Installation Operational Qualification
Serstech AB, 2021
Test 1: Dark background-hot pixel test
Equipment
Measures must be taken to prevent any ambient light from reaching the spectrometer. The
collection optics must be completely enclosed in a light-tight housing using a tightly closed
sampling accessory.
Procedure
The spectrometer collection lens is carefully enclosed to exclude any ambient light. Spectra
should be acquired at room temperature (20-25 °C). Two separate background spectra are
acquired with an integration time of 2 sec, with a noncalibrated and non-interpolated “raw
pixel” x-axis. All spectral post-processing methods (e.g. baseline correction, background
correction), as well as the laser itself, are required to be turned off. Additionally, an “offset”
spectrum is acquired with a 20 ms integration time. The spectra are saved and transferred to
suitable data analysis software (e.g. Origin). The average of the 20 ms “offset” spectrum is
calculated (skip the first and last 16 pixels) and subtracted from the first raw spectrum to
yield the “spectrum with fixed pattern noise”. In this spectrum, any pixels exceeding 2000
counts above the background mean are dropped, i.e. removed from the data set. The
arithmetic mean of the remaining points is calculated. Additionally, the two raw spectra are
subtracted from another to give the “spectrum without fixed pattern noise”. The standard
deviation of the latter spectrum is calculated. The baseline should not be skewed in any of
the cases.
Result
The arithmetic mean of the cleaned “spectrum with fixed pattern noise” is called
background mean, and the number of faulty pixels equals the number of data points
dropped during the procedure. The calculated standard deviation of the “spectrum without
fixed pattern noise” is the background noise.
Please fill in the Table 10 the information you obtained from the dark background-hot pixels
test.
