I decided to do a full runtime test so that I would know exactly what I can expect when using a protected rechargeable RCR123 cell and running the NovaTac 120P at Max output:
NovaTac 120P set to use a rechargeable 3.7 volt cell per NovaTac instructions. Using a fully charged (and reasonably new, maybe 8 charging cycles) Battery Station 900 mAh RCR123 cell that had been sitting in a 65 degree room for about 4 days. Battery reads 4.2 volts to start the test (using a very old, but reliable Radio Shack analog 'Micronta' mulit-tester).
I'm using a Luna-Pro SBC analog light meter with a fresh 9 volt battery to continually check the light level during the test. The light is sitting on the table facing the meter which is 1 foot away. Ambient temperature is 58 degrees. Times are accurate to within 15 seconds either way.
My lumens numbers were not actual readings. I used NovaTac's published rating of 120 lumens for the top level, and after observing and measuring very precise 1 stop drops each time the light level dropped (which equates to a halving of the light output), I believe the numbers to be at least representative of what I can expect.
12:10 am.........begin test with light set to highest level.......... 120.0 lumens
12:40..............1st drop...-1 stop on light meter...................... 60.0
12:42..............2nd drop..-1 stop.......................................... 30.0
12:43..............3rd drop..-1 stop.......................................... 15.0
12:44..............4th drop..-1 stop............................................ 7.5
12:45..............5th drop..-1 stop............................................ 3.8
12:47..............6th drop..-1 stop............................................ 1.9
12:52..............7th drop..-1 stop............................................ 0.94
12:53..............8th drop..-1 stop............................................ 0.47
12:57..............9th drop..-1 stop............................................ 0.23
1:00..............10th drop..-1 stop............................................ 0.12
1.01..............11th drop..light begins low battery warning pulse.. 0.08
1.10 am........ test terminated. Turned off light, turned back on and set to lowest level (.08), light began low battery warning within 10 seconds, and kept going. Turned light off. Battery tested at 3.2 volts.
Other observations: I kept close tabs on the light meter and noticed very little fluctuation of the needle during the 1st 30 minutes of the test, very flat and even. The meter was also flat between level drops, with no visible fluctuations observed, except at the very, very end of the test. What happens when the light goes into low battery warning mode is that it is mostly on, with a slight off/on every second of so.
After the light had been in the low battery warning mode for 7-8 minutes, I did observe a very slight, uneven, but not terribly prominent flicker in the LED. After one minute of that I terminated the test.
Comments: I'm not sure exactly how to interpret the results, as I've not had that much experience with this type of thing. Maybe turbodog would be willing to offer some insight into whether these readings have any relevance to the discussion at hand.
Of course, this test applies only to how the light handles rechargeable cells, so I think I need to do a similar test using a primary cell to see exactly how much juice my light will suck out of one of those, and what happens with the regulation when the primary begins to become depleted.
Personally, I was pretty happy with the fact that I got another 20 minutes of useable, dimming but still regulated light after my initial 30 minutes of excellently regulated pedal-to-the-metal full output, and that the light dropped the output levels before the battery's protection circuit kicked in, so that I never did have the light actually shut off completely. I like that. It was also cool that even the low battery warning, which seemed like it would go for a while (for sure the 9 minutes I observed), would still be quite usable if absolutely necessary.
BTW, I know the internal battery protection circuit of this particular battery is functional as it has worked in another light.
As to the slight bit of flickering I noticed after 7-8 minutes of the low battery warning.... since NovaTac suggests that the battery should be changed well before encountering the low battery warning, I feel the light performed well within it's designated parameters, and although I would rather have not had that occur, I can certainly live with it as an acceptable quirk.
Maybe someone with more experience can tell me how this compares with what the HDS did in this situation. Does 3.2 volts seem too high a point for the light to be going into the battery warning for a rechargeable? It might be nice if there were a menu included in the options section to let us select different values for the low voltage warning. Let me know what you think.