Aquatic Spectrometer & Turbidity Meter Preliminary Design Review ECE 4007 L1, Group 8 Paul Johnson Daniel Lundy John Reese Asad Hashim
Introduction & Background What is it? A device to detect the colour and clarity of a uniform flowing water sample How will it work? LED’s, a diffraction grating, a photodetector array and an on-board PC Why do we need it? Demand from Aqua-culturists and Water Regulation Authorities for a cheap and easy to use device
High Level Block Diagram
Electronic Specifications LED’s (Luxeon LXHL-NWG8) Switched on/off via control signal, through 350mA Continuous Power Supply CMOS Sensor (Kodak KAC-9630) Triggered to capture via control signal through serial interface Power Considerations
Electronic Block Diagram
Optics - Prism Higher cost Larger area required Source:
Optics – Diffraction Grating Inexpensive Easily positioned 500 grooves/mm Resolution of 0.633nm Source:
Optics – Physical Placement Requires fine sensor adjustment +/-0.1mm tolerance for 1 st order maxima Active sensor area is the limiting factor
Mechanical Hardware Secure the optical and electronic components to the enclosure Facilitate and ease the alignment process Keep the system calibrated, mechanically, as long as possible
Two Slits Assembly
CMOS Sensor Assembly
Base Support Plate Slotted holes provide movement and alignment adjustment for the distance between the diffraction grating and CMOS sensor Mates to the Vertical Support Plate
Vertical Support Plate Three Point Precision Mount (Springs and Screws) Middle screw hole utilizes negative pressure via lock down screw to secure position Provides minute adjustments in the horizontal direction Mates to both the Base Support Plate and CMOS Mount Plate
CMOS Mount Plate Attaches the CMOS Sensor to the plate via standoffs Provides the vertical alignment adjustment
Single Board Computer TS-7250 ARM9 Single Board Computer 200 MHz 32 MB RAM Programming in C Four source files: SpecMain.c SquareWave.c Process.c Networking.c Compiling with ARM9 compiler obtained from vendor Networking software with wireless networking capabilities Source:
Software Flow Chart
Photo Sensor Interfacing Sensor will be clocked at 10 MHz A 1 byte intensity value corresponds to each pixel on the sensor A serial image consists of a data out pin d[0] and three synchronization pins: d[1],vsync, and hsync Source: Kodak KAC-9630 data sheet
Normalizing the Spectrum The white light spectral response of the sensor is not perfectly flat Other factors such as LED spectral output also add distortion to white light response Source: Kodak KAC-9630 data sheet These inconsistencies are accounted for by performing spectral analysis with no sample present and multiplying the measured response of samples by the inverse of the white light response
Color Analysis - Obtaining Spectrum Values Intensity values are stored in a vector Vector is divided into 3 (or more) regions Total intensity of each region is calculated The resulting regional intensities are compared to each other and stored as ratios Ratios are compared to predetermined ratios from known algae samples to determine the algae's growth stage
Spectrum Division Visual representation of spectral division
Turbidity Analysis Regional intensities from color analysis are summed to create an overall intensity The weaker the overall spectral intensity, the greater the turbidity Intensity to turbidly conversion will be calibrated by finding the spectral intensities of various samples of water with known turbidities
Cost Analysis PartCost LEDs$9 SBC$184 CMOS Image Sensor$14 Optics Kit$10 Power Supply$20 Misc. Hardware$120 Total$357
Conclusions Electronics Schematics drawn, parts en route, prototyping in progress Optics Parts en route, calculation & experimentation stage Mechanical Mechanical drawings done, in fabrication stage. Software SBC delivered, preprogramming stage
Questions?