Malvern Mastersizer Series Getting Started page 124

C H A P T E R 1 0
P a g e 1 0 . 2
G e t t i n g S t a r t e d
It is important to realise that if these parameters are set unnecessarily the accuracy
and resolution of the analysis will be reduced.
lower end of the range to exclude data routinely unless you have good reason.
For example in some spray measurements the presence of a propellant material
imposes on the signal a scattering pattern of its own. This signal is only present at
the same time as the sample and so cannot be compensated for by the background
subtraction. In such cases the effects are frequently confined to a small angular
range and it is desirable to exclude that range from the analysis. Unexpectedly
large signals in these channels are more likely to be due to excessive bubbles in
the dispersion, bad alignment or dirty optics.
Large signals in the outer channels may well result from allowing high levels of
daylight or artificial light to fall on the receiver optics during spray measurements.
If spray measurements have to be conducted at a distance from the range lens
such that the one or more outer rings of the detector are cut off then you should
kill these rings to stop possible bias in the results. See "Avoiding lens cut off" in
chapter 4 for more details.
Up to ten data channels may be removed from the inner (lower numbered) and
outer (higher numbered) data channels. The channels are counted from each end
of the range so it is not possible to kill a lower numbered channel without also
killing the channels below it, or a higher numbered channel without killing the
channels above it. Killing a channel is not the same as putting the value to zero -
the killed channel is actually excluded from the analysis.
Below is an example of how killing data channels affects the result. The left hand
display shows the measured data and on the right is the corresponding result. The
data from the 50 microns particle distribution has been corrupted by scattering
from bubbles - showing up as high readings in the first three channels. This
produces the peak at 220 microns in the result. The lower displays show what
happens if the first three channels are killed À. The result display now shows the
expected distribution.
Do not set kill channels at the
M A N 0 1 0 1