This is a continuation of:
The goal of this project is to build a system that will sense water leaks in my home and automatically shut off the main water valve to my home. Chapter one gave an overview of the project. I want to thank mcb1, dougw , and beacon_dave for their helpful suggestions on the mechanics of actually moving the valve.
The project is progressing very well on all fronts. Today specifically I will show the simple sensors that been prototyped and tested so far.
We are all familiar with the customary water sensor that is used to measure whether plants or soil need watering.
https://www.sparkfun.com/products/13322
I wanted something similar that could be laid out on the floor in critical locations and would detect water quickly and definitely. The moisture detector relies on the conductivity of water which has a very uncertain resistance depending on levels of contamination. Since I would have no way to predetermine the resistance of any leaking water I decided to produce my own contamination in the form of salt. The idea is to have two copper electrodes wrapped in absorbent paper towels some of which has been soaked in salty water and then dried.
Here are two sheets of paper towels that have been soaked in salty water and are now being allowed to completely dry. The next step is to make the copper electrodes for the sensor. Short pieces of 18 Gauge copper wire were cut and wire pigtails were attached.
Here is a pair of electrodes separated by absorbent towels one piece of which has dried salt in it. The towels are wound tightly with the paper and then the paper rolls are secured with a couple pieces of light gauge wire.
These sensors will be placed in locations that are susceptible to leaks. It is hoped that even a few drops of water will soak into the towel, mix with the salt and provide a rich ionic conductor between the two electrodes.
Here is a short video of a test of the prototype sensor.
These sensors are one shot avalanche type. They should react to a very small amount of water, as you saw in the video, with a significant drop in resistance. Based on the tests of the prototype I will design the circuit to respond to any resistance that is less than 60 K Ohms. Since the dry sensors have very high resistance, (they are still OL on the 50 Meg range of the Fluke), they can be paralleled in the circuit. This will allow me to run one or two sensor lines with multiple sensors on each line. If there is a water leak problem I will have to manually check each sensor to see where the leak is, since the system will not identify a specific location as it is presently being designed. The activated sensor will have to be disassembled, dried, and wrapped with new paper before it is put back into service. I have made up 9 of them at this point and will be refining the locations where they will give the most protection.
The rest of the project is progressing well. I have decided how the mechanics of the water shut off will work and the construction and operation will be reported in Chapter three of this Blog, in the next few days.
John
Top Comments