Return Loss Measurement Basics

[E Kafeman]

A VNA and a time discriminator is an easy measurement method to analyze different types of cable faults but very unreliable for long cables which requires long time delay and real impedance/phase becomes hard to measure due to cable discontinuities. Using frequencies in GHz range makes it also almost impossible to get good time reference. An example about how a TDR instrument can be done with simple tools and almost no cost: TDR Circuit. It gives a more "hands on" understanding how these kind of instruments are working. Add some processing power and you have a professional TDR instrument.

 

[FvM]

There have been previous discussions on the same topic, e.g https://www.edaboard.com/threads/195492/ It has already been emphasized, that classical cable fault locators used TDR techniques. The simple design and direct readout without complex signal processing are obvious advantages. These methods are still valuable today.

Later on, instrument manufacturers (e.g HP/Agilent and Anritsu) offered time domain display software supplements to their network analyzers and started arguments about FDR being superior to TDR instruments in fault locations and similar applications.

I don't feel a need to strictly contradict or agree to these arguments. I assume, that both methods can achieve similar results in a wide range of applications.

 

[cmd]

Yes you are right FvM,

We had talked about TDR and FDR and their advantages. TDR theory is basic and easy to understand and apply. But FDR offer a selected frequency range measurment which is important for pass band applications. Also measurement at high frequency provide to see small faults on cables. With TDR you can make a measurement at high frequency with short pulses (<2-3ns) but the reflections may not arrive back to the test port because of the attenuation.

 

[cmd]

According to the distance to fault application notes from Anritsu, it has to be measured magnitude and also phase information between incident and reflectected signals from the cable in order to get transform frequency domain data to time domain data. That's why for distance to fault measurement it should be used a vector network analyzer or VNA based cable and antenna anlyzer. Scalar network analyzers and spectrum analyzer plus a tracking generator can only measure magnitude data which is not enoug for time domain transfom.
Is it true? Because i also read some documents in which is claimed that spectrum analzyer and scalar network analyzer can also distance to fault measuremet. There is a contradiction between application notes. I have a confusion because of that...

 

[FvM]

Yes, a scalar reflection measurement can't give the same informations as a TDR or vectorial FDR measurement. It should be sufficient to determine singular cable faults.

 

[cmd]

Hi FvM,
Thanks for your reply. You said the phase information is necessary for distance to fault measurements. The Anritsu's DTF aplicatin note says"The FDR technique requires a swept frequency RF signal. The Frequency Domain Reflectometry principle involves vector addition of the sources output signal with reflected signals from faults and other reflective characteristics within the transmission line"
I can't understand this part of application note. We have a reference signal from source and reflected signals with different magnitude and phase from different discontinuites in the transmission line. The reflected signals are sin waves and they are continuos. How does a VNA isolate the reflected sin waves from each other and compare them with reference signal and create a one magnitde and one phase value for all frequency points in frequency range? I can't visualize it in my mind.

 

[Amithsleek]

Hi cmd,

The formula for return loss is RL=10*log(Pr/Pi).
1)In return loss positive is loss and negative implies gain.
2) s11 parameters are used.
3) There is no relation between return loss and the cable loss. When cable length is more then the cables loss is increased.
4) Yes it affects your fault measurement accuracy.