Ask a network engineer what single test instrument they wish had been in the truck two hours before a carrier deadline, and you'll get a range of answers. In my case, the answer is almost always the same: an Anritsu Site Master, not the laboratory vector network analyzer that sits back at our shop.
I'm a senior RF commissioning tech. For the last eight years I've coordinated the field-testing side of a telecom contracting company, and I've been the person who decides which analyzer goes to which job. That includes more than 400 site turn-ups and emergency restorals, a fair number of them with a same-day or next-morning deadline attached.
Through that work, I keep running into the same purchasing question:
Should we buy the Anritsu 37397C, the big benchtop vector network analyzer, or the Anritsu S331B, the handheld Site Master?
It sounds reasonable. Two Anritsu boxes, two model numbers, one question. But the question is built on a misunderstanding, and companies waste real money acting on it. So let's compare those two tools the way I wish someone had explained them to us years ago.
The Anritsu S331B is a Site Master cable and antenna analyzer. It was built for one broad job: proving an installed transmission line and antenna system is healthy. It sweeps from 25 MHz to 3300 MHz, runs on battery power, and is portable enough to carry up a tower without a second person to help. It gives a field tech return loss, VSWR, cable loss, and distance-to-fault. If an antenna feed system is degrading, the S331B tells you before it becomes a dropped-call complaint.
The Anritsu 37397C is a different animal. It's from Anritsu's Lightning-series vector network analyzers, and in many configurations it measures from 40 MHz up into the millimeter-wave range—65 GHz depending on the test set and options. It's an AC-powered laboratory instrument that performs full vector error-corrected measurements: two-port S-parameters, gain, phase, group delay, and everything else you need when you're characterizing a component, not verifying an installation.
You can already see the issue: the word analyzer does a lot of work in that purchase order. I've watched people compare them as though they were two versions of the same product. They're not.
Here's the thing: when a crew is standing at the base of a site with a cutover in six hours, they don't need more measurement capability. They need to know whether that antenna and feed line are good, and if not, where the fault is located. The S331B was designed for exactly that workflow. Set the frequency range, calibrate, sweep, read the DTF trace and the VSWR curve. It gives a clear answer.
A bench VNA like the 37397C can also measure return loss. What it brings is power and flexibility: the ability to measure insertion loss, phase, isolation, amplifier response, and complex impedance with high accuracy. That's essential in design labs, component incoming inspection, and certain hard-to-find RF faults. But none of those tasks happen at 2 a.m. on a tower with wind blowing. The 37397C doesn't make the field measurement faster or easier.
The verdict here is simple: if your company's recurring work is maintaining and commissioning sites, the S331B is the core tool. The 37397C is for engineers who need to see deeply inside components and systems from a bench. They are complementary tools, but they rarely compete for the same day on the schedule.
By now you'd expect me to say the lab VNA is more accurate, but the handheld is more portable. That's the conventional line. It also skips an important truth about measuring antennas.
When you measure an antenna from the ground, you're not measuring the antenna alone. You're measuring through every connector, jumper, and section of feed line between the instrument and the antenna. If your reference plane isn't where you think it is, a very expensive analyzer will happily produce a very precise measurement of the wrong thing. I've seen clean-looking plots from bench instruments that didn't mean what the engineer thought they meant, because the phase reference had shifted somewhere up the line.
The S331B doesn't solve every problem by magic, but its distance-to-fault function is built for this. You enter the cable's velocity factor, and the analyzer turns the sweep into a location trace that shows you where a reflection is happening. A bad connector 137 feet up the tower shows up as a peak at 137 feet. That's not a feature you have to configure deeply; it's part of how the tool thinks. With a benchtop VNA, you can do similar work only if it has time-domain options, and only if the operator knows how to apply gating correctly. In an emergency, that extra complexity is a liability.
So let's be honest: for installed cable-and-antenna work, the S331B isn't a stripped-down compromise. It is often the more trustworthy analyzer on the job.
Then there is the cost conversation. I don't compare two instruments by their sticker price. I compare them by what they actually cost our company over a year.
A benchtop VNA has costs beyond the purchase: calibration, cables, adapters, a proper cart or bench space, temperature-controlled environment, and the training required to get a valid measurement. A field analyzer has its own costs, too. But the biggest total cost driver is usually utilization. A tool that rides in a service truck and gets used daily is easy to justify. A tool that needs to be signed out and set up in a clean room is a different financial animal.
Here's a story, and I'll admit my own part in it. In March 2024, I told our buyer we needed a spare analyzer for a weekend cutover. I meant a field unit like the S331B. What the buyer heard was give the team the best vector network analyzer we can afford. The result: an Anritsu 37397C showed up in a crate while the crews were still driving to the site with no spare field analyzer.
I said analyzer. They heard full VNA. The mismatch cost us time and a chunk of budget that should have gone to a tool the trucks could actually use. Looking back, I should have written Anritsu S331B or equivalent handheld Site Master in the request. At the time, I thought the word analyzer was specific enough. It was not.
None of that makes the 37397C a bad instrument. It makes it a bad answer to a field-service question. When you evaluate RF tools, include the cost of having the wrong one at the job site. That single item has cost our company more than any calibration bill ever did.
If your work is tower work—installing, commissioning, sweeping, maintaining antenna systems—the Anritsu S331B or its newer Site Master successors should be near the top of the list before any benchtop VNA. It is the tool your crews will reach for weekly, and the one that pays for itself the fastest.
If your work is component design, high-frequency lab measurement, or anything that requires two-port S-parameters above a few gigahertz, then a 37397C or a modern equivalent is justified. That is the environment where its capabilities actually get used.
If you do both, the honest answer is that you'll eventually want both. But if you can only fund one purchase this quarter, look at your schedule, not at the spec sheet. The network analyzer that sits in protective foam because your crews don't live in a lab is not an asset. It's an expensive reminder of a wrong question.
One thing I'll credit Anritsu Company with: it has kept solid documentation and application support for its legacy products over the years, which matters when you're buying a used S331B or keeping an older 37397C in service. That continuity is one reason both model numbers still show up in real-world searches.
So which is better, the 37397C or the S331B? Neither—they do different work. The better question is which job you're actually being paid to do. The bench analyzer belongs in your lab. The Site Master belongs in your truck. Buy the tool that can get to the problem in time, because on an RF job, being anywhere but where the signal is means you're not measuring anything.