

Number of Points, if Bandwidth, Sweep Time ?  RF Cafe Forums

mbergmann 
Post subject: Number of points, if bandwidth, sweep time ?
Posted: Tue Jun 20, 2006 5:14 am


Joined: Mon Jun 19, 2006 10:27 am
Posts: 13 
Hello,
can anyone please explain to me,
the relationship between  number of points
 if bandwidth  sweep time ?
For
example, how do I know which bandwidth I need
if the number of points and sweep time are given
?





Kirt Blattenberger 
Post subject:
Posted: Tue Jun 20, 2006 9:57 am



Site Admin 

Joined: Sun Aug 03, 2003 2:02 pm
Posts: 476 Location: Erie, PA 
Greetings mbergmann: The relationship
varies with each instrument. The usuer's guide
usually provides a formula and/or table, such
as Table 153 (page 149) in this Agilent manual
for the ESA and PSA series spectrum analyzers.
There are relational formulas throughout the
programming examples.
http://cp.literature.agilent.com/litweb ...
90482.pdf
_________________  Kirt Blattenberger
RF Cafe Progenitor & Webmaster





mbergmann 
Post subject:
Posted: Wed Jun 21, 2006 7:55 am


Joined: Mon Jun 19, 2006 10:27 am
Posts: 13 
Thanks Kirt for your answer, but does it really
depend on the instrument ? An explanation
is given on
http://kuprism.org/video/JohnPaden/network_analyzer_oper.htm
http://kuprism.org/update/SummerSeminar/MrPaden.html
According to this, the number of points
is determined by the maximum length of the impulse
response and the bandwidth of measurement:
number of points = length of impulse response
* bandwidth of measurement First of all I
don't understand this simple formula, secondly
how do I know the impulse response of my DUT
? Generally, that's unknown I would say.
The necessary IF bandwidth is determined
by the desired dynamic range (SNR).
Finally, the sweep rate is calculated by the
bandwidth, number of points and group delay
through the IF BWfilter (approximately the
reciprocal value of the chosen IF BW). The sweep
rate, together with the length of the impulse
response results in the change in frequency
during the impulse response, which must be lower
than the IF bandwidth ("the sweep time is always
okay if you are properly sampling the frequency
domain!").


Posted 11/12/2012



