F-751 Avocado Operation Manual Rev. 1/25/2023
Page
|
7
Felix Instruments
1554 NE 3rd Ave
Camas, WA 98607, USA
Phone: +1 (360) 833-8835
Fax: +1 (360) 833-1914
www.felixinstruments.com
[email protected]
Variance Due to Analytical Procedure
Every measurement method has inherent sources of error. For most
gravimetric methods (e.g., oven, microwave, etc.) error can be attributed to
unrepresentative sampling and measurement error associated with
equipment (e.g., the balance not being properly calibrated, temperature not
being monitored, etc.). The F-751 most similarly mimics the coring sampling
technique, and thus will provide the most comparable results when the user
destructively analyzes a core sample taken from the scan site.
If a sampling technique other than coring is being utilized, then extra
precautions (e.g., increasing the number of scans and scan sites on the
avocado) must be taken to ensure the average DM value given by the
instrument aligns with current destructive testing methods.
Measurement Variance Due to Instrument
The new avocado application has been created with artificial neural
network (ANN) chemometrics, which has drastically reduced the amount of
inter-instrument variation. This ensures a high level of repeatability and
accuracy when taking measurements with the F-751. However, if the user
wishes to further reduce their measurement variation, it is recommended
that they increase their sample size. Increasing the sample size decreases
error significantly. Using the average DM of ten fruit will typically deliver
30% better performance than using the average DM of five fruit. Because it
is quick to measure with the F-751, we recommend a larger sample size that
delivers the desired performance.
