Counting efficiency is the probability that an Optical Particle Counter (OPC) will sense and accurately count a particle passing through the particle counter's sample volume. This probability describes the percentage of particles counted at—or above—a specified particle size. In the adjacent figure, three plots illustrate counting efficiency versus particle size. Curve A (the vertical line) shows a counting-efficiency curve for a hypothetical optical particle counter with perfect sizing resolution. Curve B shows the curve for a real-world OPC with good resolution. Curve C shows the curve for an OPC with poor resolution1. While the signal produced by the particles is symmetrically distributed around the nominal most sensitive threshold, the exponential relationship between particle size and signal causes the counting efficiency curve to be asymmetrical.
An OPC’s resolution is the ability to resolve small differences in particle size. A number of factors combine to make an OPC’s resolution less than perfect. These factors include the illumination uniformity within the sampling volume, the quality of the optical system, the quality of the electronics in the Pulse Height Analyzer module, and the noise due to photon statistics2.
The term instrument resolution describes the number of particle size channels, integrated by the manufacturer, to span the instrument’s entire size range. The minimum possible channel size width is limited by the fundamental resolution. The instrument resolution selected for Particle Measuring Systems’ particle counters is much broader than the fundamental resolution. This is done to simplify instrument operation, minimize costs, and applications seldom require high particle resolution.
Read the full paper to learn more about the following:
- What is particle counting efficiency and why should it be 50% at the most sensitive sizing threshold?
- What is the meaning of optical particle counter resolution and why is it important?
- How will particle counting efficiency affect data for actual particle distributions in cleanrooms as compared to data from monodispersed calibration particles?
- Percent resolutions selected to illustrate good and poor are 10% and 25% respectively.
- A low signal-to-noise ratio (S/N) can also produce resolution degradation by allowing artifacts (noise pulses) to be counted as particles. For purposes of this discussion, we will assume that the S/N ratio is adequately sufficient and allow us to ignore this factor.