One of the applications I wanted to include in this road test is capacitance water level sensors because you can get an accurate depth measurement for almost no cost as long as you have a meter that can measure capacitance The Keysight meter does measure capacitance and is useful for testing and calibrating such sensors The physics behind why this technique works is a little trickier than it looks but I want to provide a bit of a feel for the technique without delving into math Basically the capacitance between 2 conductive plates is proportional to their area and the dielectric constant or relative permittivity of the intervening material It is also inversely proportional to the distance between plates
If we have a 2 conductive capacitor plates immersed in water, where one plate is bare metal and the other is insulated with a film of PVC plastic, a capacitor is formed with 2 dielectric materials between the plates - PVC and water. The permittivity of PVC is 3.4 and the permittivity of water is 80.4. The way I think of the combined permittivity is it is like the combined capacitance of a large capacitor in series with a small capacitor, so something like 3.3 in this example, which is very close to the smaller permittivity. This low permittivity doesn't help make capacitance in water high, but what the high permittivity of water does is effectively move the capacitive charge of the bare plate close to to the PVC insulation, reducing the effective plate separation distance to just the thickness of the insulation. This effect does make the capacitance dramatically higher.
If the 2 plates are in air the situation is reversed because the PVC has a higher permittivity than air, but it only effectively moves the plates closer by the thickness of the insulation.
If the plates have significant physical separation and are partly in air and partly immersed, the capacitance will be completely dominated by the immersed portion. When making a sensor that uses this property, the plates can be metallic strips, round wires, or rods of any uniform cross-section. To get higher capacitance from the sensor, make the wire diameter larger and the insulation thinner, or use insulation with a higher dielectric constant. Here are the dielectic constants of a few wire insulation materials plus air and water:
Here is a little video demonstrating the Keysight directly measuring water depth via capacitance:
If you want to develop a really accurate capacitance sensor system, you may need to minimize the influence of people, objects and stray electric fields on the sensor capacitor. There are a number of ways to tackle this:
- if you are able to put the probe into the water, the ground electrode (capacitor plate) can be a stainless steel tube. The other plate is an insulated wire or rod right down the middle of the tube. The tube acts like a Faraday shield.
- If you need the sensor probe to be outside the tank, you can partly shield it in a similar way - use a ground strip on either side of the sense electrode and put a conductive cover over the sense strip connecting to both ground strips.
In both cases there should be enough air around the sense conductor to keep the air capacitor from being too significant.
Another technique is to employ out-of-phase signals. This is explained in an app note from TI.
In summary capacitance is a great way to easily inexpensively and accurately measure water levels of any depth and the Keysight is a great asset in working with such sensors Especially because it will not be harmed if you happen to drop it in the water.
Links to other installments of this road test: