Water quality monitoring has become a critical environmental and industrial concern as pollution, industrialization, and regulatory requirements increase. It is clear that traditional manual sampling methods currently do not meet the needs of real-time decision-making, necessitating the development of portable, low-cost tools. On the other hand, spectrophotometers are precision scientific instruments that are complex in their operation. Most laboratory spectrophotometers are very expensive.
When discussing a low-cost, portable water quality monitoring tool, it is a platform used for continuous and cost-effective monitoring of critical parameters, such as pH, Turbidity, Free chlorine, and Dissolved oxygen. Low-cost, portable tools for water quality provide accurate measurements; however, their widespread application in site use cases faces several limitations, including high power consumption, complex integration, high manufacturing costs, hard calibration processes, and the need for monthly or yearly maintenance.
This project aims to design and manufacture an affordable, low-cost connected spectrophotometer, suitable for general public use in water analysis, with adequate sensitivity.
Proposed solution general explanationWater samples' chemical and physical parameters can be measured quantitatively using spectrophotometric techniques. Spectrophotometry is a technique used to measure the interaction of light with water and its dissolved chemical components. When light falls on a water sample, every chemical compound dissolved in water can absorb, transmit, or reflect light at a specific wavelength.
This project deals with the design of a simple spectrophotometer to assess water quality, potability, and track pollutants. The preliminary version of this device shows a measurement accuracy that allows us to reliably assert values for pH, Chlorine CL, Bromine Br, and Turbidity.
The block diagram below describes the main device architecture. LEDs with a specific wavelength are used as a light source to illuminate the water sample. The LED wavelength is selected to tackle specific components dissolved in water. The focal lens is used to get uniform illumination and increase light intensity. A beam splitter is used to allow our solution to get both sensor responses before and after the light hits the water sample. We are proposing three sensors attached to the water solution sample. We use an optical filter with two of them to tackle specific chemical parameters or pollutants.
The proposed design is split into four submodules:
- The sensor module is based on the TEMD5010x01 photodiode.
- The LED module is based on LXZ1 LEDs with a specific wavelength.
- The photonic board is based on ADPD4200 from Analog Devices. This module interfaces three sensor modules with one LED module.
- The motherboard used to read measures from the photonic board, execute: Analysis, EDGEAI models, and communications tasks.
The design is done with KiCad v9. The project files are attached; many thanks to NextPCB for the support.
The sensor module boardThe sensor board is built using a reverse-biased photodiode. And the photons hitting the photodiode modulate the leakage current, the schematic and the board in the figure below.
Spectroscopy is based on light; each wavelength is used to target a specific compound. In a laboratory spectrophotometer, monochromators are used to get the desired wavelength. In this design, we simplify the architecture by using LEDs with a specific wavelength. Below, the board uses LXZ-PB01 to provide lighting with a wavelength of 530nm.
The water analysis board is just 85mm x 55mm, is used to accommodate up to three sensor module boards; one photodiode is integrated into the board for correction purposes.
This board is designed using the ADPD4200 from Analog Devices, mainly used for healthcare and wearable applications.
The communication with the water analysis board is based on the SPI interface, to read inverse currents from the photodiodes and control the LED.
This board executes the software to read measurements from the photonic sensor boards through an SPI interface, inference IA models, and send data using Bluetooth low energy BLE. The main processor is a NORDIC nRF5340. The schematic and the design details are included in the attached KiCad project files.
The tube holder is 3D printed it implements a focal lens, and a light splitter.
Comments