Question:
I am confused about the frequencies in the XTrace list of my spectrum analyzer.
When e.g. the center frequency is set to 1 GHz with a span of 200 MHz, I would expect the first frequency point to be measured at 900 MHz.
After requesting "TRACE:DATA:X? TRACE1" I get "+9.001447178E+008" for the first measurement point. What is wrong here?
Answer:
When talking about measurements with a spectrum analyzer we have to keep some parameters in our mind.
Span
Defines the displayed frequency range and therewith the Start- and Stop Frequency defined by the Center Frequency +/- 0.5 *Span.
Number Of Sweep Points
Describes the division of the complete Span into subranges to be used for each sweep step.
So the bandwidth of each subrange is calculated when dividing the Span by the Number of Sweeppoints ).
First conclusion:
The analyzer will only capture a value in the middle of each definite subrange.
Second conclusion:
When having an odd number od sweeppoints, the only frequency that would match to one of the set parameters is the center frequency.
Let's take a deeper look using the following example:
Center Frequency: 1000 MHz
Frequency Span: 200 MHz
Number of Sweeppoints: 691
This defines
Start Frequency = Center Frequency - Span / 2 = 1000 MHz - 200 MHz / 2 = 900 MHz
Stop Frequency = Center Frequency + Span / 2 = 1000 MHz + 200 MHz / 2 = 1100 MHz
Step Size = Frequency Span / Number Of Sweeppoints = 200 MHz / 691 = 0.289435601 MHz
Frequency of the first measurement = Start Frequency + Step Size / 2 = 900 MHz + 0.289435601 MHz / 2 = 900.1447178 MHz.
Frequency of nth measurement = Start Frequency + Step Size * n - Step Size / 2
Frequency of 346th measurement = Start Frequency + Step Size * 346 - Step Size /2 = 900 MHz + 0.289435601 MHz * 346 -2 0.289435601 MHz / 2 = 1000 MHz
Why does it use the center frequency instead of the start frequency of each subrange?
Each measurement is done using a dedicated filter with a specified Resolution Bandwith.
If an adequate Resolution Bandwith is selected, the whole range will be covered by the measurement.
It also avoids "unsymmetric" measurement. If you would set the measurement frequency to the low end of each subrange for example, the whole sweep would have a gap at the upper end.
