Understanding the VSWR and Insertion Loss Plots
Pickering Interfaces RF product data sheets contain real plots taken from sample products. The information supplied in the plots is intended to help users anticipate the impact on the imperfections on an RF system.
On this page we examine some of the plots from a typical product and the consequences they might have on a system. For this page we have used the 40-876 terminated multiplexer as the example. However, the principals can be applied to any RF device.
How are the plots created?
We use a vector network analyzer (VNA) to create the plots. Before performing measurements we calibrate the network using a calibration kit and a set of test leads. Unless otherwise noted the measurements are performed in a 50Ohm system.The calibration process normalizes the RF measurements so that it reflects the performance of the switch as measured at the input and output connectors. The plots therefore include errors due to the module connectors, but not the characteristics of the mating connector or the connection cables. This is normal practice for measuring RF devices.
All the example plots were generated using a Keysight (formerly Agilent) 8.5GHz VNA with an LXI interface.
The VNA is used to measure the S parameters of the switch. More information on S parameters can be found in Wikipedia http://en.wikipedia.org/wiki/S-parameters.
S parameters model a device on how power flows into and out of the device terminals in a defined transmission line system. For the example used here, 40-876, the system is 50 Ohms so all the references are to that impedance. For 75 Ohm systems we use impedance adapters to perform the same measurements and calibrate the VNA accordingly.
S11, S22 and input VSWR
S11 is a measure of how close to 50 Ohms the input of the switch is when terminated in 50 Ohms. It is a measure of how much
signal power is reflected by the switch back to the source where the signal is absorbed and is a primary signal that
the VNA measures. Industry practice is to show this as the input Voltage Standing Wave Ratio (VSWR) and the VNA conveniently
converts its measurements to this format.
S22 is simply the same measurement on the output port of the switch. For switches the output and input impedance are similar, so generally we only show one port.
The VSWR is dependent on the path that the switch is set to, so on the plots we typically show a selection of the paths which include the best and worst paths measured on the sample. That can be quite a large number of plots. For the 40-876 example (excluding the terminated position) it has 4 switches each with four paths which produces 16 plots. In this case all the plots are present.
S21, S12 and insertion loss
This is a measure of how much power loss there is through the switch. The VNA assesses the ratio of the power going into the switch to the power coming out of the switch. As switches are passive devices the response is symmetric, so we simply show the loss for one direction. The VNA converts the information so the loss is expressed in dB and described as the Insertion Loss.
Again the switches can have more than one path, so again in the case of 40-876 we have 16 different plots and in this case show them all. In some larger LXI products we only show a few paths. As far as possible the 40-876 is designed to have a nominally equal loss on switches and paths, so the traces tend to be on top of each other.
An important factor to remember is that the insertion loss does not include losses due to the reflected signal from the
input impedance (S11, S22), a subject we will return to shortly.
What is important?
In an ideal system the VSWR would be 1 and the loss would be 0dB, in reality that will never happen but we try to get the best performance we can from the components we use. We try to minimize VSWR and insertion loss through the switch and maximize the usable BW. Low insertion loss does not guarantee low loss through a system, to get good performance users need both repeatable insertion loss and good VSWR as will be shown below.
Users sometimes assume that if the data sheet indicates an insertion loss of say 1dB and they have cables with a loss of 1dB then their system will have a loss of 2dB. Unfortunately this is not true, real systems do not have the same view of a component as a VNA does.
Below is a table copied from our PXImate book (an overview of the PXI standard together with useful information about the technology behind the switching and instrumentation modules a typical chassis can contain), the publication also has tables for uncertainty due to a source and load VSWR. (Download a free copy of our PXImate)
Let’s assume we have a signal at 1.5GHz which is conveniently the midpoint on the 40-876 plots horizontal axis. The VSWR is about 1.25 on the worst path and the insertion loss is 0.6dB
The path loss that might be measured in a system however will not be this. Assuming that the source and load are perfect then when the signal from the source arrives at the switch some of that signal is reflected. The reflection could be because the switch has either low or high impedance or have the right magnitude but not be entirely resistive compared to 50 Ohms, and this has a direct impact on how much signal actually goes through the switch to the load.
As the switch VSWR increases the added uncertainty of the signal levels increases rapidly. For a test systems requiring good RF accuracy we recommend using switches with typical VSWR of around 1.5 or less. VSWR numbers above 2 can only be reliably used for functional applications as the level uncertainties tend to rise and measurement become less meaningful.Other sources of VSWR (cables, connectors, signal source and load impedances) can add to the uncertainty so users need to select their connecting components carefully. Many cables and connectors can exhibit significant VSWR errors unless chosen carefully, and as always cheap solutions may not offer the best performance.
In some applications users use attenuators to help mask VSWR effects, in principle the insertion of a 3dB will improve return loss by 6dB.In many systems users apply correction data at the system level on the assumption the paths will be repeatable.
Bandwidth numbers are often treated as primary indicators of the frequency over which a switch can be used, but this is not always the case. The BW is usually quoted when the path loss reaches 3dB, however some designs can be used to well beyond their 3dB BW while others cannot. An example is clearly a length of coaxial cable which has an insertion loss of greater than 3dB if it is long enough but it can clearly pass the signal and is perfectly usable, especially when the user applies system level correction values to a path. What is usually important is that the frequency response is smooth (has no resonant structures) and the VSWR is reasonably low (generally <1.6) so reflections do not create additional uncertainty.
In some case it maybe crosstalk or some other parameter that controls the usable frequency range for a specific application. See our article:
Understanding Crosstalk and Isolation Plots
If good RF performance is required with accurate management of RF signal levels then it is important to select a RF switching products which have a good input VSWR and a connector with good RF characteristics. The availability of sample RF plots of Pickering Interfaces switching products helps users to assess the impact of the RF switches on their system and aids making key decision about switch selection and connector types.