the voltage increases, so does the current that will flow through the LED. Taking into account that the visble light emitted from LEDs is due to the current flowing through the LED, as the voltage increases, so will the brightness.

Also, what I’ve noticed with many batteries when they are not under-load and you measure the voltage of the battery it will usually display a higher voltage than that of when the batteries are measured when the battery is on load. I think, when the battery is on load, the voltage decreases due to the load…