beautypg.com

2 wavelength linearization, transmission probes, Wavelength linearization, transmission probes – Metrohm NIRS XDS Process Analyzer – DirectLight/NonContact User Manual

Page 88

background image

86

▪▪▪▪▪▪▪

11.4.2

Wavelength Linearization, Transmission Probes

The first test to be run is Wavelength Linearization. This provides initial wavelength locations for the
sample spectrum, prior to final wavelength calibration.

Wavelength Linearization uses internal wavelength materials to determine a set of internal, arbitrary
peak positions that the instrument will use to maintain repeatability of wavelength response.
Wavelength Linearization is performed on the internal reference fiber located in the XDS Process
enclosure.

The wavelength positions of these peaks
appear as shown.

The scale of this display is marked in
encoder pulses, which do not relate to
nanometers directly.

From the peaks, a linearization is
performed, which allows assignment of
nanometer values. (Actual spectral shape
depends on fiber configuration.).

The spectrum shown above does not exhibit any noise effect from fiber attenuation. Actual fiber
optic probe spectra will exhibit noise at the lower and upper ends of the spectrum, based upon
length of fiber used. This is normal.

A Wavelength Linearization is performed in the forward direction of grating scan, then again in the
return direction. The resulting information is adjusted by Vision into one corrected spectrum.

These peak positions are not meant to be
traceable, as the wavelength calibration of the
instrument is done on an external standard,
traceable to NIST.

The internal wavelength materials are used to
maintain the external wavelength registration by
use of software adjustment for any external effects
on the instrument.

Select Wavelength Linearization from the
Diagnostics menu. The instrument will scan the
reference, which is the fiber optic that runs from
the monochromator to the detector area inside the
instrument.

Peak positions for the reference materials are located using a peak-finding algorithm. These “found”
peaks are compared to the nominals. This is done twice, one for each direction of the grating
motion. No operator interaction is required.