What does a Vector Network Analyzer actually measure???

[Joe]

Hello there, I have a quick question.

I'm an undergraduate physics major learning a few basics on Microwave
theory with regards to communications. I've been reading many papers
and I understand what S-parameters are and signal analysis along
transmission lines. My question is quite fundamental.

What does a Vector Network Analyzer (VNA) actually measure? All the
papers and tutorials I've been reading describe the VNA as measuring
the incident & reflected/transmitted signal, but none actually have
said what that signal actually is. Is it a voltage, current, or power
signal?? Or is it something else??

Thanks for your time

Joe

 

[Don Pearce]

Hello there, I have a quick question.

I'm an undergraduate physics major learning a few basics on Microwave
theory with regards to communications. I've been reading many papers
and I understand what S-parameters are and signal analysis along
transmission lines. My question is quite fundamental.

What does a Vector Network Analyzer (VNA) actually measure? All the
papers and tutorials I've been reading describe the VNA as measuring
the incident & reflected/transmitted signal, but none actually have
said what that signal actually is. Is it a voltage, current, or power
signal?? Or is it something else??

Thanks for your time

Joe

The signal is just RF power, as would come from a signal generator. It
is delivered at a series of spot frequencies - enough to look like a
continuous sweep on the screen. The power is delivered from a very
well matched 50 ohm source at port 1 of the analyser. The power at
that port is measured in both the forwards (out of the port) and
reverse directions. The measurement is done for both amplitude and
phase.

A second port is provided for items that need to connections -
filters, amplifiers etc. This is identical to port 1, and uses the
same power source and detection system.

During a measurement ports 1 and 2 will be used alternately as source
and load in order to end up with the four S parameter measurements
(S11, S21, S12 and S22). S21, for example means "what appears at port
2 as a result of applying a signal from port 1). S11 is the power
going back into port 1 from a signal coming out of port 1.

d

_____________________________

http://www.pearce.uk.com

 

[Paul Burridge]

During a measurement ports 1 and 2 will be used alternately as source
and load in order to end up with the four S parameter measurements
(S11, S21, S12 and S22). S21, for example means "what appears at port
2 as a result of applying a signal from port 1). S11 is the power
going back into port 1 from a signal coming out of port 1.

Would you say It's a highy useful - if expensive - piece of kit to
have around for the keen experimenter then, Don?

 

[Don Pearce]

Would you say It's a highy useful - if expensive - piece of kit to
have around for the keen experimenter then, Don?

Back in my designing days, I wouldn't have been without one. I started
out doing the job manually with slotted lines, diode detectors and a
pad of graph paper, but with the advent of the VNA, a day's work was
done in a few seconds.

You can make a scalar analyser quite easily with directional couplers
and a spectrum analyser, but if you really wan't to design properly,
and get things matched and stable, you do need the phase component of
the measurement.

So the answer is yes - expensive but useful.

d

_____________________________

http://www.pearce.uk.com

 

[John Larkin]

Would you say It's a highy useful - if expensive - piece of kit to
have around for the keen experimenter then, Don?

If you like sine waves. The time-domain equivalent is Time Domain
Reflectometry, TDR. A TDR is a fast 50-ohm step generator combined
with a fast sampling scope (3-20 GHz overall bandwidth is common.)
Data from a VNA or a TDR can, in theory, be Fourier transformed into
the opposite domain.

John

 

[Leon Heller]

Hello there, I have a quick question.

I'm an undergraduate physics major learning a few basics on Microwave
theory with regards to communications. I've been reading many papers
and I understand what S-parameters are and signal analysis along
transmission lines. My question is quite fundamental.

What does a Vector Network Analyzer (VNA) actually measure? All the
papers and tutorials I've been reading describe the VNA as measuring
the incident & reflected/transmitted signal, but none actually have
said what that signal actually is. Is it a voltage, current, or power
signal?? Or is it something else??

Paul, N2PK, has just set up a web page about a nice DIY VNA he has designed:

http://users.adelphia.net/~n2pk/index.html

Leon

 

[John Jardine]

Paul, N2PK, has just set up a web page about a nice DIY VNA he has designed:

http://users.adelphia.net/~n2pk/index.html

Leon

A first class link.
Pauls made a magnificent effort there. His project writeup is a pleasure to
behold!.
regards
john

 

[SioL]

A first class link.
Pauls made a magnificent effort there. His project writeup is a pleasure to
behold!.
regards
john

Indeed, a real joy. I am considering building the thing. Perhaps we should
organize something with the boards, if there's enough interest?

SioL

 

[Jeroen Belleman]

[...]
Data from a VNA or a TDR can, in theory, be Fourier transformed into
the opposite domain.

Oh, not just in theory. In practice too. Our HP8753D does it.
A wonderful piece of equipment, to be sure. Now if they could
only think of a more sensible way to get measurement results
into a PC...

Jeroen Belleman

 

[Paul Burridge]

Indeed, a real joy. I am considering building the thing. Perhaps we should
organize something with the boards, if there's enough interest?

Excellent suggestion.

 

[SioL]

Excellent suggestion.
--

Ok, we've got 2 people interested so far

SioL

 

[Active8]

Ok, we've got 2 people interested so far

SioL

i was planning on designing something like this any way. what kind of
"organizing" do you have in mind?

mike

 

[SioL]

i was planning on designing something like this any way. what kind of
"organizing" do you have in mind?

mike

Well, reading the instructions it turns out that there's some kind of a file
that you can send to the online PCB shop, Pcad 2001 won't read it so
I'm stuck. Can it be exported to some usable form from PCBExpress?

Anyway, here is an excerpt from the instructions:

Here's one way to obtain the main VNA printed circuit board from ExpressPCB.
ExpressPCB call this board size their "Miniboard." The minimum Miniboard order
is three identical boards for a total of $62 (US only) - i.e. three VNAs.

So one way to organize this is to submit the PCB and split it three ways (since there's
a minimum of three boards).

The other one would be to organize purchase of parts (those a bit harder to come by- ad etc).
I don't see any parts list anywhere, could that be in the ExpressPCB pcb files?

SIol.

 

[Active8]

Well, reading the instructions it turns out that there's some kind of a file
that you can send to the online PCB shop, Pcad 2001 won't read it so
I'm stuck. Can it be exported to some usable form from PCBExpress?

Anyway, here is an excerpt from the instructions:

i'm still dnlding the stuff *and* cleaning up the old, useless, 20 mile
flame wars posted here that show up in my reader, but i'm thinking he
might have a gerber X file (which can be read in a free gerber viewer)
or he used software from ExpressPCB.

So one way to organize this is to submit the PCB and split it three ways (since there's
a minimum of three boards).

that's one good idea. i'm betting he's using the ExpressPCB software.
another option is

https://secure.pcbfabexpress.com/index.jsp

$13 ea for 2 layer boards 5 min. they need gerber files and they
panelize the boards, so, depending on what size limit they have for this
board option, it may be possible to get a better deal.

The other one would be to organize purchase of parts (those a bit harder to come by- ad etc).
I don't see any parts list anywhere, could that be in the ExpressPCB pcb files?

if anyone can get parts as cheap as randy of

http://www.glitchbuster.com

that would be great. i think his PICs are not as cheap as the mfgs
"budgetary price", but i don't know what their min order is and all.

randy won't have a lot of parts for this project, but i'm using him as
an example of "good deal." he got my order yesterday and ran it to the
post office. OTOH timeline-inc.com won't ship my LCDs 'til next week
(once a week shipping!) they have me worried.

mike

 

[Active8]

On Fri, 17 Oct 2003 17:08:17 +0200, SioL, said...

The other one would be to organize purchase of parts (those a bit harder to come by- ad etc).
I don't see any parts list anywhere, could that be in the ExpressPCB pcb files?

the parts list is in the part 2 PDF, which mentions part 3 which i can't
find.

mike

 

[SioL]

On Fri, 17 Oct 2003 17:08:17 +0200, SioL, said...

the parts list is in the part 2 PDF, which mentions part 3 which i can't
find.

mike

I guess it is time to write to the author. Perhaps he wants to see
who is building it

Siol

 

[SioL]

I guess it is time to write to the author. Perhaps he wants to see
who is building it

Siol

I sent him an email. Let's see what he says.

SioL

 

[Russell Shaw]

Hello there, I have a quick question.

I'm an undergraduate physics major learning a few basics on Microwave
theory with regards to communications. I've been reading many papers
and I understand what S-parameters are and signal analysis along
transmission lines. My question is quite fundamental.

What does a Vector Network Analyzer (VNA) actually measure? All the
papers and tutorials I've been reading describe the VNA as measuring
the incident & reflected/transmitted signal, but none actually have
said what that signal actually is. Is it a voltage, current, or power
signal?? Or is it something else??

A sine-wave voltage source launches an incident wave to the DUT (device
under test). A directional coupler diverts a portion of the forward
wave to a VNA input and is used as the incident voltage measurement.
The reflected voltage wave from the DUT input is measured by the VNA from
the reflector-port output of the same coupler. The ratio
reflected/forward voltage is S11 (a complex number). The DUT output
goes thru another directional coupler then terminated in Zo (50ohms).
S11 on the output of the DUT can then be measured by switching the
source in place of the load (gives S22). The ratio
voltage-thru-DUT-into-load/incident-voltage-wave gives S21 and S12.
Scalar analysers use diodes to measure voltage or thermistors to
measure power, but phase is lost. Your research shows you're already
better qualified as a real EE than 95% of EE graduates these days,
all of which couldn't do a decent job of biasing a BJT.