Agilent Technologies N9010A - Manual (Page 24)

24 Chapter 1
Agilent EXA Signal Analyzer
Frequency and Time
Description Specifications Supplemental
Information
Frequency Readout Accuracy
±(marker freq × freq ref accy. + 0.25%
× span + 5% × RBW a
+ 2 Hz + 0.5
×
horizontal resolution
b
)
Single detector only
c
Example for EMC
d ±0.0032% (nominal)
a. The warranted performance is only the sum of all errors under autocoupled conditions. Under
non-autocoupled conditions, the frequency readout accuracy will nominally meet the specification
equation, except for conditions in which the RBW term dominates, as explained in examples below.
The nominal RBW contribution to frequency readout accuracy is 2
% of RBW for RBWs from 1 Hz to
390 kHz, 4
% of RBW from 430 kHz through 3 MHz (the widest autocoupled RBW), and 30% of RBW
for the (manually selected) 4, 5, 6 and 8 MHz RBWs.
First example: a 120 MHz span, with autocoupled RBW. The autocoupled ratio of span to RBW is
106:1, so the RBW selected is 1.1 MHz. The 5
% × RBW term contributes only 55 kHz to the total fre-
quency readout accuracy, compared to 300 kHz for the 0.0.25
% × span term, for a total of 355 kHz. In
this example, if an instrument had an unusually high RBW centering error of 7
% of RBW (77 kHz) and
a span error of 0.20
% of span (240 kHz), the total actual error (317 kHz) would still meet the computed
specification (355 kHz).
Second example: a 20 MHz span, with a 4 MHz RBW. The specification equation does not apply
because the Span: RBW ratio is not autocoupled. If the equation did apply, it would allow 50 kHz of
error (0.25
%) due to the span and 200 kHz error (5%) due to the RBW. For this non-autocoupled RBW,
the RBW error is nominally 30
%, or 1200 kHz.
b. Horizontal resolution is due to the marker reading out one of the sweep points. The points are spaced
by span/(Npts –1), where Npts is the number of sweep points. For example, with the factory preset
value of 1001 sweep points, the horizontal resolution is span/1000. However, there is an exception:
When both the detector mode is “normal” and the span > 0.25
× (Npts –1) × RBW, peaks can occur
only in even-numbered points, so the effective horizontal resolution becomes doubled, or span/500 for
the factory preset case. When the RBW is autocoupled and there are 1001 sweep points, that exception
occurs only for spans > 750 MHz.
c. Specifications apply to traces in most cases, but there are exceptions. Specifications always apply to the
peak detector. Specifications apply when only one detector is in use and all active traces are set to Clear
Write. Specifications also apply when only one detector is in use in all active traces and the "Restart"
key has been pressed since any change from the use of multiple detectors to a single detector. In other
cases, such as when multiple simultaneous detectors are in use, additional errors of 0.5, 1.0 or 1.5
sweep points will occur in some detectors, depending on the combination of detectors in use.
d. In most cases, the frequency readout accuracy of the analyzer can be exceptionally good. As an exam-
ple, Agilent has characterized the accuracy of a span commonly used for Electro-Magnetic Compatibil-
ity (EMC) testing using a source frequency locked to the analyzer. Ideally, this sweep would include
EMC bands C and D and thus sweep from 30 to 1000 MHz. Ideally, the analysis bandwidth would be
120 kHz at
−6 dB, and the spacing of the points would be half of this (60 kHz). With a start frequency
of 30 MHz and a stop frequency of 1000.2 MHz and a total of 16168 points, the spacing of points is
ideal. The detector used was the Peak detector. The accuracy of frequency readout of all the points
tested in this span was with
±0.0032% of the span. A perfect analyzer with this many points would
have an accuracy of
±0.0031% of span. Thus, even with this large number of display points, the errors
in excess of the bucket quantization limitation were negligible.