In order to measure the change of water level as small as 0.01mm, I made a water level sensor composed of two concentric metal cylinders inside and outside (outer surface insulation treatment), the diameter of the outer cylinder is 20mm, and the diameter of the inner cylinder is 6mm.The inner cylinder is connected to the CIN1 (+) pin of the AD7746, and the outer cylinder is connected to the EXCB pin.The height of the submerged part of the sensor is calculated from the variation of the measured capacitance.But after testing, two problems were found:
1. The capacitance value measured by the AD7746 is not relatively stable, but gradually increasing under the condition that the depth of the sensor submerged in water remains unchanged.Appear this kind of phenomenon, I use the EXCB Excitation Voltage (Excitation Voltage Level) to ±VDD/2. However, if the excitation voltage is changed to ±VDD×3/8, the phenomenon of continuous increasing will be alleviated.What may be the reason for this and how can we avoid it?
2. I use metal or glass measuring cylinder to hold water. In the process of measurement, as long as people are close to the measuring cylinder, the measured capacitance value will change greatly, reaching about 0.4pf. According to the calculation, when the water level changes by 0.01mm, the sensor I used only brings about a capacitance change of 0.04pF. Hand interference is 10 times higher than water level changes, making measurement accuracy impossible to guarantee. Is there any way to reduce the interference of this hand? I tried to add a grounding shielding layer to the outer cylinder of the sensor, which was effective when the sensor was not immersed in water, but the interference was not reduced after the sensor was immersed in water.
Thank you very much for your guidance!
The above problem occurred when I used eval-ad 7746ebz evaluation board and supporting software.