Hey Liz!

Wow -- this looks great! Since this research note was originally posted, there are a few insights that have emerged:

• Silicone doesn't seem to work very well as a sealant with the typical water bottle plastic -- it tends to leak -- and there's also the worry that some silicone sealants might include materials that would tend to corrode the probe materials over time (provide a conductive path between them). To avoid this, we're currently focusing on trying to use metal screws / bolts with rubber washers on both sides of the plastic, and a nut to tighten the assembly down -- as seen below:

• We've also realized that running the circuit at > 1.2 volts might induce electrolysis (to some unknown extent) in the liquid, affecting the measurement. So we might try to use a 555 timer with the lowest operating voltage we can find -- thus far, this one seems to run at 1V. We'll have to figure out how we might easily prototype this on a breadboard (might just required a voltage divider to drop whatever battery voltage we're using down to 1 V). Not sure how big an effect this is, or whether 1V is sufficiently low to avoid the problem. The circuit as-is does generally respond to changes in conductivity, so it's still useful as a demonstration of the concept.

Thanks for the great feedback!