Water tank level sensor: 4–20mA current loop, ESPHome, and Home Assistant
Our house water comes from a rain water tank. I built an ESPHome device, based on an Olimex ESP32-PoE2, to interface a 4-20mA current-loop liquid pressure sensor, and keep track of the water level in Home Assistant.
#clkMode #electronics #ESPHome #HomeAssistant #OlimexESP32PoE2 #waterTankSensor
In search for a better filter
https://blog.narf.ssji.net/2025/05/10/in-search-for-a-better-filter/
Linear vs. Polynomial calibration to get water level from sensor reading are pretty much the same.
Low-pass filtering too slow to react to actual changes
Works well enough for immediate reading and daily consumption aggregate
Too noisy for hourly readings and other event detection
Water delivery took hours to register the full scale
Kalman?
I thought it was for sensor fusion, but it also works for single-sensor filtering.
Good videos: https://www.youtube.com/watch?v=HCd-leV8OkU https://www.youtube.com/watch?v=qCZ2UTgLM_g
Home Assistant has some integration for Kalman filters, but they seem unmaintained https://community.home-assistant.io/t/kalman-filter/58483
ESPHome: https://esphome.io/components/sensor/combination.html
Dug out and translated the sheet that came with the current loop liquid pressure sensor. It looked like an ad, but actually contained useful info on the characteristics of the sensor.
Likely useful to calibrate the Kalman filter
Also has a “debugging” / calibration method, which would allow to get a wider range.
Adding a Kalman-filtered entity in ESPHome
Error on the sensor: Got the process stddev from overnight data from Home Assistant (when the water level is expected not to change), and chucked it into LibreOffice’s STDEV function: ~600L
Standard deviation on the process: change per second. For lack of a better estimate, using the baseline of 400L/day use: 400/(24*3600) = 0.00462963
This seem very low as “A low value here will place high importance on the current state and be slow to respond to changes in the measured samples.”
Also trying a larger value for likely changes, e.g., one tap running for a minute at 9L/min: 0.15
[…]
Linear vs. Polynomial calibration to get water level from sensor reading are pretty much the same. Low-pass filtering too slow to react to actual changes Works well enough for immediate reading and daily consumption aggregate Too noisy for hourly readings and other event detection Water delivery took hours to register the full scale Kalman? I
Finagling the water sensor parameters
https://blog.narf.ssji.net/2025/02/22/finagling-the-water-sensor-parameters/
Now that I have the hardware and ESPHome OtA setup for the water tank sensor, I can play a bit. The ESPHome dashboard did not detect the remote sensor until I set ESPHOME_DASHBOARD_USE_PING=”true” (or any non-empty value) in the docker-compose environment (as hinted by https://community.home-assistant.io/t/esphome-devices-offline-but-only-for-dashboard-status/403141)
First, with taking all possible built-in sensors for a spin:
uptime
internal_temperature (bad luck this one gives a well-known invalid value)
Then, I can try to find the right filter https://esphome.io/components/sensor/#sensor-filters. There are many combination just with the basic filters, so let’s start simple.
[…]
Now that I have the hardware and ESPHome OtA setup for the water tank sensor, I can play a bit. The ESPHome dashboard did not detect the remote sensor until I set ESPHOME_DASHBOARD_USE_PING="true" (or any non-empty value) in the docker-compose environment (as hinted by https://community.home-assistant.io/t/esphome-devices-offline-but-only-for-dashboard-status/403141) First, with taking all possible built-in sensors for a spin:
Putting it all together
https://blog.narf.ssji.net/2025/02/15/putting-it-all-together/
I’m finally putting the water tank sensor together for a test
PIN 35 on the ESP32-POE(2) is not Vcc, but the battery level, when a jumper is in place. Rather than putting the jumper, I created a short cable from the 3.3V pin to pin 35. It’s at 3.14V when powered over Ethernet.
the adc needs attenuation: auto, which works pretty well for both VCC and the current loop sensing resistor (the default is 0dB, which doesn’t)
Found a nice Lovelace card for Home Assistant lovelace-fluid-level-background-card
Also using icon mdi:hydraulic-oil-level, which seems more telling than the plain mdi:water
Crimped and patched the CAT6 cable to the water tank. There are two standards for Ethernet connectors, but Australia mostly uses T-568A (green first)
Updated the sensor config to use Ethernet for connectivity, in addition to power (rather than Wi-Fi)
The ESP32-POE2 uses GPIO0 for the clock, same as the first version.
Does DHCP by default, like Wi-Fi did.
Took a leap of faith and OTA’d it. It still work. Pfew.
esphome upload –device water-tank-sensor.example.net water-tank-sensor.yaml
Also moved passwords and related parameters to .gitignored secrets.yaml
Got some data in! Pretty limited, so far, with one calibration point. Re-enabled the moving average filter to reduce the noise. Now checking with air pressure to see if changes have much of an impact.
Next steps: add calibration points, maybe add an air pressure sensor if needed, finish the PCB as a generic base for current loop sensors.
[…]
I'm finally putting the water tank sensor together for a test PIN 35 on the ESP32-POE(2) is not Vcc, but the battery level, when a jumper is in place. Rather than putting the jumper, I created a short cable from the 3.3V pin to pin 35. It's at 3.14V when powered over Ethernet. the adc
Clamping and back to ESPHome
https://blog.narf.ssji.net/2025/02/14/clamping-and-back-to-esphome/
Obtained a better understanding of clamp diodes thanks to Alan. When they start conducting, they prevent the current from getting into the ADC. So a clamp diode from the ADC line to +VCC will protect from over-voltages above Vcc+Vd, the same from GND to the ADC would protect from voltage reversals from -Vd. Zeners reverse voltage would provide a tighter clamping, but a 1N4148 would at least do the above
I have plenty of 1N4148, and found one with a notably low Vd at around 0.5v.
Used that as between the ADC input and Vcc.
Put the water tank sensor together and played in ESPHome
No Ethernet connection (power over USB for now), so no 24V for the sensor for real
Could test ADC reading using the adc platform for the sensor component
[…]
Obtained a better understanding of clamp diodes thanks to Alan. When they start conducting, they prevent the current from getting into the ADC. So a clamp diode from the ADC line to +VCC will protect from over-voltages above Vcc+Vd, the same from GND to the ADC would protect from voltage reversals from -Vd. Zeners reverse
Re-working the 4-20mA adapter
https://blog.narf.ssji.net/2025/02/02/re-working-the-4-20ma-adapter/
I finally got back to my water-pressure meter. As I fried the previous esp32-poe, and I didn’t want to do this to my brand new esp32-poe2. I’ll need better protection on the ADC input, and I’m flat out of the right Zeners for clamping.
I could, however, rewire the adapter around a 150 Ω parallel resistor, and a 1 kΩ into the ADC. Testing this in isolation into a jar of water show hopeful progress!
Next, I need more Zeners.
[…]
I finally got back to my water-pressure meter. As I fried the previous esp32-poe, and I didn't want to do this to my brand new esp32-poe2. I'll need better protection on the ADC input, and I'm flat out of the right Zeners for clamping. I could, however, rewire the adapter around a 150 Ω parallel
ESP32-POE-two
https://blog.narf.ssji.net/2024/11/09/esp32-poe-two/
Mouser added the ESP32-POE2 to their catalogue, but with no stock
Ended up ordering straight from Olimex https://www.olimex.com/Products/IoT/ESP32/ESP32-POE2/open-source-hardware
Playing around with esphome, it seems to work with the same config as the esp32-poe
the EXT1 header has both VPP (my 24v), and GPIO36 (my ADC) […]
Mouser added the ESP32-POE2 to their catalogue, but with no stock Ended up ordering straight from Olimex https://www.olimex.com/Products/IoT/ESP32/ESP32-POE2/open-source-hardware Playing around with esphome, it seems to work with the same config as the esp32-poe the EXT1 header has both VPP (my 24v), and GPIO36 (my ADC) https://github.com/OLIMEX/ESP32-POE2/blob/main/DOCUMENTS/ESP32-POE2-user-manual.pdf VPP@1 GND@2,4,6,8 GPIO36@12 some doc on ESP32 ADCs https://randomnerdtutorials.com/esp32-adc-analog-read-arduino-ide/
New blog post: Random discoveries
Olimex now has the ESP32-POE2 https://www.olimex.com/Products/IoT/ESP32/ESP32-POE2/open-source-hardware, which has a nice configurable 12/24V line that would make it substantially easier to build my water tank sensor. Waiting for Mouser to carry it.
I finally settled for EH connectors for the CTCSS module for my AWA RT85. They were surprisingly hard to find. I […]
New blog post: Power and protection
More work on the 4–20mA Current Loop Rain Water Tank Sensor
Need to drop the voltage for the 78L24
46 to 24 -> 22V
Drop 11V in each stage -> 11V zener
Splitting voltage drop to dissipate power equally in each component
Current control resistor: 5mA at 46V (before established regime): ~10k
Need to protect the ADC […]
New blog post: Converting 48v to 40v for 78L24
I got pretty much all my components wrong to power my 4–20mA Current Loop Rain Water Tank Sensor: PNP instead of NPN transistor, and non-power components (I may be able to get away with this one)
Though I should be able to swap the VCC and GND in the circuit and get the PNP to work
Other suggestions to test (and be more careful about component […]
https://blog.narf.ssji.net/2024/03/05/converting-48v-to-40v-for-78l24/
Olivier Mehani