Appendix A: Accuracy calculations
Model 6517B Electrometer Reference Manual
A-4
6517B-901-01 Rev. C / August 2015
Calculating resistance/resistivity accuracy and repeatability using the
alternating polarity method
From the specifications, accuracy and repeatability when using a Model 8009 test fixture are
calculated as follows:
Repeatability:
I
BG
R/ V
ALT
+ 0.1 % (1
)
(instrument temp. constant ±1 °C)
Accuracy: (V
SRC
Err + I
MEAS
Err
R)/V
ALT
Where:
I
BG
is a measured, typical background current noise from the sample and fixture over the
measurement time programmed, under normal measurement conditions (1 PLC, same range,
filters off, etc.)
V
ALT
is the alternating polarity voltage used
V
SRC
Err is the accuracy (in volts) of the voltage source using V
ALT
as the setting
I
MEAS
Err is the accuracy (in amps) of the ammeter using V
ALT
/R as the reading
The following example shows how to calculate accuracy and repeatability for a 10
13
ohms sample
measured in a Model 8009 test fixture, with 50 V stimulus and a background current of 4 pA. The
user-selected measurement time for the alternating polarity measurement is 15 seconds.
I
BG
is measured with no voltage applied, under normal measurement conditions, and 15 seconds
(according to the user’s choice for the measurement) of readings are stored to the buffer. Normal
conditions require 1 PLC, the same amps range that is used for the alternating polarity measurement
and similar preconditioning to the normal measurements (if the samples are normally inserted into the
fixture and immediately measured, the stored readings should be performed similarly after lid
closure). After the storage is complete, the standard deviation (1 sigma) of these values is taken. This
can be done by retrieving the standard deviation of the buffer over the bus or by pressing
RECALL
once and
NEXT
repeatedly, scrolling through the second-line displays to the standard deviation. This
is
I
BG
. Assume for example, that the result was 55 fA or 5.5
10
-14
A.
Repeatability = 5.5
10
-14
A
10
13
/ 50 V + 0.1 % = 0.011 + 0.1 % or 1.2 %
Note that with normal ohms methods, the 4 pA total background current contributes as much as 400
% error.
V
SRC
Err = 0.15 %
50 V + 10 mV = 0.085 V (from 100 V V-source range specs)
I
MEAS
Err = 1 %
(50 V / 10
13
) + 30 counts
10
-16
A / count = 5.3
10
-14
A (from 20 pA spec)
Accuracy = (0.085 V + 5.3
10
-14
A
10
13
) / 50 V = 0.0123 or 1.23 %
