EXFO IQS-1500 Power Meters for IQS-500/600 - Manual (Page 75)

Calibrating Other Power Meters with the IQS-1500
Power Meters 67
Entering Calibration Uncertainties
Type A uncertainties are characterized by taking a series of measurements
and calculating the standard deviation. The standard deviation of the
calibration factor for the DUT is automatically calculated and displayed in
the Standard Deviation box of the calibration summary (see Viewing
Calibration Summary on page 76).
Type B uncertainties must be half width, U
Bi
, of a rectangular distribution
from −U
Bi
to +U
Bi
(in percentage).
The Source power stability is very important since the calibration is
sequential; that is, measurements are first taken on the IQS-1500
Calibration Power Meter, and then on the power meter under test (DUT).
This first error source can be evaluated by measuring the variation of the
source power on the Calibration Power Meter over the time of calibration
at one wavelength (about 30 seconds).
The Reference Meter spectral sensitivity slope, in %/nm, around the
wavelength of the source influences calibration because of the uncertainty
about the exact knowledge of the source wavelength. If not already
known, the slope can be calculated using the following procedure:
To calculate the reference meter spectral sensitivity slope:
1. With a stable light illuminating the detector, set the IQS-1500 to the
wavelength of the source, for instance, 1550.0 nm, and record the
displayed power, for instance 1.000 mW.
2. Increase the wavelength of the IQS-1500 by the uncertainty of the
wavelength of the source, for instance, if ± 0.1 nm, set 1550.1 nm.
3. Record the second displayed power, for instance, 1.001 mW.
4. Calculate the slope. In this case, the variation is
([1.001−1.000]/1.000)/0.1 nm=1 %/nm; the uncertainty is then 0.1 %.
The DUT spectral sensitivity slope has a similar effect. It can be evaluated
the same way if the spectral resolution of the DUT is also 0.1 nm. If not, an
estimate can be obtained by setting the second wavelength as close as
possible, for instance, 1540 nm. The evaluation will then be less precise,
and judgment is important to have a good estimate.