VNA FAQ: What are the Key Network Analyzer Specifications?
Engineering is hard. RF engineering is even harder. Complicated concepts, tons of terminology—it’s information overload. Even the simplest of questions have convoluted answers, making it difficult to know how to even get started learning.
But don’t worry, you’re in the right place. Welcome to the third installment of the “VNA FAQ” blog series! In each blog of this series, I’ll provide bite-sized answers to a handful of related, common VNA questions. Today, we'll take a look at key network analyzer specifications, or “specs” for short, including:
- frequency range
- dynamic range
- output power
- number of ports
- trace noise
- sweep time
Vector network analyzers act as both a signal generator and receiver, so while these six specs definitely don’t cover absolutely everything, they provide the context you need to begin your search for the perfect VNA.
Why do we care about frequency range?
You’d probably guess that the frequency range of your vector network analyzer (VNA) is important because it allows you to test the full frequency range of the device you want to characterize. And you’d be right. If you test broadband components, like those in a high-speed digital system, then you need broadband coverage from your VNA. While “broadband” does not have a strict definition, in this context, think from low megahertz to gigahertz.
If your device under test (DUT) operates in a specific frequency band, for example, up to 6 GHz let’s say, then you need a VNA that can reach at least 6 GHz. Actually, for a 6 GHz amplifier, engineers often like to look at spurious harmonics that fall outside of the DUT’s operation band. Doing so helps them design or choose filters with the necessary stopband attenuation. So, for an amplifier working at 6 GHz, engineers might analyze to 12 or 18 GHz.
With this in mind, should you always buy the VNA with the highest max frequency? A 67 GHz max is clearly much better than a 3 GHz max every time, right? Well, I know every engineering student hates this answer, but…it depends. I know, I’m sorry! But before you grab your pitchforks, let me explain.
With high frequency comes high cost. So, you should only pay for the frequency range you need. If you are developing advanced millimeter wave (mmWave) systems or even sub-terahertz tech, then it is worth the investment to get high-frequency equipment. But, if you aren’t on the bleeding edge of the market, then you probably don’t need the highest frequency coverage out there. To make it simple, here's the general rule of thumb: when picking the VNA frequency range you need, get a network analyzer with a maximum frequency of at least three to five times the maximum operating frequency of your DUT.
Why do we care about dynamic range?
Engineers refer to the power range over which they measure the DUT response as the dynamic range.
Figure 1 shows two different ways of looking at this spec. The system dynamic range is the value used for instrument specifications.
- System dynamic range indicates the instrument’s power capability without booster amplification or DUT gain. The instrument's maximum source power, or output power, is the maximum power level, Pref.
- Receiver dynamic range is the instrument's dynamic range with power amplification. Rather than using the source power as the maximum power level, this specification is based on the maximum power the instrument's receivers can measure, Pmax.
Figure 1: Visual comparison of system and receiver dynamic range.
Essentially, dynamic range tells your VNA’s output power and receiver noise floor capabilities in a signal spec. Why do you care? Well, imagine you were characterizing a filter. You want to see both the passband and stopband response on the same instrument. This means that the network analyzer receiver’s noise floor has to be low enough for you to reliably measure the stopband response where your filter’s output power is very low.
This concept also applies when you test high-gain devices like amplifiers. If you measure the S-parameters of a high gain amplifier, you want to use a low input power so that you best test the low input power the high gain amplifier is likely to deploy in (it also helps to not damage the VNA with the output power level). But this means that your S11 measurement will have low signal-to-noise ratio (SNR). In low SNR scenarios like this, it is helpful if you have a low receiver noise floor so that you still capture the low power response.
Think of it this way, for the same reason that you want a frequency range high enough to capture your DUT’s operation frequency and 5th-order harmonics, you also want a high dynamic range so that your VNA detects low power signals.
Why do we care about output power?
The output power indicates the maximum power your VNA can send into the DUT. It is expressed in dBm and references a 50-ohm impedance to match the characteristic impedance of most RF transmission lines.
Engineers value high output power because it helps improve the measurement SNR. It also allows engineers to determine the compression limit of their DUT.
Many active devices, such as amplifiers, require challenging linear and nonlinear high-power measurements that exceed the power limits of network analyzers. In this case, you would connect a dedicated signal generator to the backplane of the VNA and supply the power level you need to your device.
Why do we care about trace noise?
Trace noise describes noise that you see superimposed on DUT responses because of random noise in the test system. Higher trace noise makes the signal look less smooth, or even jittery (Figure 2).
Figure 2: Trace noise before applying averaging.
Engineers mitigate trace noise either by increasing test power, which lowers the receiver’s bandwidth, or by sweep averaging (Figure 3). In the past, trace noise was more of a concern because network analyzer synthesizer schemes varied a lot, however, most modern VNAs produce very low trace noise.
Figure 3: Trace noise after applying averaging.
Why do we care about the number of ports?
Your VNA needs at least as many ports as your DUT. In the early days of network analysis, all measurements were focused on 2-port S-parameters. As the capabilities of VNAs expanded, the ability to test power splitters, mixers, differential devices, and more led to the evolution of the 4-port VNA. Today, many components include multiple functions integrated into a single component. The number of ports on these components continues to increase with device complexity. Engineers call devices with more than four ports multiport devices.
Figure 4: The number of component ports continues to increase with device complexity.
Usually, multiport measurements are used when:
- the DUT has more than one path through it
- when you want to measure more than one DUT at a time
- when you have some combination of the two above situations
For the multipath situation, think of a transmit-receive (T/R) module where you have both a transmit and a receive path. Or a transmitter that uses multiple different channels. Each channel has its own transmit path, each at different frequencies. Each path means another port to test.
Why do we care about point sweep time?
Sweep time communicates how fast the VNA conducts measurements. Bottom line: the faster, the better. Production engineers really care about this spec as they have to test numerous DUTs as fast as they are manufactured. Time is money after all.
However, design and development engineers also care about sweep time. Think of it this way. Do you want to wait several minutes for a search result to resolve, or for a webpage to load? Every. Single. Time. No? Neither do RF developers every time they need to run a measurement.
- All testing has to happen upfront—once the satellite launches…pretty difficult to get it back.
- Not only is there a lot of testing, but all of it must happen before the launch, so engineers must fit every measurement in by the deadline.
- Since the satellite deploys in outer space, developers need the test environment kept under very special, highly controlled—extremely expensive—conditions.
As you can imagine considering these constraints, satellite developers consider test speed of critical importance. Slower tests waste valuable time and money.
Conclusion
Engineers assess a bunch of specs when picking out the perfect VNA. If you don’t believe me, check out our PNA-X network analyzer’s datasheet. But all you really need to get started in your search are these six key specs:
- frequency range
- dynamic range
- output power
- number of ports
- trace noise
- sweep time
There’s a lot to learn in the field of RF engineering. Sometimes it seems like too much. Instead of getting overwhelmed and throwing in the towel, check out Keysight Learn for free educational courses and boot camps. You can also check out the previous “VNA FAQ” posts, “An Introduction” and “What are Linear Devices.” Stay tuned for the next “VNA FAQ” blog!