Calibration Procedure for New Products
This procedure is outlined for snack foods that may or may not be ground in the lab for analysis or powder and bulk solid or similar applications. Double-Blind Samples are recommended for calibration and validation. This allows the sensitivity or span (linear regression slope) to be accurately determined in the laboratory to speed up the calibration of on-line transmitter when it’s difficult to create a range of process moisture samples on-line.
Step 1: Collect sufficient product for about four or five samples of 200 grams each.
Step 2: Measure 100 grams +/- 2 grams and place into a sealed Bag or container.
This will be sample 1.
Step 3: Measure 100 grams and spray about 2 grams of water on the sample. Mix
the sample and place into a sealed bag or container. This will be sample 2.
Step 4: Measure 100 grams and spray about 4 grams of water on the sample. Mix
the sample and place into a sealed bag or container. This will be sample 3.
Step 5: Measure 100 grams and place in an oven with a temperature of 100 degrees C.
Leave this sample in the oven for 10 minutes*. Remove from the oven and
Place into a sealed Bag or Container. This will be Sample 4.
Note: If a fifth sample is used repeat Step 5, but keep in oven for 20 minutes*.
Step 6: Leave these samples in the Sealed Bag or Container for at least four hours to equlibrate.
Step 7: Take Sample 1 and place under the MCT. Record the moisture measurement.
Step 8: Do the Normal Lab analysis for this sample. A Confection Oven or
Vacuum Oven is recommended.
Step 9: Repeat Steps 7 and 8 for the remaining samples. When all the samples are
Analyzed, we can begin the MCT Calibration Routine. This can either be performed
via the MCT Operator Interface panel or by using the PSC Viewer Suite software. Entering the Lab vs. MCT Near Infrared Transmitter Data allows for a linear regression to be performed and the new Span/Sensitivity (Linear Regression Slope) and Zero/Offset (Linear Regression Y- Intercept) to automatically be generated for correct calibration.
Step 10: Confirm calibration by collecting sample on-line within target set point values and only make a zero or offset adjustment based on double-blind lab sample. Congratulations!
* Theseare variable.
Some products may need more or less water. Some products may take more or less time. Please contact PSC to discuss any questions or to request our assistance.