Page 220 - Photodetection and Measurement - Maximizing Performance in Optical Systems

Source: Photodetection and Measurement

Chapter
10

Measurand Modulation

10.1 Introduction
In earlier chapters we investigated techniques for optimizing the detection
sensitivity of weak optical signals and for achieving high repeatability of
measurements of relatively strong signals. The primary technical difficulties
addressed were, respectively, noise and stability, which we treated as operating
independently. Here we will look at an important measurement situation where
the two errors are strongly coupled. This concerns the performance of an instru-
ment in detecting very small variations in its measurand, that is, the instru-
mental characteristic of limit-of-detection or LOD.
Consider first the design of a high-sensitivity on-line scattered light mea-
surement system, such as a turbidimeter used for analysis of drinking water
(Fig. 10.1a). A high-power modulated light source is projected through the
sample to a beam dump, while a high sensitivity receiver collects the scattered
light, synchronously detects the source modulation, and displays the turbidity
reading V m. During commissioning the water is independently verified to be
particle-free, so the relative light scattering efficiency d and photocurrent I p
should be almost zero. As the signal is so weak, repeated readings of the instru-
ment’s display V m are made and shown to be distributed around zero, as
expected. This is represented by the fuzziness of the measurement result at the
calibration shown as Day 1.
The next day V m is found to have increased by a small amount DV to just
above the first measurement value, causing us to ask “has the turbidity
actually increased?” Again, repeating the measurement a few times gives the
expected variation in results, which suggests that it wasn’t a one-off error. The
smallest change in signal DV that can be detected, and hence the LOD of the
turbidity or other physical measurand, is clearly a function of the noise level of
the detected signal. To be confident that the intensity has changed, for example
between the Day 1 and Day 2 measurements, we require DV to be several times
greater than the standard deviation of the noise, the measurement uncertainty.

213
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2004 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.