Simple questions about RBW setting on spectrum analyzer...

[dwluv3333]

Hi,

Have a question regarding spectrum analyzer and RBW setting...

On a spectrum analyzer, say you have a signal at 1 GHz and your resolution bandwidth is 300 kHz. If you use the 'marker' function and put the marker at 1 GHz, and it says -50 dBm, what exactly does this mean? Does it mean that the power in a bandwidth located at 1 GHz within a 300 kHz bandwidth (or, equivalently, 1 GHz +/- 150 kHz) is -30 dBm?

Also, I've heard that if you want to do an accurate power measurement of a signal, you need to set the RBW to the same bandwidth as the signal (so a 3.84 MHz WCDMA signal needs to have an RBW set to 3.84 MHz, etc). Is this true? I'm having a hard time understanding why the RBW isn't always best when set to its lowest possible value.

Thanks everyone in advance for your advice and help!!!!

 

[keith1200rs]

RBW is the width of the filter used to detect the RF - it doesn't give any information about the spread of the signal within the filter bandwidth. You could look at two different signals with a wide bandwidth and they could look the same, but as you reduce the RBW you could see that one is a single frequency and the other is made of multiple frequencies.

So, your marker of -50dBm means that if it is a single frequency, that is the size of the signal but you need to reduce the RBW and check you don't see something different.

Keith.

 

[dwluv3333]

Re: Simple questions about RBW setting on spectrum analyzer.

Thanks for the reply keith.

So my understanding is that the setting of the RBW affects the ability to resolve closely separated frequency signals. What about the level of the power measured? If you have -30 dBm at a frequency, does it mean that you are seeing -30 dBm/(RBW) Hz with that frequency at the center of that bandwidth?

Also if you could clarify any further on the other question I had regarding how wide to set the RBW to get an accurate power measurement, it would be much appreciated.

Thanks again!

 

[keith1200rs]

You can get an accurate power measurement with a wide RBW provided you know there is only a single frequency there. However, to make sure that is the case you need to reduce the RBW first to check. You should be able to check this yourself - feed a signal generator in and change the RBW and the power level won't change. Make sure you are using auto sweep time otherwise you can start to introduce errors.

A spectrum analyser is simply a mixer with a swept oscillator with the output going through a filter and peak detect. So you can imagine how it could be fooled with two frequencies close together which are within the filter bandwidth.

In interpreting the power you see, if you measure -30dBm, then that is the power - the RBW is irrelevant. If you try to take into account the RBW you are trying to integrate the result i.e. take into account the area but the area is artificially created by the RBW.

I could try and post some screenshots to help if you like - assuming you haven't a spectrum analyser & signal generator to try out.

Keith

 

[FvM]

Re: Simple questions about RBW setting on spectrum analyzer.

I understand, that the question is related to power measurement of wideband signals, e.g WCDMA. As a first simplified answer, to see the total signal power of the wide band signal, the RBW must be wide enough to allow the signal to fit the analyzer filter without attenuation. And detector should be operated in averaging (RMS) mode.

A more verbose answer has to confess, you possibly never get a correct power measurement of a wideband, time varying signal with a standard spectrum analyzer. Because the analyzer is scanning the band at a rate unrelated to the WDCMA symbol rate. Dedicated signal analyzers or analyzing options for recent spectrum analyzers should be able to, however.

 

[keith1200rs]

This is probably worth a read:

www.ictregulationtoolkit.org/en/Document.3588.pdf

Keith.

 

[snafflekid]

Interesting question. I am almost certain that the width of the bandpass filter makes no difference on the power measurement. The power is normalized to a bandwidth of 1 Hz.

 

[FvM]

Re: Simple questions about RBW setting on spectrum analyzer.

The power is normalized to a bandwidth of 1 Hz.

Not with any usual spectrum analyzer. It displays power, not power density.