# Using a Spectrum Analyzer to measure VSWR

I need to measure the VSWR of a coaxial wire that has a length of 100 metres and a characteristic impedance of 50 ohms using a Spectrum Analyzer (I don't have a Network Analyzer). I found this website: https://rigol.desk.com/customer/en/portal/articles/2270018-how-do-i-measure-vswr-using-a-spectrum-analyzer-

And I don't know if I understood right the testbench. Would it be something like this?

If that is OK, what do I have to put in the green box? A short circuit? Open circuit? Why?

The coupler is a unidirectional, isn't it? I understand how a coupler works but I don't know how the tracking generator and the coupler work together.

Thanks

• put the 50 ohm load on the end of the wire. – Jasen Aug 5 '18 at 20:57
• ok. Can you please explain me why? I understand that, ideally, I would not have any reflected signal. – Dylan Aug 5 '18 at 20:57

I need to measure the VSWR of a coaxial wire that has a length of 100 metres and a load of 50 ohms ... what do I have to put in the green box?

The green box is the 50 ohm load you said you want to have on the end of your cable when you measure its VSWR.

Why?

Because you said you want the VSWR of a cable with a 50 ohm load, not the VSWR of a cable with an open or short load.

The coupler is a unidirectional, isn't it?

We usually call it a directional coupler.

The port connected to the input of the spectrum analyzer is the one that gets most of the energy from the signal returning from the device under test (DUT).

Internally there is another port that gets mostly energy coupled from the generator. But this port is already terminated with 50 ohms inside the coupler device so you see the coupler as a 3-port device rather than the 4-port device it actually is.

I understand how a coupler works but I don't know how the tracking generator and the coupler work together.

The tracking generator generates a signal at the same frequency that the spectrum analyzer is measuring at any given moment (it tracks the spectrum analyzer measurement).

So the generator sends a signal out to the DUT. The reflection from the DUT is coupled by the coupler to the spectrum analyzer input and its power is measured.

If you know the power sent by the generator (you'll have to do another measurement to find this) and the coupling ratio of the coupler (yet another calibration measurement), then you know the reflection coefficient ($\Gamma$) of the DUT.

And from $\Gamma$ you can find the VSWR.

Since the tracking generator and spectrum analyzer sweep their operating frequency together, you can use one sweep of the spectrum analyzer (and some post-processing) to measure the VSWR of the DUT as a function of frequency.

Edit

the 50ohms is the characteristic impedance of the coaxial cable, not the load. SOrry.

Then how you should terminate it depends what you want to know about the cable.

If you want to know how well matched it is to 50 ohms, use a (very accurate) 50 ohm load.

If you want to know how long it is, or what's its attenuation, then use a high-reflection load. A well-made short will probably provide slightly better accuracy for these measurements than just leaving the cable end open.

• THe Photon, the 50ohms is the characteristic impedance of the coaxial cable, not the load. SOrry. So, how can I send some amount of power and see the power that is returned? Is there any option in the Spectrum Analyzer? In the link provided by Rigol, they do something with some markers but I don't understand what – Dylan Aug 5 '18 at 21:28
• @Dylan Well what load do you want to have attached for the VSWR measurement? VSWR measures the power reflected from a load. What load do you want to measure the power reflected from? – immibis Aug 5 '18 at 22:56
• I don't know. I suppose that if I'm using a directional coupler I need a load of 50 ohms so, ideally, I don't have any reflected signal – Dylan Aug 6 '18 at 4:04
• @Dylan, the reflected signal is what you want to measure. If you have no reflected signal you have nothing to measure (but if you're trying to measure the cable Z0, you want to measure "how small is the small reflection?") The point of the directional coupler is to deliver as much of the reflected signal to the measurement instrument as possible, and as little of the forward going signal from the generator as possible. – The Photon Aug 6 '18 at 4:08
• @ThePhoton so, I would put a load of 50 ohms (the same of the characteristic impedance of the cable). Is that ok? – Dylan Aug 7 '18 at 3:00

if you're trying to measure the VSWR of the wire put the 50 ohm load on the end of the wire.

if you want to see more reflections leave the end of the wire open or short it. That should cause a reflection peak at around 1Mhz and harmonics thereof. the exact frequency will tell you how fast signals travel in that wire (I'm guessing about 2c/3 )

if you measure with no wire and then re-measure with the wire with no terminator the difference could tell you how much signal is lost in 200m of wire (100m to the end and 100m back to the measuring device)

so there's a bunch of different measurements you can make with that setup.

• Yes, I understand that. If I put 50 ohms in the green box, I could use a unidirectional coupler. If I put an open circuit I would have reflections, so I would need a bidirectional coupler. Is that ok? – Dylan Aug 5 '18 at 21:06
• what are you trying to measure? – Jasen Aug 5 '18 at 21:29
• The VSWR and the attenuation of a coaxial cable that has a characteristic impedance of 50 ohms and a length of 100 metres – Dylan Aug 6 '18 at 4:04

In your question, you state that the load is 50 ohms. The green box is the load. Or, perhaps you meant that the impedance of the coaxial line is 50 ohms...?

• *In your case, the green box would be a 50 ohm non-inductive resistor of sufficient wattage to dissipate the power generated by the RF source. – Randy Venable Aug 5 '18 at 21:02
• Yes, I modified the main post. 50 ohms is the characteristic impedance – Dylan Aug 5 '18 at 21:05