Can Infrared Sensors Effectively Monitor Water Level in a Tank?

[cyeokpeng]

I am not an experienced engineer as I am still studying in the university. I wonder if I want to design a system which can monitor the water level of a tank, what transducer can I use? Infrared sensors?

Let's say I use infrared sensors, and I use a 10-series LED IC chip to determine the water level, (LED light up means full, 5 LED light up means half full, and 1 LED light up means going to be empty, and 0 LED light up means empty). Does that mean I have to use 10 infrared sensors located at different levels of the tank? Do anybody has more efficient idea to design such a system without using 10 infrared sensors?

Byt the way, how does infrared sensors work? When it detects that the water level has gone past it, what signal does it produce which can be fed back to the microcontroller or circuit to output the necessary 10-series LED strength indicator?

Sorry about the bombardment of questions, because I just want to add additional features to my project on remote controlled chemical tanker to punch in more quality in the design.

Thanks

 

[chroot]

Most infrared sensors will not tolerate being submerged in water; you'll destroy them.

The most common solution is a float attached to a rheostat or some other device, but that's complex and mechanical.

You might want to consider something like an ultrasound range finder mounted in the top of the tank, or a simple scale under the tank to weigh its contents.

- Warren

 

[Integral]

Floats have a nasty habit of sinking and are not real good for a continuous level feed back.

We use a ultrasonic detector to give an analog measurement of a liquid level. The sensor is at the top of the tank looking down at the surface. Fortunately the liquid (TMAH, it may look like water...but it aint) is stored in a rectangular tank so we can read out the volume directly.

Another way that is pretty common is to use capacitive sensors on the outside of your tank, and yes you need one of these at each level you wish to sense. You may be able to do a similar arrangement with the IR sensor looking through the wall of your tank (This will not work with all tanks).

 

[cyeokpeng]

Floats have a nasty habit of sinking and are not real good for a continuous level feed back.

We use a ultrasonic detector to give an analog measurement of a liquid level. The sensor is at the top of the tank looking down at the surface. Fortunately the liquid (TMAH, it may look like water...but it aint) is stored in a rectangular tank so we can read out the volume directly.

Another way that is pretty common is to use capacitive sensors on the outside of your tank, and yes you need one of these at each level you wish to sense. You may be able to do a similar arrangement with the IR sensor looking through the wall of your tank (This will not work with all tanks).

How does capacitive sensors work? I mean what physical theory do they use to detect the water level as the water rises past these sensors? It seems from the description that it has something to do with capacitance, but still how can capacitance of a capacitor change when water level rises past them? Doesn't seem logical to me though.

 

[NoTime]

The water will change the value of the capacitance.

IIRC there is also a coaxial sensor.
Basically a hollow conductive tube with a wire running down the center.
A pulse will get reflected at the fluid level because of the dielectric change.
I don't remember what the design constraints were on these.

 

[Gokul43201]

How does capacitive sensors work? I mean what physical theory do they use to detect the water level as the water rises past these sensors?

The capacitor itself is a cylindrical (coaxial) capacitor standing vertically in the tank.

The dielectric constant of water is about 80 times that of air. This increases the capacitance by a factor of 80 if completely submerged. The capacitance is a nearly linear function of height between empty and full (assuming end effects have been engineered out).

 

[Integral]

The capacitor itself is a cylindrical (coaxial) capacitor standing vertically in the tank.

The dielectric constant of water is about 80 times that of air. This increases the capacitance by a factor of 80 if completely submerged. The capacitance is a nearly linear function of height between empty and full (assuming end effects have been engineered out).

Actually they are frequently mounted horizontally outside of the tank. Of course this works best with plastic tanks. The same principle applies. When the fluid level reaches the sensor, the change in capacitance creates a small current flow.

I have not seen a veridically mounted capacitive sensor, that seems like it would work pretty good as long as you were only dealing with a small change in level. How far from the surface can you be and still resolve meaningful changes?

 

[FredGarvin]

All of our large tank gauging is done by magnetostrictive floats. They are very good and with temperature compensation, have very good resolution. These floats are vertically mounted and do span the entire height of the tank. On a 10,000 gallon tank, we can see the volume down to a total volume of about 100 gallons.

 

[Gokul43201]

Actually they are frequently mounted horizontally outside of the tank. Of course this works best with plastic tanks. The same principle applies. When the fluid level reaches the sensor, the change in capacitance creates a small current flow.

I have not seen a veridically mounted capacitive sensor, that seems like it would work pretty good as long as you were only dealing with a small change in level. How far from the surface can you be and still resolve meaningful changes?

Okay, I guess you are talking about a two-state level sensor; that tells you whether the water level is above or below a certain fixed point - just like a float sensor.

What I'm describing is the design of a liquid level sensor that tells you the position of the liquid level at any time. Of course, I should add that I've never heard of this design being used for water - there would definitely be concerns of shorting the electrodes, which may not be easily resolvable. These sensors are frequently used to monitor liquid levels in non-polar and cryogenic liquids - I recently built a liquid helium level sensor, and that's about as hard as they get in terms of signal; the dielectric constant of liquid helium is only 4% higher than that of the cold gas.

Of course, the design is ridiculously expensive and excessive if all you want is to know is when to turn off a pump (or something of that nature).

 

[cyeokpeng]

I have just realized that putting IR sensors outside the tank is of no use too, even though the tanks are made transparent. IR cannot pass through most transparent materials.