This is a question regarding the measurement of a fluid level (in my case water) using a capacitive sensor. Capacitive sensors are a common topic, but I have one specific question: It is necessary for both electrodes to be insulated from the fluid?

This is the basic concept that I am intending to implement: Capacitive Fluid Level Sensor (Instructables)

In that sensor design, both the inner and the outer electrodes are insulated from the fluid. (Note the remark that the sensor only worked correctly when the outer electrode was the ground pole.

Instructables design Original Instructables design

I would like to implement a simplified design where the outer electrode is an aluminium tube, and the inner electrode is sealed inside in a PVC pipe. (obviously with an electrical connection though!)

Concept design with cutaway Concept design with uninsulated outer electrode

Hence my question: Is it necessary to insulate the outer electrode? What effect will it have on the capacitance of the sensor if the dielectric medium (i.e. the water) is effectively connected to the outer electrode? Bear in mind that we are only trying to achieve a measurable change in capacitance.

The following topics here are of interest:

  1. Water Level Sensor Circuit (Curd's answer)
  2. Water level capacitive probe topology
  3. Water level: BIG capacitive level sensor with steel bars?
  • \$\begingroup\$ Proposed dimensions for interest. Probe length: 1000mm to 2500mm, Outer electrode ID: 28mm, Inner electrode OD 16mm, Electrode spacing: 3mm air/water + 3mm PVC \$\endgroup\$
    – Nicolas
    Jun 21, 2018 at 22:30
  • \$\begingroup\$ If you introduce a metal into water that is connected to other different metal in same body of water galvanic action will corrode the more active metal in time. Not worth the pain if you want long term reliability. Fine for testing but characteristics will change when additional insulation is later added. \$\endgroup\$
    – KalleMP
    Oct 16, 2019 at 6:55
  • \$\begingroup\$ @KalleMP - Note that the inner electrode is insulated. There should be no galvanic action because there is no electrical path between the two electrodes. \$\endgroup\$
    – Nicolas
    Oct 17, 2019 at 14:03

1 Answer 1


There is no electrical reason to add a gap from water to any electrode except for rust. Any extra plastic (4) or air (1) gap reduces sensitivity. But then so does rust or lime or calcium buildup if it insulates the metal at some significant thickness .

Perhaps pouring inside a rust-proofing paint and turning it to coat the interior metal surfaces and curing hard may slow down the rust a bit. Then pour out the excess . Might be messy.

The capacitance will be proportional to the mutual surface area of two electrodes and the gap should be smaller for low impedance (~50 ohms).

The interior pipe or rod needs to be coated as well as possible for the same reason and centred such as an anodized aluminum pole.

Then use a Schmitt Inverter as a Relaxation Oscillator to measure the height of water by pulse interval time 1/f . You could even slow it down with some CD4040 counter and LED.

  • \$\begingroup\$ Thanks @Tony-Stewart-older-than-dirt. The electrodes are aluminium, so significant corrosion is not expected. Also, the inner electrode will be completely sealed to avoid continuity with the water. \$\endgroup\$
    – Nicolas
    Jun 21, 2018 at 22:16
  • \$\begingroup\$ Regarding circuit design for actually performing the measurement - that is another matter which is not considered in this question. Different solutions can/will be tested and optimised. \$\endgroup\$
    – Nicolas
    Jun 21, 2018 at 22:21
  • \$\begingroup\$ Ok but it will react depending on pH and O2 content of water and coating process. \$\endgroup\$ Jun 21, 2018 at 22:41
  • 1
    \$\begingroup\$ Water supplier in Toronto suggests all pipes will get buildup unless the water is “softened” with salt. Preferably . KCl \$\endgroup\$ Jun 21, 2018 at 22:56
  • 2
    \$\begingroup\$ There is no electrical reason to add a gap from water to any electrode except for rust. Obviously if BOTH electrodes are not insulated the the water would effectively short the capacitor, rendering it useless. \$\endgroup\$
    – Nicolas
    Jul 11, 2018 at 18:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.