Author Topic: SA video bandwidth affects signal level?  (Read 1798 times)

0 Members and 1 Guest are viewing this topic.

Offline jmwTopic starter

  • Frequent Contributor
  • **
  • Posts: 296
  • Country: us
SA video bandwidth affects signal level?
« on: March 21, 2023, 06:15:05 pm »
I have a R&S FPC-1500 and looped the tracking generator to the input. Evidently the signal level drops when I lower the VBW setting. Goes from about -1 dB (1 kHz VBW) to -1.5 dB (100 Hz) to -6 dB (30 Hz). I thought VBW just smooths the displayed detector output. Is this a bug in the SA?

 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3656
  • Country: fr
  • Analog, magnetics, Power, HV, Audio, Cinema
    • IEEE Spectrum
Re: SA video bandwidth affects signal level?
« Reply #1 on: March 22, 2023, 01:03:17 pm »
suggest that you read and learn about Spectrum Analyzers.

The sweep speed, res BW and freq span are related. As video filters,are,applied the observed level is affected.

our SA are,all vintage TEKTRONIX and HP without FFT,

Ask the manufacturer of your SA.

Jon
An Internet Dinosaur...
 

Online Kean

  • Supporter
  • ****
  • Posts: 2468
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: SA video bandwidth affects signal level?
« Reply #2 on: March 22, 2023, 02:13:09 pm »
My understanding is that lowering the VBW causes additional filtering/smoothing that affects the displayed trace.  You can see an example of this in the FPC getting started video, but not explained in any detail.  This is similar to the effect of reducing RBW, but it affects only display instead of the measurement.



I don't know that Paul has done any videos that go deeper into affects of VBW or details of the FPC tracking generator specs, but I do seem to remember a post of his about this topic.  I cant find it right now, so maybe I am misremembering who wrote it.

I am admittedly an RF novice, so I may be using the wrong terminology - best to watch some of the related videos.

Ping @pdenisowski
 

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3656
  • Country: fr
  • Analog, magnetics, Power, HV, Audio, Cinema
    • IEEE Spectrum
Re: SA video bandwidth affects signal level?
« Reply #3 on: March 22, 2023, 02:34:10 pm »
Rebonjour Cher Monsieur

The relationship between RBW, sweep speed, accuracy and sweep band is,well known.

tHe best application notes,are,the old HP, and Tektronix Circuits Concept books on Spectrum Analyzers circuits and SA measurement Concepts.

http://www.davmar.org/concepts.html

Jon
An Internet Dinosaur...
 
The following users thanked this post: jmw, Hexley, French_Nikolas

Offline switchabl

  • Frequent Contributor
  • **
  • Posts: 445
  • Country: de
Re: SA video bandwidth affects signal level?
« Reply #4 on: March 22, 2023, 04:11:25 pm »
The relationship between RBW, sweep speed, accuracy and sweep band is,well known.

According to the screenshot, the sweep time is set to "Auto", so the SA should adjust it according to span, RBW and VBW to maintain a calibrated display. Also, if this were an analog swept-tuned SA, I am not sure how an insufficient sweep time could result in what appears to be an offset trace, rather than a distorted or smoothed-out one.

I suspect that this might be a bug. However, AFAIK the FPC has a purely digital IF with both swept and FFT modes and the tracking generator is actually a programmable synthesizer, so it is hard to be sure what exactly is going on and I may be missing something.
 

Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: SA video bandwidth affects signal level?
« Reply #5 on: March 22, 2023, 05:15:34 pm »
I have a R&S FPC-1500 and looped the tracking generator to the input. Evidently the signal level drops when I lower the VBW setting. Goes from about -1 dB (1 kHz VBW) to -1.5 dB (100 Hz) to -6 dB (30 Hz). I thought VBW just smooths the displayed detector output.

Yes, normally VBW only effects the "smoothness' of the trace.  If you're measuring the amplitude of a CW signal, changing VBW shouldn't change the amplitude at all.

I don't have an FPC with the TG option in my lab at the moment, but here's an example on the FSW (our flagship analyzer). Essentially no change (< 0.1 dB) with VBW set automatically to 3 kHz, 1 Hz, and 80 MHz.

I'm working on getting an FPC + TG to try this out and see if I can reproduce the problem, but 6 dB of difference based on changes in VBW seems very, very strange to me.

I've also pinged our Technical Support Center on this and will let you know if they've ever seen this before (I haven't).

« Last Edit: March 22, 2023, 05:44:10 pm by pdenisowski »
Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 
The following users thanked this post: Kean, jmw

Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: SA video bandwidth affects signal level?
« Reply #6 on: March 22, 2023, 05:28:03 pm »
tHe best application notes,are,the old HP, and Tektronix Circuits Concept books on Spectrum Analyzers circuits and SA measurement Concepts.

We also have a nice (modern) spectrum analyzer fundamentals primer:

https://www.rohde-schwarz.com/us/products/test-and-measurement/analyzers/signal-spectrum-analyzers/spectrum-analyzer-fundamentals_253351.html

It's essentially a shortened version of our "Fundamentals of Spectrum Analysis" (Grundlagen der Spektrumanalyse) book by Rauscher. 

Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 
The following users thanked this post: jmw

Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: SA video bandwidth affects signal level?
« Reply #7 on: March 22, 2023, 05:43:03 pm »
One additional observation that I probably should have made in my earlier post:  as many others have mentioned, RBW, VBW, and sweep time are normally automatically coupled in modern spectrum analyzers.

If sweep time is too low, the video filter will not have time to settle and the result will be reduced amplitude and/or a frequency offset in the received signal.  This can happen when VBW is < RBW, which is the case in the screenshot.  I think the "(uncal)" message is complaining about sweep time being too short for the configured parameters (although I do see SWT showing "Auto").

Is there a reason why you want to manually set the VBW when using the tracking generator (which has a CW output)? 
« Last Edit: March 22, 2023, 05:50:21 pm by pdenisowski »
Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 
The following users thanked this post: Kean

Offline switchabl

  • Frequent Contributor
  • **
  • Posts: 445
  • Country: de
Re: SA video bandwidth affects signal level?
« Reply #8 on: March 22, 2023, 06:03:34 pm »
If sweep time is too low, the video filter will not have time to settle and the result will be reduced amplitude and/or a frequency offset in the received signal.  This can happen when VBW is < RBW, which is the case in the screenshot.  I think the "(uncal)" message is complaining about sweep time being too short for the configured parameters (although I do see SWT showing "Auto").

Some SAs definitely display "UNCAL" when you set an insufficient sweep time but I think on the FPC it just means that TG response is not normalized, no? In any case, I don't know why this should ever happen on Auto.
 

Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: SA video bandwidth affects signal level?
« Reply #9 on: March 22, 2023, 06:06:01 pm »
If sweep time is too low, the video filter will not have time to settle and the result will be reduced amplitude and/or a frequency offset in the received signal.

Here are some screenshots (again, from an FSW) showing this phenomenon.  With RBW = 500 kHz and SWT held constant=~ 1 ms, you can see that when VBW is manually lowed to 30 kHz, the measured amplitude drops 2 dB and when VBW is manually lowed to 10 kHz, the measured amplitude drops almost 15 (!!!) dB.

This is what I presume is happening in your case.  I'll see if I can reproduce this on an FPC.

(You can also see the frequency inaccuracy when VBW is too low for sweep time)

Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 

Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: SA video bandwidth affects signal level?
« Reply #10 on: March 22, 2023, 06:06:40 pm »
Some SAs definitely display "UNCAL" when you set an insufficient sweep time but I think on the FPC it just means that TG response is not normalized, no? In any case, I don't know why this should ever happen on Auto.

Agree.  This is something I need to check on.
Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 

Offline Hexley

  • Regular Contributor
  • *
  • Posts: 209
  • Country: us
Re: SA video bandwidth affects signal level?
« Reply #11 on: March 22, 2023, 06:40:37 pm »
Some experimental results from an FPC1500 with TG:
  • The OP's results could be duplicated when the SA was first powered up and the TG was enabled. Changing the VBW affected the amplitude.
  • pdenisowski's results could also be duplicated when the SA was changed from TG to CW mode, and a single 550 kHz tone was measured. In this case, changing the VBW did not affect the amplitude.
  • When the SA was then changed from CW back to TG, the amplitude no longer changed as VBW was varied.

This sequence seems to be reproducible. After powering down the SA, I could repeat the sequence above. Sort of smells like a data initialization bug somewhere.
 
The following users thanked this post: Kean, pdenisowski, jmw

Offline jmwTopic starter

  • Frequent Contributor
  • **
  • Posts: 296
  • Country: us
Re: SA video bandwidth affects signal level?
« Reply #12 on: March 22, 2023, 06:58:46 pm »
Wow, that's interesting Hexley! I didn't think to try that sequence - I'm in a different building at the moment, but will see if I can confirm a little later today. Edit: able to reproduce the sequence & behavior described by Hexley. Once source is set to CW and back to TG, I'm actually seeing the opposite effect: the TG signal appears to get stronger at very low VBW. At 10 Hz VBW, the -20 dBm TG output reads as -19.81 dBm.  :-//

Yeah, it says "uncal" only because I didn't run a S21 normalization aka response calibration before taking the measurement. The sweep time effect on filter charge-up is a valid point, I wonder if I manually increase sweep time whether it will change the offset level. I'll give that a try too. Edit: TG forces Auto sweep time, can't force it higher or lower. With CW output, forcing a low sweep time distorts the signal but it also puts a red warning indicator on the screen to tell you that. No such indication when the TG is active with auto sweep time.

The reason I am cranking down the VBW is to clean up traces of measurements close to the noise floor. I'm doing shunt-thru measurement of milliohm resistors. S21 expected be about -70 to -90 dB below the TG output.
« Last Edit: March 22, 2023, 08:19:00 pm by jmw »
 
The following users thanked this post: pdenisowski

Offline Detlev

  • Regular Contributor
  • *
  • Posts: 59
  • Country: de
Re: SA video bandwidth affects signal level?
« Reply #13 on: March 22, 2023, 07:29:40 pm »
Hello,

I just recreated it, I hope all the parameters are the same.

First VBW to 1kHz, normalized and marker set and then VBW to 30Hz, the attenuation is 0.13dB on my FPL:



A good document on SAs is from Rohde & Schwarz

https://www.rohde-schwarz.com/de/produkte/messtechnik/analyzers/signal-und-spektrumanalysatoren/educational-note-fundamentals-of-spectrum-analysis-register_252824.html

(Thank you R&S for that  :) )



This post is "Made in Germany" 😎
 
The following users thanked this post: pdenisowski

Offline pdenisowski

  • Frequent Contributor
  • **
  • Posts: 929
  • Country: us
  • Product Management Engineer, Rohde & Schwarz
    • Test and Measurement Fundamentals Playlist on the R&S YouTube channel
Re: SA video bandwidth affects signal level?
« Reply #14 on: March 22, 2023, 09:09:41 pm »
First, many thanks to jmw, Hexley, and Detlev for looking into this.

I've been working with the R&S North America support team this afternoon and this issue does seem strange enough that we'd like you to open a case.   

Could you send an email to Support@rohde-schwarz.com and describe the issue?  Please let them know that you've been working with me (Paul Denisowski) and include your description of the issue (as well as what Hexley found).  I'm assuming you're on the latest FW since they definitely will ask that.

Once they get the email they'll open a case and most likely they'll also contact me (internally) directly.  I've PM'ed you my email address already, so please reach out to me directly anytime as well.

Thanks again for the detailed description and all the investigation: by the time I had arranged for remote access to an FPC you had all beat me to it :)
« Last Edit: March 22, 2023, 09:13:18 pm by pdenisowski »
Test and Measurement Fundamentals video series on the Rohde & Schwarz YouTube channel:  https://www.youtube.com/playlist?list=PLKxVoO5jUTlvsVtDcqrVn0ybqBVlLj2z8
 
The following users thanked this post: Kean, jmw, Hexley, Detlev

Offline Richard Head

  • Frequent Contributor
  • **
  • Posts: 685
  • Country: 00
Re: SA video bandwidth affects signal level?
« Reply #15 on: March 24, 2023, 12:34:47 pm »
The TG and SA need to track properly in order for the amplitude to be correct. At narrow RBW settings (<100Hz) it gets quite critical as a very small freq error between the TG and SA receiver can cause the TG output to fall outside (or on the filter skirt of) the IF filter passband. The calibration procedure just requires a coax between the TG and the SA input and when the cal button is pressed the instrument will align the two so the trace is level. This is normally only necessary when very narrow RBW's are used such as when checking the response of crystal filters. The newer SA's probably do it automatically now.
 
The following users thanked this post: pdenisowski

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 3656
  • Country: fr
  • Analog, magnetics, Power, HV, Audio, Cinema
    • IEEE Spectrum
Re: SA video bandwidth affects signal level?
« Reply #16 on: March 24, 2023, 01:52:10 pm »
I cannot testify on the R&S, or,modern FFT SA.

A TG is not a,good test, instead use a single tone, with any traditional swept /LO/mixer/IF analog SA, eg Tektronix 7L5, 7L12, 7L13 or HP 4195A, HP8558B, etc.

change sweep speed with no video,filter observations of shape of tone displayed and level at center.

As,sweep speed is increased, at some,point level at peak decreases and shape is distorted and asymmetrical

The relationship between sweep speed, RBW, sweeping width of a SA is very well documented
See the already mentioned Tektronix Circuits Concept book Spectrum Analyzers.

In Signal processing, the same limitations must apply to FFT Analyzers.

With Kind Regards

Jon
An Internet Dinosaur...
 

Offline jmwTopic starter

  • Frequent Contributor
  • **
  • Posts: 296
  • Country: us
Re: SA video bandwidth affects signal level?
« Reply #17 on: March 24, 2023, 03:08:25 pm »
The TG and SA need to track properly in order for the amplitude to be correct. At narrow RBW settings (<100Hz) it gets quite critical as a very small freq error between the TG and SA receiver can cause the TG output to fall outside (or on the filter skirt of) the IF filter passband.
This could explain why the FPC constrains the RBW to a minimum of 1 kHz when using the tracking generator.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf