Mantracourt DSC User Manual page 44

CLX5 = 449.98
(Note that these are the raw reading values, not the known true values.)
Now calculate all the residual errors, and set up the correction factors –
3
×(x1 – c1) ≈ 1000×(000.00–0.0010) = –1.0
CLK1 = 10
3
CLK2 = 10 ×(x2 – c2) ≈ 1000×(100.13–100.44) = –310.0
3
×(x3 – c3) ≈ 1000×(199.72–200.57) = –850.0
CLK3 = 10
3
CLK4 = 10 ×(x4 – c4) ≈ 1000×(349.97–349.75) = +220.0
3
×(x5 – c5) ≈ 1000×(450.03–449.98) = +320.0
CLK5 = 10
The CELL output values will now have the required values at all these 5 points.
Note on the Example
If you graph the errors from the above example, the results look like this
0.4
0.2
0
0 100
-0.2
-0.4
-0.6
-0.8
-1
This doesn't show any very definite linear trend, so the calibration is okay.
However there is a big jump between points 3 and 4, which might be worth a more detailed investigation: Some
important features of the error curve could have been missed by the test.
43 Micron Meters
200 300
DCell & DSC User Manual
400 500