Calibration
We need to consider ways to improve the accuracy we can
expect with pulsed measurements. The primary considera-
tion when calibrating is to have the analyzer completely
set up before the calibration is begun. The pulse modula-
tor, triggering circuit, and all other hardware should be in
place, and all instrument settings should be complete
before starting a calibration. The only thing absent from
system should be the DUT. This ensures that additional
errors will not be introduced into the system after the cali-
bration is complete.
The type of calibration chosen depends upon the type of
measurement being made and the accuracy required. We
will consider three types of calibration: response calibra-
tion, enhanced response calibration, and full two-port cali-
bration.
Response calibration (normalization) is most often used
for transmission measurements, though it is also possible
for reflection measurements (using a short as a calibration
standard). It removes only the frequency response error;
however this is often sufficient since this is usually one of
the largest errors in the measurement system. The advan-
tages of this method are that it is easy and quick and
requires only a single calibration standard. The disadvan-
tage is that the measurement uncertainty is higher for this
method than for other methods
1
. This is because the accu-
racy of any calibration depends in some measure upon the
port match of the analyzer and the DUT. Placing well-
matched 6 dB pads on the ports before calibrating will
sometimes solve this problem, but at the expense of
dynamic range.
Enhanced response calibration is a new feature offered on
the 8720ES. This technique corrects for frequency
response error, and in addition corrects for source match.
This is an advantage because it removes any ripple in the
response due to the raw port match of the analyzer. This
in turn significantly improves the measurement uncertain-
ty. The disadvantages of this method are that it requires
four calibration standards (short, open, load, thru) and
that it requires more time and attention to detail than a
simple response cal. Once calibrated, the analyzer sweeps
just as quickly as it would with a response cal.
Full two-port error correction provides the highest accura-
cy possible using short, open, load, through (SOLT ) cali-
bration standards. As the name implies, it corrects for
errors at both ports, including frequency response, source
match, load match, transmission tracking, reflection track-
ing, and if desired, cross talk. The disadvantages of this
method are that it requires up to seven calibration stan-
dards (when both sexes are considered), it requires twice
as many connections as enhanced response cal (with the
attendant opportunity for mistakes) and once completed,
it runs at half the speed of an analyzer using enhanced
response cal (because both a forward and a reverse sweep
are required before the parameters can be updated).
2
Table 2 shows the various combinations of measurement
type and calibration that are compatible with pulsed meas-
urements.
10
Table 2-Calibration options for various measurement configurations
3
Operating Mode Measurement Type Calibration Type
4
Standard 8720 8720 with Option 085
5
Modulation during Calibration
Low PRF Transmission Response OK OK OFF
Pulse Profile Enhanced Response OK OK OFF
Full two-port OK OK OFF
Reflection Response Not Advised OK OFF
Full one-port Not Advised Not Advised Not Advised
Full two-port Not Advised Not Advised Not Advised
Low PRF Transmission Response OK OK Either
Point on Pulse Enhanced Response OK OK Either
Full two-port OK OK Either
Reflection Response OK OK Either
Full one-port OK OK Either
Full two-port OK OK Either
High PRF Transmission Response OK OK Either
Enhanced Response OK OK Either
Full two-port OK OK Either
Reflection Response Not Advised OK ON
Full one-port Not Advised OK ON
Full two-port Not Advised OK ON
1. A measurement uncertainty calculator is available on the web at www.agilent.com (Search for uncertainty calculator)
2. More details on error correction in are covered in Application Note 1287-3, Applying Error Correction to Network Analyzer Measurements, which is available
at www.agilent.com (Search for the number 1287-3)
3. Some of these entries are counterintuitive until they are viewed in terms of the measurement uncertainty equations. All combinations were verified empirically.
4. Two-port calibration assumes the DUT is bilateral, for example it transmits in both directions
5. Assumes the pulse modulator is inserted in the RF access loop provided by these instruments
To Next Page
Loading...
Need help?
General
