Summary
Water tanks are often permanently installed in campers and are difficult to access. Reliable measurement of the fill level is therefore important, regardless of whether it is fresh, gray or waste water.
Previous solutions often require contact with water and are limited to an old analog display. In 2025, this will really be better.
I was looking for a good solution to meet my requirement of being able to read all sensors etc. in one interface in Home Assistant.
The different types of level measurement
I have tested these modern sensors and was unfortunately dissatisfied with all of them:
Ultrasonic sensor
This sensor emits sound waves that are reflected by the water surface. The time it takes for the waves to return enables an accurate calculation of the fill level.
The classic HC-SR04 ultrasonic sensor was out of the question here. Although it is very inexpensive, it does not have a housing and would corrode sooner or later if it came into contact with water.
- Ultrasonic distance measurement from 2 cm to 3 m accurate to approx. 3 mm
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
Better suited: The A02YYUW as a waterproof ultrasonic sensor. Quite expensive, but at least gave me quite reliable readings.
- Precise distance measurement: With a range of 3 cm to 450 cm, this ultrasonic distance sensor provides accurate and reliable measurements that are not affected by the environment. It ensures precise distance detection for various applications.
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
The basic problem: The sensor needs a clear view of the water surface, which means that condensation quickly forms on the sensor, resulting in unreliable readings. Another problem: Due to the rather narrow format of my water tank, the sound waves were reflected by the side walls, resulting in incorrect measurements.
TOF sensors (Time-of-Flight)
These sensors work in a similar way to ultrasonic sensors, but use light instead of sound. Here I have tested the TOF10120 sensor. It is officially supported by ESPHome and can therefore also be easily read out.
It is cheaper than the ultrasonic sensor, but ultimately had similar problems. The readings were good as long as no water came into contact with the sensor. In the end, I quickly discarded this idea too.
- 【Features】: Single power supply unit; UART I2C interface for device control and data transmission; Lead-free, RoHS-compliant.
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
Other measurement options
Pressure sensors measure the hydrostatic pressure at the bottom of the tank. By measuring the pressure at the bottom of the tank, the fill level can be calculated. This method is probably very accurate, but also requires a tank probe that remains permanently in the tank and therefore also comes into contact with the drinking water. I wanted to avoid that. In addition, this seemed to me to be the most expensive solution, as I could only find one sensor that was suitable for configuration with ESPHome.
- High precision: input level sensor, high-precision diffusion of silicon, more accurate monitoring of the water level.
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
Also exciting: A newer solution from VOTRONIC with a tank sensor. I don't know whether this could be read out with ESPHome. But maybe I'll test it in the future.
- Permissible operating voltage: 12V and 24V DC (suitable for continuous operation) - For mounting on the top or side of the tank - Suitable for fresh and waste water tanks
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
My experience with the various systems
In practice, ultrasonic and TOF sensors in particular have presented a number of challenges:
- Condensation will sooner or later distort the measured values.
- The moving water surface while driving makes precise measurements almost impossible.
- Intervention in the tank necessary. Without at least a small hole in the top of the tank, no measurement was possible.
The smart do-it-yourself solution: XKC-Y25-NPN Capacitive sensors on the outer wall of the tank
After a lot of experimenting, I found the perfect solution: XKC-Y25-NPN capacitive sensors in combination with an ESP for reading and transmitting to Home Assistant. This combination has been working very reliably for me for over three years. And since there is no contact with water, it is also absolutely wear-free.
- 【The durable】 liquid level sensor is suitable for the outer wall of non-metallic pipes without direct contact with liquids and is not corroded by corrosive liquids such as strong acid and alkali and is not affected by scaling or other debris.
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
How capacitive measurement works
The sensors work through the tank wall and can detect liquids in plastic, glass or ceramic containers up to 20 mm thick. The measuring principle is based on the change in electrical capacitance. When the water level reaches the height of the sensor, the capacitance changes - the sensor switches. Each sensor can therefore signal: "Yes, there is water behind the wall at my height - or not".
By combining several sensors, it is possible to say, for example, that the tank is at least 75% full.
I decided to use 5 sensors, which I stuck to the outside of the tank at 10%, 25%, 50%, 75% and 100% of the fill level. This precision is sufficient for me personally, but theoretically you could also use 10 sensors and measure in 10% increments.
In practice, it is particularly interesting when the fill level is critical, i.e. low for fresh water and high for gray water. For this reason, I also reduce the distances in the low % positions.
The advantages of my solution
- Non-contact measurement: the sensors are attached to the outside of the tank
- No contamination: No contact with drinking water
- Maintenance-free: no moving parts, no corrosion
- Smart home integration: Full integration into Home Assistant
A major disadvantage
- Measurement only in stages: The number of sensors determines how precise the measurement can be.
Reading out the values
I then connected all the sensors to a D1 Mini:
- ✅ D1 Mini NodeMCU with ESP8266-12F. A versatile board that is characterized by its technical data.
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
Power supply
I used a step-down converter, such as the one from Bauer Electronics, to convert the usual 12V in the camper to the 5V required by the D1:
- The modules are widely used in the fields of motor vehicles, photovoltaics, industrial plants, trucks, boats, agricultural machinery, camping, mobile facilities, hobby and DIY as well as home and garden.
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
How to install it
Positioning sensors on the tank
Correct positioning is crucial for meaningful measured values. Like this, for example:
- 0% (empty): Approx. 5 cm above the bottom of the tank
- 20%: At 1/5 of the tank height
- 40%: At 2/5 of the tank height
- 60%: At 3/5 of the tank height
- 80%: At 4/5 of the tank height
- 100% (full): Approx. 5 cm below the top edge
Tip: I first tested the position with masking tape. If everything fits, you can then simply use double-sided adhesive tape for mounting.
Cabling
Each XKC-Y25-NPN sensor has three connections:
- Brown: VCC (+5V supply voltage)
- Blue: GND (ground)
- Black: Signal (to the GPIO of the ESP32)
More information can be found in the data sheet.
Calibration for special tank shapes
Important tip: If your water tank has a special shape (e.g. tanks above the wheel arch with sloping walls), I recommend the following calibration method:
- Empty tank completely
- Fill step by step - use a measuring cup or canister so you know exactly how many liters are already in the tank
- Mounting - Glue the first sensor to the tank wall at e.g. 10%, etc.
In my opinion, this seemed to be the simplest and most practical solution for calculating the fill level.
Integration in Home Assistant
Once all the sensors have been installed and wired to the D1 Mini, it's time to configure the software. I use ESPHome and the following code for configuration:
esphome:
name: waterlevel-d1
friendly_name: waterlevel_d1
esp8266:
board: d1_mini
# Enable logging
logger:
# Enable Home Assistant API
api:
encryption:
key: ENCRYPTION_KEY
# Enable Over-The-Air updates
ota:
- platform: esphome
password: OTA_PASSWORD
wifi:
ssid: !secret wifi_ywbab_ssid
password: !secret wifi_ywbab_password
# Enable fallback hotspot (captive portal) in case wifi connection fails
ap:
ssid: "Waterlevel-D1 Fallback Hotspot"
password: AP_PASSWORD
captive_portal:
binary_sensor:
- platform: gpio
pin:
number: GPIO5
inverted: false
mode: INPUT_PULLUP
name: waterlevel_10
id: waterlevel_10
filters:
delayed_on_off: 20ms
- platform: gpio
pin:
number: GPIO4
inverted: false
mode: INPUT_PULLUP
name: waterlevel_25
id: waterlevel_25
filters:
delayed_on_off: 20ms
- platform: gpio
pin:
number: GPIO12
inverted: false
mode: INPUT_PULLUP
name: waterlevel_50
id: waterlevel_50
filters:
delayed_on_off: 20ms
- platform: gpio
pin:
number: GPIO14
inverted: false
mode: INPUT_PULLUP
name: waterlevel_75
id: waterlevel_75
filters:
delayed_on_off: 20ms
- platform: gpio
pin:
number: GPIO13
inverted: false
mode: INPUT_PULLUP
name: waterlevel_100
id: waterlevel_100
# Example configuration entry
sensor:
- platform: template
name: "waterlevel_0_to_100"
unit_of_measurement: "%"
accuracy_decimals: 0
update_interval: 5s
lambda: |-
std::array<bool, 5> states = {
id(waterlevel_10).state,
id(waterlevel_25).state,
id(waterlevel_50).state,
id(waterlevel_75).state,
id(waterlevel_100).state
};
// Count active sensors from highest to lowest
for (int i = 4; i >= 0; i--) {
if (states[i]) {
switch (i) {
case 4: return 100;
case 3: return 75;
case 2: return 50;
case 1: return 25;
case 0: return 10;
}
}
}
return 0; // No sensors active
If your displayed fill level jumps back and forth too frequently, the following setting in ESPHome can help:
delayed_on: 100ms- Sensor must detect water continuously for at least 100msdelayed_off: 3s- Sensor must detect no more water for 3 seconds before it switches over
Display on the dashboard
I used the Fluid-Level-Background-Card in combination with the Big-Number-Card to display the measurement values
- type: custom:fluid-level-background-card
entity: sensor.waterlevel_d1_waterlevel_0_to_100
card:
type: custom:bignumber-card
title: Water
entity: sensor.waterlevel_d1_waterlevel_0_to_100
scale: 30px
from: bottom
level_color:
- 48
- 30
- 138Automations
There are of course many different ideas. Here is an example of automatically switching off the pump when the fill level is low:
alias: Pumpe aus bei leerem Tank
trigger:
- platform: state
entity_id: binary_sensor.tank_leer
to: 'off'
condition: []
action:
- service: switch.turn_off
target:
entity_id: switch.wasserpumpe
- service: notify.dein_smartphone
data:
title: "🚰 Wasserpumpe abgeschaltet"
message: "Tank ist leer - Pumpe wurde zum Schutz ausgeschaltet."Another option: Flow measurement
What I have not yet tested is a display using flow measurement. I find this, probably very precise, option very exciting. But here, too, the problem remains: the flow meter is not really food-safe and will become contaminated over time.
- Suitable for kettles, automatic water dispensers, etc.
Last updated on 12.11.2025 / Advertising / Images from the Amazon Product Advertising API
Conclusion: Smart, affordable and reliable
With a total budget of less than 100 euros, you have a level measuring system that works without contact. I am not yet aware of any commercial solutions that offer this option.
After more than three years of continuous use, I can say that the system works absolutely reliably and offers real added value compared to conventional solutions thanks to additional automation.
Write a comment