Decrease water detection time

[ntlhui]

50 ns fall time is great. We should test how low we can push WATER_DETECT_EN_TIME_US and still get reliable wet/dry detection. 1 µsec = 1,000 ns (obviously not telling you something you don't know!) so do you think #define WATER_DETECT_EN_TIME_US 1 would likely work?

In my understanding, the lower we can push that value, the less electrolytic corrosion we'll have and the more durable that sensor will be over time.

Thanks, Phil

[ntlhui]

Test Procedure:

With Rev C boards, do 100 trials at each of the following settings for WATER_DETECT_EN_TIME_US

• 1000
• 800
• 600
• 400
• 200
• 100
• 50
• 25
• 12
• 6
• 3
• 1

In each trial, short the wet/dry sensor. Wait at most 5 seconds for a water detect indication.

Requirement is > 99% accuracy.

[SUPScientist]

Recommend doing with a variety of resistances meant to emulate those expected at a range of natural water conditions (e.g., conductivity = conductivity(Temp, Salinity)) as well as the dead short.

[ntlhui]

Assigning to REU 2023 students for evaluation