MOSCOW BAUMANN STATE TECHNICAL UNIVERSITY
DEMO VERSION!!!

BACK

AIM OF THE WORK

EQUIPMENT

   Spectrometer

   Optical behch design

   CCD-detector

   Microprocessor board

   Optical resolution

   Stray Light and Second Order Effects

   Fibre optic cable

   Light passing through fibre

   Fibre design

   Optical fibre types

   Collimating lens

   Objects of research

   AVALight lamp

   Halogen and deuterium lamps

SPECTROMETER CONTROL PARAMETERS

   Integration time

   Average

   Saturation detection levels

   Smoothing

   Correct for dynamic dark

EXPERIMENT

   Experiment scheme

   Àcquaintance with control interfaces

   Order of carrying out the work

   Stage 1

   Stage 2

   Stage 3

   Practical part
 
Saturation detection levels
The 14-bit A/D converter in the AvaSpec results in raw Scope pixel values between 0 and 16383 counts. If the value of 16383 counts is measured at one or more pixels, then these pixels are called to be saturated or overexposed.

Since saturated pixels can disturb the measurement results, a lot of attention has been given to detect saturation and to notify the user if a measurement contains saturated pixels. This notification is done in such a way that the user can always decide to ignore the saturation, for example if the saturation happens at pixels that are not in the wavelength range where the user is interested in.

Saturation can usually be solved by selecting a shorter integration time.

In AvaSpec 2048, different levels of saturation detection can be set.
  • Level "0" " - No saturation detection.
  • Level "1" - Detects pixels with 16383 counts.
  • Level "2" - Detects and corrects also saturated pixels with values < 16383. The reason for this is that if the detector is heavily saturated (at a light intensity of approximately 5 times the intensity at which saturation starts), it will return values less 16383 counts. The data for theses saturated pixels will be corrected to their saturation level.
Normally, you don't need to use level "2", but when measuring a peaky spectrum with some heavily saturated peaks, the autocorrect can be used. To illustrate this, a strong peak from the AvaLight-CAL was heavily saturated at 435.84 nm. This caused the most heavily saturated pixels to return inverted ( less 16383 counts) pixel values.

In the next figure, the saturation detection has been set to the third level, which will not only detect the saturated pixels at 16383 counts, but also detect and correct the inverted saturated pixels. Disadvantage of the autocorrect detection level is that processing the saturated scans by the application takes more time.

A spectrum can contain saturated pixels also when this is not directly obvious from the graph. Examples are:
  • Averaging.
  • Smoothing. The maximum pixel value of a peak can be saturated, but is averaged with neighbor pixels which may not be saturated.
  • The correct for dynamic dark algorithm subtracts the dark values that are measured at the optical black pixels from the spectral data. Therefore, the saturation level of 16383 counts will never be reached with correct for dynamic dark ON. The saturation detection is done before the data is corrected for dynamic dark, so it will also detect saturation with dynamic dark ON.