The short answer: Any good spectrum analyzer measures signal strength vs. frequency. The real differentiator—especially with Rohde & Schwarz—is how it handles the edge cases that turn a routine measurement into a debugging nightmare.
I say this as someone who's been on the receiving end of decisions made from 'good enough' measurements. Over the past few years, I've reviewed roughly 200+ test reports annually for compliance. The ones that come back to bite us always share a pattern: the measurement looked fine on paper, but something was off in the real-world scenario. The analyzer's ability to handle dynamic range, phase noise, and interference—especially under less-than-ideal conditions—is where the £20,000 difference between a mid-range and a premium instrument like a Rohde & Schwarz FSW shows up.
For context: I work in quality for a mid-sized electronics manufacturer that supplies components to aerospace and telecom integrators. We're not a massive defense contractor, but we do have to meet their standards. The specs we're given are often tight: a -110 dBm noise floor, a 1 Hz RBW, or a specific phase noise at 10 kHz offset. Any analyzer can claim to meet these. But when you're trying to spot a -120 dBm spurious emission sitting 2 kHz away from a 0 dBm carrier, the instrument's residual phase noise floor and its local oscillator's purity become the deciding factors. A cheaper analyzer might show you a bulge in the noise floor that looks like a signal—or it might completely mask a real one.
The Sticker Price vs. The Real Cost of a Wrong Measurement
I've seen this firsthand. In a past project, the engineering team used a 'value' spectrum analyzer for EMC pre-compliance. The thing passed their internal checks. The product went to a certified test house for formal EMI certification. It failed. Broadly. The re-test and re-design cost us about £22,000 and delayed the launch by two months. When we dug into it, the pre-compliance analyzer was simply not sensitive enough. It showed the noise floor as being clean. The Rohde & Schwarz ESRP we later bought showed the actual switching harmonics from a power supply that were 15 dB higher than the cheaper unit indicated. The fundamental issue wasn't that the cheap analyzer was broken—it was that its noise floor and dynamic range were insufficient for the task. The datasheet had the numbers, but the real-world performance couldn't handle the signal's crest factor.
The lesson: an analyzer's 'Displayed Average Noise Level' (DANL) is a start. But the 'Third Order Intercept' (TOI) and 'Phase Noise' specs tell you how it will perform when you're mixing a large signal and a tiny spurious one. For the Rohde & Schwarz FSW, the phase noise spec at 10 kHz offset for a 1 GHz carrier is typically around -140 dBc/Hz (1 Hz) or better. On a budget unit, that might be -110 dBc/Hz. That 30 dB difference is the difference between seeing a spurious emission and not.
Personal Bias and a Specific 'Don't' for the 'Blood Pressure' and 'Phones' Keywords
Honestly, this is where the general advice stops and my specific context kicks in. If you're searching for a spectrum analyzer to test a phone's transmitter, the advice above is spot on. You need the dynamic range and phase noise to see modulation imperfections and spurious emissions. For a product like the R&S CMW500 (a tester for mobile devices), the critical spec isn't just the frequency range—it's the error vector magnitude (EVM) and the ability to generate and analyze complex 5G NR signals.
But something like 'blood pressure'? I have to be very careful here. I've seen engineers try to use a general-purpose spectrum analyzer to measure the radiated emissions from a medical device's wireless module. That's valid. But do not try to use a spectrum analyzer to interpret blood pressure data or diagnostic waveforms. It's designed for electromagnetic spectrum analysis, not biopotential signals. If you're working on medical electronics, the analyzer is for the wireless link and the switching power supply noise. The physiological measurement needs an oscilloscope with the right probes. I'm not 100% sure about the latest medical EMC standards, but I know we had to use a dedicated medical isolation transformer and a specific grounding setup for our EMC tests on a patient monitoring system.
The 'Connector' You Never Think About
This leads to a surprisingly practical point. Someone searching 'what is a connector' in a test & measurement context isn't asking about USB-C. They're asking about RF connectors like N-type, SMA, or 3.5 mm. The connector on the front of your Rohde & Schwarz analyzer isn't just a mechanical interface. It's a calibrated reference plane. The spec for the analyzer's input VSWR (voltage standing wave ratio) includes the connector. If you use a cheap, counterfeit SMA connector, you can introduce an impedance mismatch that adds 0.5 dB of uncertainty to your measurement. I've rejected a batch of 50 cables because the cheap SMA connectors were visually off-spec—the center pin diameter was 0.1 mm too small against our R&S standard spec. The vendor argued it was 'within industry tolerance,' but for a 40 GHz measurement, that's a massive error. We rejected them. Now every contract for cables includes the requirement for a 3.5 mm connector or a verified Amphenol / Rosenberger supplier.
So when you're reading the Rohde & Schwarz datasheet, pay attention to the input connector type and the VSWR spec. For an analyzer covering up to 26.5 GHz, an N-type connector is common. For higher frequencies, you're looking at 3.5 mm or 2.92 mm (K). A 'sloppy' connector is a source of error that no amount of calibration in the software can fix.
The Honest Caveat
Here's the part where I admit my own bias. My experience is heavily weighted toward EMI pre-compliance and high-frequency signal analysis for defense-adjacent applications. If you're mainly testing a 2.4 GHz WiFi module for output power, and you have a very stable budget, the law of diminishing returns kicks in hard. A Rohde & Schwarz FSH (a handheld analyzer that starts around £8,000-10,000) is an incredible instrument that will do 95% of what you need for general RF work. The £50,000+ FSW is overkill unless you genuinely need that phase noise floor or 1 Hz resolution bandwidth for close-in spurious measurements.
Take this with a grain of salt: I've heard from colleagues in the cellular base station installation field that the FSH is more than enough for antenna and cable sweep testing. The heavy-lifting lab analysis requires the bigger bench instruments. So your mileage may vary significantly depending on if you're doing R&D prototyping or production line pass/fail testing.
