Metrohm 757 VA Computrace - Manual (Page 131)

5.8 Data processing and evaluation
757 VA Computrace – Software
121
Calibration curve calculation
The content determination using a calibration curve is performed
in two steps:
• First the relation between the mass concentration c of a sub-
stance and the evaluation quantity EV is determined by
measuring different reference solutions.
• Finally, the sample is measured and its mass concentration
c(s) determined using the recorded calibration curve.
In the content determination using a calibration curve, the follow-
ing procedure is used to calculate the sought mass concentration
c(s) of the sample:
1.
Measurement of calibration solutions
The calibration solutions with known mass concentration
c(n)
are each measured several times (defined by
No. of replica-
tions ). This gives:
EV(n) Evaluation quantity of a single measure-
ment for calibration solution
n
mean(n) Mean value of all evaluation quantities
for calibration solution
n
Std.dev.(n) Standard deviation of the individual value
EV(n)= s(n)
c(eff,n) Effective mass concentrations of the
calibration solutions (calculated from
c(n) taking the dilution into account)
2.
Determination of calibration curve
The calculation curve is calculated according to one of the
four possible model functions selected under
Regression tech-
nique on the Substances tab:
y = a + bx Straight line = Linear Regression
y = bx Straight line through zero point =
Linear Regression (trough Zero)
y = a + b x+ d x 4 Nonlinear curve of 4 th degree =
Nonlinear Regression
y = b x+ d x 4 Nonlinear curve of 4 th degree through
zero point = Nonlinear Regression
(trough Zero)
The parameters a, b and d of the regression curves are calcu-
lated by weighted least square minimization with y = EV and
x = c(eff). The weight factor for each point is the standard
deviation obtained from the replications. The parameters are
displayed in the
RESULTS window and have the following
meaning: