Return Loss – A new take on VSWR measurement

 

Well, first some background – and definitions

I’ll give some basic definitions and descriptions… if you want the full detailed technical description, try here http://microwaves101.com/encyclopedia/index.cfm

Network Analyzer

A network analyzer is essentially a spectrum analyzer with a signal generator that follows it – both of them are locked in frequency. Many service monitors have this option, some spectrum analyzers have what is called a “tracking generator”, which makes them essentially the same as a scalarnetwork analyzer. There is also a device called a vector network analyzer, which adds some other functionality… more detail than is needed here.

Directional Coupler

A directional coupler is a device that goes inline with an RF feedline – we’ll use coax in this case. A directional coupler has either 3 or 4 ports:

  • an Input
  • an output
  • a forward coupled port
  • (Optional)a reverse coupled port

The forward coupled port will have a sampled level of the RF going form the input to the output, and will NOT have significant amounts of energy relating to what is going form the output to the input port. The reverse port will be exactly the opposite.

The (forward) coupled port will have a signal that is proportional to, but reduced from, the amplitude of the signal on the input – this coupled port will have a rating in dB – the loss from the input to the coupling output. Most RF work with equipment like this is done in terms of dB units; i.e. dBm, dBuV, etc. Please see the Microwaves101 Encyclopedia for a description of Decibels and log units, such as dBm.

 

Back to the story – I was introduced to the concept of return loss while working on a business project with Mike Mertel, K7IR,of SteppIR antennas in 2012. Mike was using return loss as a way to monitor SWR on an antenna, using a dual directional coupler. We had spectrum analyzers connected to both the forward and reflected ports, and were monitoring the difference between forward and reverse ports. The difference between the two ports is the return loss.

In 2013, I was contemplating the purchase of a spectrum analyzer with a tracking generator. One option that was available was an SWR bridge. This got me to thinking – what exactly is an SWR bridge… well, then I digressed into what this passive device would have been doing to display SWR as a function of frequency. It had been a while since I thought about what SWR was. Several years ago, I had even written a TI-89 program to calculate SWR based on forward and reflected power. At the time, I had a Bird wattmeter, and wanted to easily determine SWR based on those measurements – not just tune for a minimum reflected power. (I’m digressing again). The formula to calculate SWR is as follows:

SWR Formula

(From http://emc.toprudder.com/)

What this ends up meaning is that less reflected power means lower SWR (yeah, I know… very simple concept). It dawned on me that an SWR bridge is not much more than a directional coupler. After some calculation and messing about with the overall concept, it made a lot of sense. The final diagram that I put together to illustrate my idea is below:

SpecAnSetup

Now – one question that might arise is “why?” – an SWR meter or directional wattmeter is simple enough to use. This application has several benefits over a KISS (Keep It Simple, Stupid!) setup. First and foremost, it’s VISUAL. It’s easy to see how different things affect the SWR quickly. Set a marker or two on your analyzer, and the nulls will show where the system is resonant, or at least matched. Second, this application can generate an overall SWR map of a system over a wide frequency range as quickly as the analyzer can sweep it (milliseconds – how long does it take you to measure every frequency in the 2 meter ham band?). Finally, this method, is highly accurate, so long as one understands how to take into account the losses of the coupler, coax, etc.

The one downside to this method is that one needs to do math in order to get the actual SWR measurements. This is where excel, or a similar tool comes in handy. The table below uses a 0 dBm signal out, and converts the reflected power level to an equivalent SWR reading. Note that this does NOT take into account the directional coupler coupling factor. What this means is that the reflected power level will actually be higher than what the analyzer sees and reports (some analyzers have a correction function that will account for this in the reading). If a 10 dB coupler is being used, then, in order to get -10 dBm reflected power, the analyzer would need to read -20 dBm – the coupler factor would need to be ADDED to the displayed measurement.

 

 

nput
measured
VSWR
dBm
mW
dBm
mW
0
1
0
1
#DIV/0!
0
1
-3
0.501187
5.848044
0
1
-5
0.316228
3.569771
0
1
-6
0.251189
3.00952
0
1
-9
0.125893
2.099878
0
1
-10
0.1
1.924951
0
1
-12
0.063096
1.6709
0
1
-15
0.031623
1.432581
0
1
-20
0.01
1.222222
0
1
-25
0.003162
1.11917
0
1
-30
0.001
1.065311
0
1
-35
0.000316
1.036209
0
1
-40
0.0001
1.020202
0
1
-45
3.16E-05
1.01131

Note the DIV/0 in the first column – this means the SWR is essentially infinite, as ALL of the forward power is being reflected. Normally, I would do all my work in terms of dB, but since the formula above is in terms of linear units, I had to convert to linear units from dBm – thus the mW measurements – but most spectrum analyzers default to dBm as a measurement scale.

Despite the arrowheads shown in the diagram, there are several things to keep in mind.

  • First, the input and output ports work together to create the “Thru” path. This path is bidirectional, and essentially a straight wire.
  • Second, the coupled ports on directional couplers are frequency dependent – that means that a 1 – 2 GHz directional coupler will not perform well at 30 MHz. It may show SOME power – but the coupled port will not be matched, and the coupling loss will be nowhere near the coupler’s specified loss value.
  • Third, while not pertinent to this discussion, but to spur on others to consider ways to utilize equipment – the coupled ports can also be used to inject signals onto the thru path. They will experience the same loss (so injecting a 0 dBm signal into a 10 dB coupler would result in a -10 dBm signal appearing on the thru path).

 

I have not put this concept to use yet, simply because I don’t have an analyzer at home yet (as of the writing of this article), but it is a sound concept, which I have seen parts of used before. One challenge to many people who might want to test this out is where to find directional couplers. Typically, www.minicircuits.com has a decent selection, and www.surplussales.com has a few in-stock from time to time. Two items to keep in mind when shopping for a directional coupler:

  • Power rating: do not expect to use a coupler rated at 1W average, 2W peak with your 100W HF transceiver. It will no longer function when you are done. The above situation uses low powers, such as 0 dBm / 1 mW for a reason.
  • Frequency range: as previously stated, directional couplers are frequency sensitive – a single coupler can only cover a small portion of the overall DC to Daylight spectrum. Make sure you are using an in-band coupler to get accurate results.

 

Leave a Reply