Water Filtration: DIY Gravity System & Sensors

Learning Goal: Design and construct a multi-stage gravity-fed water filtration system, analyzing purification efficiency using TDS, pH, and turbidity sensors.

Estimated Hours: 16 hours
Prerequisites: Basic knowledge of electrical circuits (voltage, current, resistance), familiarity with algebraic equations, and comfort working with hand tools (drills, utility knives).

Module 1: Introduction to Water Contaminants & Filtration Physics

This module establishes the scientific foundation of water purification. You will examine the physics and chemistry behind physical barriers, adsorption dynamics using activated carbon, and the mechanics of slow sand filters. Understanding these core processes is critical before designing physical layers in later stages.

Module 2: Designing and Building the Multi-Stage Gravity Filter

Translate theoretical physics into a physical filtration column. This module guides you through selecting materials, cleaning media, and arranging physical filtration layers sequentially to maximize gravity-fed throughput and particulate capture.

   [Raw Influent Water]
             │
             ▼
  ┌──────────────────────┐
  │ Layer 1: Gravel      │ <── Coarse sediment capture
  ├──────────────────────┤
  │ Layer 2: Fine Sand   │ <── Micro-particle straining
  ├──────────────────────┤
  │ Layer 3: Charcoal    │ <── Chemical adsorption (odor/toxins)
  ├──────────────────────┤
  │ Layer 4: Cloth/Mesh  │ <── Media retention barrier
  └──────────────────────┘
             │
             ▼
  [Filtered Effluent Water]

Module 3: Microcontrollers & Water Quality Sensors

This module transitions from physical construction to digital telemetry. You will learn the fundamentals of the Arduino platform, configure basic analog-to-digital reading paths, and interface raw physical analog probes (pH, TDS, and Turbidity) with microcontroller pins.

Module 4: Sensor Calibration & Programming

Raw analog values (0–1023) are meaningless without calibration. This module covers the mathematical principles and practical programming steps required to convert voltage values into engineering units: Standard pH, Nephelometric Turbidity Units (NTU), and Parts Per Million (PPM) of TDS.

Calibration Math & Coding (Addressing Video Gaps)

Important Note on Video Limitations: While basic analog reads are well documented in standard videos, the actual non-linear physics of turbidity and the temperature-dependent electrical conductivity of TDS are rarely fully explained in single tutorials. Below is the complete mathematical and programming framework you must implement.

1. Turbidity Non-Linear Curve Calibration (Voltage to NTU)

Unlike standard linear sensors, light scattering in water is highly non-linear. Standard gravity-based turbidity sensors (such as the TS-300B) use a parabolic regression model to convert voltage ( $V$ ) into NTU (Nephelometric Turbidity Units).

When voltage ( $V$ ) is less than $2.5\text{V}$ , the turbidity is considered too high to measure accurately. When $V \ge 4.2\text{V}$ , the liquid is essentially clear ( $0.0\text{ NTU}$ ).

The standard regression model is: $\text{Turbidity (NTU)} = -1120.4 \times V^2 + 5742.3 \times V - 4353.8$

Example Arduino Code Implementation:

const int turbidityPin = A1;

float readTurbidity() { int rawADC = analogRead(turbidityPin); // Convert 10-bit ADC to voltage float voltage = rawADC * (5.0 / 1023.0);

float ntu = 0.0; if (voltage >= 4.2) { ntu = 0.0; // Clear water threshold } else if (voltage < 2.5) { ntu = 3000.0; // Max limit of detection } else { // Parabolic regression curve conversion ntu = -1120.4 * sq(voltage) + 5742.3 * voltage - 4353.8; } return ntu; }

2. Total Dissolved Solids (TDS) Calibration Math

TDS sensors measure electrical conductivity ( $EC$ ). Dissolved salts conduct electricity, and their concentration in parts per million ( $\text{PPM}$ ) is proportional to the $EC$ value.

However, water temperature ( $T$ ) dramatically affects conductivity (increasing by roughly $2\%$ per ${^\circ}\text{C}$ ). To get accurate readings, we must normalize the conductivity to a $25\text{ }^\circ\text{C}$ reference baseline.

Temperature Compensation Formula:

$EC_{25} = \frac{EC_T}{1 + 0.02 \times (T - 25)}$

Voltage-to-TDS Conversion Math:

$\text{TDS (PPM)} = (133.42 \times V_{comp}^3 - 255.86 \times V_{comp}^2 + 857.39 \times V_{comp}) \times 0.5$ Where $V_{comp}$ is the temperature-compensated voltage value.

Example Arduino Code Implementation:

const int tdsPin = A2; const float temperature = 25.0; // Use a constant or read a DS18B20 temp sensor

float readTDS() { int rawADC = analogRead(tdsPin); float rawVoltage = rawADC * (5.0 / 1023.0);

// Apply temperature compensation factor float compCoefficient = 1.0 + 0.02 * (temperature - 25.0); float compVoltage = rawVoltage / compCoefficient;

// Convert compensated voltage to TDS in PPM float tdsValue = (133.42 * pow(compVoltage, 3) - 255.86 * sq(compVoltage) + 857.39 * compVoltage) * 0.5; return tdsValue; }

Module 5: Testing, Data Logging, and Efficiency Analysis

This final module synthesizes physical construction and electronic sensing. You will run contaminated water mixtures through your physical filter column, record real-time telemetry from all three sensors, write that data to an SD card, and calculate the overall purification efficiency of your system.

Purification Efficiency & Performance Analysis (Addressing Video Gaps)

                   [RAW INFLUENT WATER]
                Measure: TDS_in, Turb_in, pH_in
                           │
                           ▼
                ┌──────────────────────┐
                │                      │
                │   Gravity-Fed Filter │
                │                      │
                └──────────────────────┘
                           │
                           ▼
                  [FILTERED EFFLUENT]
                Measure: TDS_out, Turb_out, pH_out

To mathematically calculate the performance of your custom gravity-fed filtration system, implement the following formulas when processing your sensor logs.

1. Particulate and Solid Removal Efficiency (%)

Removal efficiency measures how effectively your physical sand, gravel, and charcoal layers trap suspended and dissolved solids. Calculate this efficiency value for both Turbidity (NTU) and TDS (PPM):

$\eta_{\text{removal}} = \left( \frac{C_{\text{in}} - C_{\text{out}}}{C_{\text{in}}} \right) \times 100$

Where:

$C_{\text{in}}$ = Raw influent concentration value (Turbidity or TDS)
$C_{\text{out}}$ = Filtered effluent concentration value (Turbidity or TDS)

Example:

Influent Turbidity ( $Turb_{\text{in}}$ ) = $450.0\text{ NTU}$
Effluent Turbidity ( $Turb_{\text{out}}$ ) = $12.5\text{ NTU}$ $\eta_{\text{turbidity}} = \left( \frac{450.0 - 12.5}{450.0} \right) \times 100 = 97.22\%$

2. pH Neutralization Index (pH Target Deviation Reduction)

Unlike physical solids, pH is measured on a logarithmic scale, and the goal of a water filter is typically to bring the water closer to a neutral pH of 7.0. We calculate how effectively our system neutralizes acidic or basic influent water using this deviation formula:

$\eta_{\text{pH}} = \left( 1 - \frac{|pH_{\text{out}} - 7.0|}{|pH_{\text{in}} - 7.0|} \right) \times 100$

Where:

$pH_{\text{in}}$ = Influent pH value
$pH_{\text{out}}$ = Effluent pH value

Example:

Influent acidic runoff ( $pH_{\text{in}}$ ) = $5.2$
Effluent after running through activated carbon and sand ( $pH_{\text{out}}$ ) = $6.8$ $\eta_{\text{pH}} = \left( 1 - \frac{|6.8 - 7.0|}{|5.2 - 7.0|} \right) \times 100 = \left( 1 - \frac{0.2}{1.8} \right) \times 100 = 88.89\% \text{ Neutralization Efficiency}$

3. Analyzing Data Logs

Once you have written your data to an SD card in CSV format, import the file into a spreadsheet tool (like Microsoft Excel, Google Sheets, or a Python script) to plot performance over time. This analysis will help you spot when physical layers are wearing down or starting to clog:

Timestamp(ms),TDS_In(PPM),TDS_Out(PPM),Turb_In(NTU),Turb_Out(NTU),pH_In,pH_Out,TDS_Eff(%),Turb_Eff(%) 10000,120.5,12.2,450.0,12.5,5.2,6.8,89.87,97.22 20000,120.5,12.4,450.0,12.8,5.2,6.8,89.70,97.15 30000,120.2,14.5,450.0,15.2,5.2,6.7,87.93,96.62

A sudden drop in your calculated efficiency scores indicates channeling (where water cuts direct pathways through media without filtering) or media exhaustion (where your activated carbon layers can no longer absorb contaminants).

Course Map

Below is the recommended sequence for your study. Ensure all prerequisites are completed before starting physical construction and programming modules.

Key People Index

Dr. Andrew Huberman (Stanford University School of Medicine)
- Context: Discusses research-backed health perspectives on tap water quality, mechanical filtration processes, and chemical adsorption methods.
Paul McWhorter (D'Arbonne Analytical Laboratories)
- Context: Created fundamental open-source educational videos for Arduino and electronics prototyping.

Final Self-Assessment

Complete this final checklist before concluding the project:

Explore Further

Related Engineering Roadmaps

View All→

engineering

Water Filtration: DIY Gravity System & Sensors

Module 1: Introduction to Water Contaminants & Filtration Physics

Recommended Videos

Module 2: Designing and Building the Multi-Stage Gravity Filter

Recommended Videos

Module 3: Microcontrollers & Water Quality Sensors

Recommended Videos

Module 4: Sensor Calibration & Programming

Recommended Videos

Calibration Math & Coding (Addressing Video Gaps)

1. Turbidity Non-Linear Curve Calibration (Voltage to NTU)

Example Arduino Code Implementation:

2. Total Dissolved Solids (TDS) Calibration Math

Temperature Compensation Formula:

Voltage-to-TDS Conversion Math:

Example Arduino Code Implementation:

Module 5: Testing, Data Logging, and Efficiency Analysis

Recommended Videos

Purification Efficiency & Performance Analysis (Addressing Video Gaps)

1. Particulate and Solid Removal Efficiency (%)

Example:

2. pH Neutralization Index (pH Target Deviation Reduction)

Example:

3. Analyzing Data Logs

Course Map

Key People Index

Final Self-Assessment

Related Engineering Roadmaps

ECG Monitor Design: Op-Amps, Filters & Noise

Soft Robotics (Molding, Arduino & Control)

Building EMG Prosthetics: Arduino & 3D Print

Building a CNC Plotter: GRBL & G-Code Setup