Author Topic: Can Spectrum Analysers display phase if they have a Tracking Gen?  (Read 2226 times)

0 Members and 1 Guest are viewing this topic.

Online rs20

  • Super Contributor
  • ***
  • Posts: 2142
  • Country: au
I feel like this is a really n00bish question, hence the beginners section. Whether calculating the properties of a filter by hand, or doing simulations in LTSpice, you can come up with two figures for any given frequency: the gain, and the phase change (I've even forgotten the proper term for this).

If I build up that same filter in real life, and hook it up to, say, a Rigol DSA815-TG, can it show me the frequency vs phase change graph (as opposed to the standard frequency vs gain graph)? Or does it have to be a vector signal[/network] analyser for it to have that ability?
« Last Edit: February 23, 2015, 03:41:04 am by rs20 »
 

Offline w2aew

  • Super Contributor
  • ***
  • Posts: 1727
  • Country: us
  • I usTa cuDnt speL enjinere, noW I aR wuN
    • My YouTube Channel
Re: Can Spectrum Analysers display phase if they have a Tracking Gen?
« Reply #1 on: February 23, 2015, 03:43:59 am »
I feel like this is a really n00bish question, hence the beginners section. Whether calculating the properties of a filter by hand, or doing simulations in LTSpice, you can come up with two figures for any given frequency: the gain, and the phase change (I've even forgotten the proper term for this).

If I build up that same filter in real life, and hook it up to, say, a Rigol DSA815-TG, can it show me the frequency vs phase change graph (as opposed to the standard frequency vs gain graph)? Or does it have to be a vector signal[/network] analyser for it to have that ability?

It generally has to be a vector network analyzer. The DSA-815-TG and most all SAs do not have phase coherent receivers.
======================================
YouTube channel: https://www.youtube.com/w2aew
FAE for Tektronix
 

Online rs20

  • Super Contributor
  • ***
  • Posts: 2142
  • Country: au
Re: Can Spectrum Analysers display phase if they have a Tracking Gen?
« Reply #2 on: February 24, 2015, 09:27:55 am »
It generally has to be a vector network analyzer. The DSA-815-TG and most all SAs do not have phase coherent receivers.

Thanks! So if I have a control loop sitting in front of me, and I want to get a Bode plot / check the phase margin of that control loop, what's the standard way to do this if a network analyser can't do it (by itself)? I'd be surprised if a network analyser is the standard method to do that; those things seem to go up to GHz and cost megabucks. Obviously I'm missing something.
 

Online Howardlong

  • Super Contributor
  • ***
  • Posts: 4656
  • Country: gb
Re: Can Spectrum Analysers display phase if they have a Tracking Gen?
« Reply #3 on: February 24, 2015, 10:02:44 am »
What frequency are you interested in?

You can measure phase shift and thus margin with a signal generator and a scope assuming that the bandwidth of the scope is sufficient. Doing a Bode plot for both gain and phase can also be done on a scope and signal generator too but it'll take some time to do it manually.

As a minor aside, for lower frequencies the Analog Discovery may do what you want, it will do mag and phase to a few MHz.
 

Online rs20

  • Super Contributor
  • ***
  • Posts: 2142
  • Country: au
Re: Can Spectrum Analysers display phase if they have a Tracking Gen?
« Reply #4 on: February 24, 2015, 10:21:07 am »
What frequency are you interested in?

You can measure phase shift and thus margin with a signal generator and a scope assuming that the bandwidth of the scope is sufficient. Doing a Bode plot for both gain and phase can also be done on a scope and signal generator too but it'll take some time to do it manually.

As a minor aside, for lower frequencies the Analog Discovery may do what you want, it will do mag and phase to a few MHz.

Probably only a few MHz, yes. But the average lab wouldn't have an Analog Discovery, and doing the sweep by hand seems laborious. I guess I'm just overestimating the need for Bode plots of real life circuits; I suppose I had just invalidly assumed that it was standard practice to test this sort of thing.
 

Offline jpb

  • Super Contributor
  • ***
  • Posts: 1615
  • Country: gb
Re: Can Spectrum Analysers display phase if they have a Tracking Gen?
« Reply #5 on: February 24, 2015, 12:29:56 pm »
Probably only a few MHz, yes. But the average lab wouldn't have an Analog Discovery, and doing the sweep by hand seems laborious. I guess I'm just overestimating the need for Bode plots of real life circuits; I suppose I had just invalidly assumed that it was standard practice to test this sort of thing.
It depends how often the job needed doing. If it was frequently then it would probably be worth writing some software to automate a function generator and a scope.

Also high end counters can measure phase, so if you had a two channel function generator with one output going to the filter and the other being a reference bypassing it and attached them to the two channels of a counter that measures phase you could then get a phase plot fairly easily (I can't vouch for this as my counter is a cheap model that doesn't but the high end ones do and also can display both phase and frequency I think.) You could do it with one channel of course and just split the signal before the filter.

PS : the AD8302 measures gain and phase and you can buy a development board for it.
« Last Edit: February 24, 2015, 01:05:02 pm by jpb »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf