RAPT Pill Hydrometer & Thermometer Quick Start Guide
Page
12
of
18
Last Updated 6/05/2022 16:49 PM
3: Custom Calibration
Warning! Custom Calibration is for advanced users only
.
This mode can provide the most accurate readings across a wider range than the other calibration
modes.
Only proceed if you are comfortable with spreadsheets and complex curves. If you are working with
solutions with gravities higher than 1.080, then custom calibration is advised.
For advanced calibration, you will need to take readings at various gravity points roughly 10 gravity
points apart, and make a note of the angle of the Pill at each point. We recommend readings from
1.080 down to 1.000, 10 points apart (so, 1.080, 1.070, 1.060 etc). Note that it does not matter if the
gravity readings are not exactly 10 points apart, as long as the gravity reading is correct. Use of a
hydrometer that is known to be accurate is quite important.
To perform Custom Calibration, it is easiest to start with the strongest solution and dilute to
progressively weaken the solution. It is important to use a calibrated hydrometer to ensure accurate
gravity readings.
1.
Place the Pill in the first solution of known gravity. Wait for the angle to stabilise, then make
a note of the angle and the gravity.
2.
Repeat this step for solutions of different gravity, making a note of the angle and the gravity
at each step. It is easiest to dilute the starting solution in steps of approximately 10 points.
Finish with a reading in a solution with a gravity of 1.000 (IE water).
Once you have sufficient readings, the angle and the gravity need to be entered on an Excel
spreadsheet (or similar such as Google Sheets) as a graph wit
h gravity on the ‘y’ axis and angle on
the ‘x’ axis. The more readings you take, the more accurate this calibration will be. We recommend a
minimum of 6 samples, with 10 or more providing a greater degree of accuracy.
When you have created the graph, create a polynomial trend line onto the graph with the degree of
the polynomial being
at least
one less than the number of data points. If the trend line creates an
irregular curve (such as not passing through the data point at 1.000 correctly or not following a
polynomial shape) reduce the number of data points (ensuring that there are still more data points
than the degree of the polynomial) until the curve matches.
With 10 data points, a polynomial degree between 5 and 8 will typically give the best results.
