Customer Solutions
Creating an Integrated Spectral Water Analysis System
Author(s):
Eric Lyness, Mink Hollow Systems, Inc.
Industry:
Life Science
Product:
Data Acquisition, LabVIEW
The Challenge:
Determining the multiple biological and chemical contents of ocean water by analyzing the spectral water signatures at different wavelengths and the light scattering at different angles.
The Solution:
Controlling a single multifunction data acquisition board with National Instruments LabVIEW, with which we modulated three lasers at different wavelengths and measured the spectral response of the water using optical detectors; an ocean optics spectrometer; and a multichannel custom Kaitech spectrometer; while we combined and analyzed all of the data with a separate NI LabVIEW application to determine the interrelationships of the water contents and optical signatures.
Postprocessing User Interface
|
Identifying Several Constituent Elements
Existing spectral water-content-measuring instruments typically look for a single ingredient with a relatively low-resolution sensor. Scientists looking for multiple contaminates need to carry multiple instruments to the water. Kaitech designed an instrument that could identify several constituent elements by using three laser light sources, two spectrometers, and photo detectors in a single box. Each laser stimulated the water differently, causing the water to fluoresce differently. The resulting fluorescence signature could characterize the elements in the water. The photo detectors measured the optical power that scattered in front of, beside, and behind the laser light path. The ratio of the power scattered in each direction provided a signature indicating the size of particles in the water.
The Kaitech Integrated Spectral Water Analyzer is an operational prototype that generates signature data and provides a comprehensive user interface to analyze the data. Currently, the laboratory instrument is undergoing beta test at Dauphin Island Sea Lab in Alabama. Using the analysis, scientists can determine the relationships between water contents using their signatures. By essentially combining several instruments in a single box, the resulting data set is greater than the sum of the parts. Scientists can find more complex signatures through the relationship of scattering and spectral response at different frequencies.
Measurement Design: Accuracy and Precision
In an experiment, water flowed through a small spherical measurement vessel, and light from each laser entered the vessel through fiber optics. As the laser light passed through the water, lenses, mirrors, and fibers within the measurement vessel and directed the light to the detectors, ocean optics spectrometer, and Kaitech customer spectrometer.
The system used five functions of the NI PCI-6025E multifunction data acquisition (DAQ) board -- digital input, digital output, analog input, analog output, and timer output. As part of the Kaitech custom spectrometer, the system used a driver for a P-CAM camera attached to the NI PCI-BUS and another driver for an ocean optics spectrometer attached to the USB port.
The unique water ingredient signature can be very small relative to the overall signal power. Thus, it was imperative that we made accurate and precise measurements. The key to making good spectral signature and scattering signature measurements is calibration and synchronization. Calibration provides accuracy by eliminating extraneous ambient light. Synchronization provides precision by increasing the signal to noise ratio. Using LabVIEW and PCI-6025E multifunction, we performed the calibration and synchronization required to make the measurements.
We achieved calibration by measuring the ambient light when no lasers were enabled. The LabVIEW application performed a calibration for each spectrometer and each detector at the beginning of a test and stored the calibration data with the output data so we could trace measurements back to the raw values. For optical sensors, the signal-to-noise ratio decreased when the sensor integrated a dark image while the light stimulus was not enabled. Synchronization ensured that the optical sensor was only exposed while the light source was active. The digital I/O and timer outputs, with the flexibility of the PCI-6025E, made synchronization possible.
We read the photo detectors using three analog inputs. We were most interested in the photo detector signature of the optical power in each direction (forward scatter vs. side scatter; forward scatter vs. backscatter; side scatter vs. backscatter). This step required very precise measurements as the ratios could be quite small when overall power signal was not small. Also, it was important that all of the detectors read at the same instant so that we knew that the relative values of each reading passed through the exact same sample of water.
For the most precise measurements, we tied the PCI-6025E analog input trigger lines (which were dynamic at run time) to the output modulation lines so that the system read the analog inputs only while the laser was illuminated. We enabled laser A and configured the analog input to trigger on input PF1, which we wired to the timer output modulating laser A. When we enabled laser B, we triggered on PF2, and so on. This synchronization improved our signal-to-noise ratio for good scattering measurements.
The laser modulation frequency did not affect the slower responding spectrometer CCDs. But it was important that the spectrometer exposure not begin until the laser reached full power and that the exposure ended before we turned off the laser. We hardware-triggered each spectrometer using a digital output on the multifunction DAQ board after the laser modulated for a warm-up period. Digital inputs indicated when the exposure was complete and the laser could be extinguished.
Because bubbles in the water or flow problems could cause bad measurements, the test sequence typically repeated 50 measurements with each laser. We implemented the sequencing in LabVIEW using a nonblocking state machine with which we could abort the test at any time.
Our fundamental measurement sequence included:
1. Initialize I/O and instrumentation.
2. Calibrate detectors and spectrometers.
3. Configure analog I/O and spectrometers for
synchronization.
4. Enable and modulate the laser.
5. Wait the delay time.
6. Trigger the spectrometers.
7. Wait for the spectrometer exposures to
complete.
8. Disable the laser.
9. Read buffers.
10. Store data.
11. Repeat steps 3 through 12 for each laser.
12. Repeat steps 3 through 13 until 50
measurements have been made.
A single test sequence run resulted in 150 data files (50 files per laser), each about 35 kB, and a total of over 5 M of data. This data was unwieldy and not very useful. Using the excellent LabVIEW data presentation abilities, we created a separate postprocessing program that condensed the data and helped us browse and combine data to search for clear signatures.
While 3D surfaces mean quick trend identifications, an adjustable boxcar-averaging technique helps us effectively increase the signal-to-noise ratio by averaging the results over adjacent runs.
When we identified a clear water constituent signature, we exported an image of the graph along with a spreadsheet file containing the graph data.
NI Products Provide Flexibility and Cost Savings
Using LabVIEW and DAQ hardware, we rapidly developed software to make precise, calibrated, synchronized water properties measurements using spectral techniques and presented and analyzed the resulting data. We saved thousands of dollars by using a single instrument instead of several. And, the high spectrometer resolution, in conjunction with the data analyzing software flexibility, provided far more scientific information than was possible with several instruments.
For more information, contact:
Eric Lyness
Mink Hollow Systems
6880 Mink Hollow Rd.
Highland, MD 20777
Tel: (301) 854-1579
Fax (310) 854-9746
