Projects

This page is a running log of electronics projects I’m building and refining over time. Entries include design notes, implementation details, and what I’m currently testing or planning next.

PCB Humidity Sensor for Plants

In Progress

I decided to return to electronics and work on small projects. My academic background includes electrical engineering and computer science, although most of my career has been in software engineering. For this project, I’m building on open-source soil-moisture sensor designs and documenting the process end to end. I’m interested in whether the design could eventually be commercially viable, but the main goal is to build something well and understand the trade-offs between quality and cost.

Fortunately, I still had some materials from my teaching days: an Arduino Uno, a breadboard, and several electronic components. They were enough to begin testing the basic principles.

I’ve also started buying equipment for this and future projects: a multimeter, a soldering station, and storage for electronic components. A complete electronics workbench will require more investment, but this is enough to get started.

I wanted a simple way to test the sensor, so I came up with this circuit:

Schematic of the Arduino soil-moisture sensor test circuit.

The Arduino’s analog input reads the sensor value. Digital pin D3 sends a 5 V pulse to the circuit each time we take a measurement.

R1 protects the circuit from current spikes.

The probe is the sensor: the soil between its conductors acts as the dielectric. As soil moisture rises, the dielectric constant increases, changing the probe’s capacitance.

The capacitor C2 smooths the incoming pulses and stores charge, while R2 slowly discharges C2. Together they form a low-pass filter that converts rapid changes into a stable voltage that can be measured by A0.

The Schottky diode protects the circuit while capacitor C2 discharges, preventing current from flowing in the wrong direction.

I built the circuit on a breadboard and confirmed that it works.

This was the easy part, but it confirmed that the same principle could work on a custom PCB.

It was time to move to KiCad and turn the design into something more sophisticated.

I decided to start small. Since this is my first PCB, I chose to recreate only the sensor so I could test it and practise soldering. This is the schematic.

Schematic for the first custom PCB for the soil-moisture sensor.
Three-dimensional preview of the soil-moisture sensor PCB.

Current focus

I’m iterating on the first revision of the PCB, concentrating on stable analog readings in a humid, corrosive environment. I’m also validating component choices to balance accuracy, lifetime, and cost.

  • Refining electrode geometry for consistent moisture response
  • Testing different protection strategies against soil corrosion
  • Designing the analog front-end for low-noise, repeatable readings
  • Planning the interface to a microcontroller for data logging

Design notes

This is still an early-stage board, so the emphasis is on learning from each prototype rather than optimizing for production. I’m logging:

  • How trace routing and ground layout impact measurement stability
  • Drift over time when the sensor stays in soil for days or weeks
  • Calibration approaches that stay valid across different soil types
  • Assembly notes for hand-soldering vs. small-batch assembly

Once the readings are consistent enough, the next step will be integrating temperature compensation and defining a simple protocol for external controllers.

Next updates planned

  • Documenting the first PCB revision, including schematic and layout rationale
  • Capturing measurement data across different soil types and watering cycles
  • Adding photos and notes from early breadboard and probe tests

Other Projects

  • Why I replaced a working amplifier chip

    Why I replaced a working amplifier chip

    A working LM386 amplifier taught me that a technically satisfying redesign is not always the most efficient fix. For a car audio project, speaker sensitivity and enclosure design mattered as…

  • Building a Low-Power Audio Amplifier

    Building a Low-Power Audio Amplifier

    A build log for a 5 V LM386N audio amplifier, explaining its coupling, filtering and gain components from schematic through breadboard testing.