Diy - detecting water leaks or water level....

found  this great post on a homemade water level sensor using a pvc (or whatever pipe) and some resistors.  looks promising.  they used it with an arduino so it should would with the webcontrol.
 
http://lifeboatfarm.wordpress.com/2009/12/28/arduino-water-level-gauge/

I don't want to appear unnecessarily negative, but:
 
Might appear to work for a little while under ideal conditions, but it won't work for long.
Pure water is a poor conductor of electricity for a start.
Resistors like that are not designed for high humidity, far less permanent immersion!
The DC bias will eventually "eat away" the conductors
Any ground leakage between your WC power supply/ground will give you wildly varying numbers.

oh i agree.  just thought it was interesting.  some of these diy approaches are very good.  others - so so.
 
this site is where he got the idea and he has it mounted to a metal pipe (or with an aluminum strip on opposite side of resistors)
 
Why couldn't you use that metal tubing they use for fridge water dispensers (the non copper)

or heres the other thing.  could you electroplate a zinc coating onto the resistors?  would that make them last longer?  There are plenty of diy ways to do that.

Resistive sensors are simply NOT suitable for trying to measure how much water is in a tank.
There are too many variables. Particularly contaminants in the water (especially salts).
 
Capacitive works better. But still not great.
If it is practical to access the bottom of the tank, an inexpensive pressure transducer is quite effective.
 
What I ended up with in my own 105,000 liter tanks is an ultrasonic *distance* sensor. Measuring from the tank top to the surface of the water is highly reliable, accurate, stable and long-lived. Running 24/7 for 8 years now without a glitch. (OK, it's more expensive than the system shown - but there are VERY cheap ultrasonic distance measuring kits for the arduino etc that could do the job too, and could work with the webcontrol I'm sure!)

If you use water and metal foil sealed in the plastic tube to form a capacitor, then connect this capacitor to a frequency generator. Then use the frequency counter in WebControl, you can actually measure the water level that way.

I did the exact thing (resistive) a decade ago with a basic stamp to measure the draft of coal laden barges in a Caribbean sea port. In practice it works longer than you would think. My resolution was 1 foot (10 foot range) and I'd typically get 10-12 months out of them before they'd fail in the salty sea water. The nice thing was that when they failed, they did so predictably so it was easy to design around that. Also they only cost about $20usd each to make and parts could be sourced locally (Colombia) so we were never low on inventory. Swapping them out just became part of the routine PM schedule for the barges, of which there were about 100.

Why not just use a few float sensors?  I have 3 I put into my sump crock.  using 3 I can calculate how fast water is rising to determine how much time I have to get home to fix water problems.
 
Similar to this one on the homeseer forums:

 
--Dan

That is true, if you use lowest sensor for switching pump off, middle sensor for switching pump on, and top sensor for sending out warning notice, with few lines of PLC code, it can automatically do the work.

even if you do not use this for control of the pump, what I do is monitor that the pump is GOING to go on (the lowest is set to just under the height of the sump's trigger point).
 
The middle is, something is wrong with the pump, and the difference in time between the lowest and the middle is XX, which means you have XX amount of time to get home and resolve this.  The middle is mounted 1/2 way up the crock, between the lowest sensor and the top of the crock.
 
The top is you have 1" of water before the crock is overflowing...better fix it now!!!

I see, you have lowest sensor to turn on pump, if it got to the middle sensor, then pump must have problem.
You could have a water pressure sensor to turn off pump. When pump output has no pressure, turn the pump off.

How about a pressure xmitter. 0.43psi per foot. You just have to find the correct one for your application.  http://sensing.honeywell.com/index.php/ci_id/142173/la_id/1/document/1/re_id/0  Heres an idea of what to search for. Try Digikey. For inground or tanks that have no accesible plumbing on the bottom, you could always pot the board with waterproof potting and then leave the pressure port opn and vent the other to atmosphere. Drop the unit to the bottom of the tank. Keep the vent clean and dry. As for the DITY part, you'll get to solder it to your board and then pot it.

todster said:

How about a pressure xmitter. 0.43psi per foot. You just have to find the correct one for your application.  http://sensing.honeywell.com/index.php/ci_id/142173/la_id/1/document/1/re_id/0  Heres an idea of what to search for. Try Digikey. For inground or tanks that have no accesible plumbing on the bottom, you could always pot the board with waterproof potting and then leave the pressure port opn and vent the other to atmosphere. Drop the unit to the bottom of the tank. Keep the vent clean and dry. As for the DITY part, you'll get to solder it to your board and then pot it.

Click to expand...

 
 
I have used freescale pressure transducers connected to my cai webcontrol board to measure water depth.  It works.  I use the 1.45psi model and I need to measure about 30 inches.

Dr. Lou set up this pump control with the sensor he mentioned and WebControl running for years.

CAI_Support said:

Dr. Lou set up this pump control with the sensor he mentioned and WebControl running for years.

Click to expand...

 
Hold up there, more like a little over 1 year, not years.  Check back with me in 2015 and I'll let you know.
 
I am potentially pushing my luck withe the transducer because it is located below the water level, meaning that water could get into the sensor and contact the guts of the device.  In fact, I am quite certain water has gotten inside of the devices.  The specs say you shouldn't do that.  But they keep working.  I actually have 6 transducers that are all probably wet on the inside.  None have failed in over a year of being wet.  I also tried a GE sensor, it failed early on.  I don't know why, but I suspect it was the water.  I never tried a second GE model.
 
If failure is not an option, I suggest using two sensors in parallel and set some sort of alarm if the two don't agree.  Failure would probably result in a grossly abnormal reading such that you could write your plc code to ignore it, use the other sensor, and alert you to the problem.
 
In fact, no matter what system you are using, if failure is not an option, I suggest some method of backup.  Perhaps just an alarm system water contact from GRI set high enough that it should never get wet.  They are pretty much passive devices that if they get wet they close the circuit.  This could easily wire up to the digital inputs on the cai, while the pressure transducer goes to the analog, giving you a precise level.