next topic »

[on:]

We had a bit of rain here recently, so the next project is a sensor to warn when the water level gets too high in the sump pump crock.

Will be doing some experimenting, but thought I might as well ask first.  I've checked out the method used by Botanicalls, and have seen similar approaches one or two other places, so that might be one place to start.  Years ago, I designed a TTL-based water level controller that used LM1830 Fluid Detectors.  Those worked well, but are no longer made, and I haven't been able to find anything similar.  Maybe that's overkill anyway.  A float switch is another approach, but the no-moving-parts solution is appealing.

All input appreciated!

[Reply #1 on:]

With water level do you mean ground water level or water above ground level e.g. flood?

water level can be measured with some hollow pipe in which some cork thing can float - sort of bobber (like fishers use)

water can be measured by electric resistor ladders, the more resistors under water the lower R will be

[Reply #2 on:]

With water level do you mean ground water level or water above ground level e.g. flood?

Water above a certain level.  The sump pump sits in a cylindrical crock that is below the basement floor level maybe two feet or so.  Under normal operation, the pump will start when the crock is maybe half full.  All I'd need is a single sensor positioned a bit above the high water level at which the pump normally starts.  So this would indicate pump failure.

[Reply #3 on:]

I drafted a design for a water level indicator a while back, but never got around to actually build it. The plan was to use a low pressure sensor (say 0-1 psi full scale) and mount the sensor to one end of a pipe (plastic tube). The other end of the pipe is open and goes into the tank. When water level rises, the air column trapped inside the pipe will get compressed and so pressure will increase. The pressure (relative to open air) will be proportional to water level.

[Reply #4 on:]

I drafted a design for a water level indicator a while back, but never got around to actually build it. The plan was to use a low pressure sensor (say 0-1 psi full scale) and mount the sensor to one end of a pipe (plastic tube). The other end of the pipe is open and goes into the tank. When water level rises, the air column trapped inside the pipe will get compressed and so pressure will increase. The pressure (relative to open air) will be proportional to water level.

Interesting.  Would there be any concern about the air column slowly going into solution in the water over time, thereby changing the sensitivity?  I have a laundry tub pump that works on the same principle, but the bottom of the sensing tube is out of the water every time the tub is pumped out.

[Reply #5 on:]

There are a lot of sexy methods, but remember that crock is going to be very dirty and perhaps have oils, etc. in it; these are going to quickly foul any delicate sensors.  You'd be hard pressed to do better than a basic float switch made for bilge pumps in boats--these have to be super reliable and tolerate of all sorts of gunk.  Another option would be a tethered tilt switch like that on many sump pumps; you could build something like that yourself.  Any switch that has close-tolerance moving parts is risky because gunk can cause sticking and failure.  Good luck!

[Reply #6 on:]

"Bubblers" are often used for sewage.
http://en.wikipedia.org/wiki/Level_sensor#Air_bubbler
You should be able to build something for less than $50 using a differential air pressure sensor and aquarium air pump.

[Reply #7 on:]

Have you considered using the PING))) ultrasonic sensor?  If "no moving parts" is more important than cost, this is a method to consider.  In industrial applications water level is frequently determined with ultrasonics, but if there is likely to be turbulence, it might be a good idea to use a "stilling well" to keep the surface of the water directly underneath the PING))) as quiet as possible.  You also want to make sure that the PING))) is mounted higher than any possible water level.

[Reply #8 on:]

SparkFun has a Liquid Level Sensor that I was looking at a little while ago when I was considering making a 'smart' water bottle, it may be helpful to you here.

http://www.sparkfun.com/products/10221