On Mon, 20 May 2019 18:56:59 -0700, Jeff Liebermann
wrote:

My guess(tm) is that these
lights used the usual AMC7135 current regulator IC which is rated at
350ma from 2.7 to 6.0v.
https://www.electroschematics.com/wp-content/uploads/2016/04/acm7135-datasheet.pdf

Duh. Maybe I should have looked at the data sheet before I continued
to guess. See graph on Pg3 "QUIESCENT CURRENT vs. SUPPLY VOLTAGE".
The graph only goes to 4.1v which is a subtle hint that 6v might be a
bad idea. They could have used two or three AMC7135 chips in parallel
to obtain the necessary current handling, but then the leakage current
would be 2x or 3x as high. Killing the batteries with this leakage
current in 3 months is quite possible.

However, much depends on whether the Duracell Durabeam Ultra 350 uses
an AMC7135, and whether power is always applied to the AMC7135. I
can't tell from here without disassembling the flashlight, reverse
engineering the schematic, and making a few measurements.

Deconstructing a flashlight:
https://kandepet.com/deconstructing-a-flashlight/

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558