I must admit a few weeks ago getting the enclosure wet for the first time, at 100 feet the leak detector sounded. In short a small fiber, most likely from a lens wrap, was on the main o-ring. I missed it. That said there was only a few drops of water in the bottom of the housing and the leak detector picked it up doing its job. The fact that there is an audible alert ensured that I immediately reacted to the situation, heading for the surface.
That said after rinsing out the enclosure and in theory drying it, I locked it up for the night without desiccant. At some point in the middle of the night the alarm went off again. It would appear enough residual moisture was present, most likely between the wires and the heat shrink tubing used at the end of the sensor circuit board.
Note I updated the firmware on Instructables to include a power savings mode which should extend the battery life to a year of standby time. The leak detector in the power savings mode polls the sensor every four seconds. When on travel, test at the start of the trip. The detector shifts to polling the sensor once per second, shifting back to power savings mode after two weeks. The older source code file is still present on Instructables which should yield three months of continuous monitoring (e.g. once per second polling), using a CR2032 battery. The updated file is what was in use when the small flood occurred, thus I'm rather sure the code is working.
Edited by vsc, 29 August 2017 - 04:25 AM.