Every spectrum analyser datasheet has a dynamic range spec. Most engineers glance at it, nod, and move on to the frequency range and DANL figures instead. That’s a mistake — especially once you push measurements into the millimetre-wave region. At mmWave frequencies, dynamic range stops being a background number and becomes the single figure that decides whether your measurement is trustworthy or misleading. Dynamic Range is a Trade-Off Dynamic range is usually defined as the span, in dB, between the noise floor and the highest signal level the analyser can measure without distortion. But that simple definition hides three different limits fighting each other: Noise floor (DANL): how small a signal you can see at all. Compression point: how large a signal you can measure before the front end starts to distort it. Spurious-free dynamic range (SFDR): how far below the fundamental you can see a real spurious tone before it disappears under the analyser’s own internally generated mixing products. A wide dynamic range spec sounds like good news across the board, but in practice you’re rarely operating across the whole range at once. You’re usually asking one specific question — “can I see this weak harmonic sitting 80 dB below my carrier?” — and the answer depends on where in that range you’re working, not the headline number alone.   Why Your Spectrum Analyser Struggles So Much Harder Above 26 GHz Below about 26 GHz, most spectrum and signal analysers work natively — the mixer, IF chain and detector are all designed and calibrated for that range, and the manufacturer’s dynamic range spec applies more or less as printed. Above that, almost nobody has a native front end. To reach the E, W, or D-bands, the signal has to be down-converted first, typically with an external harmonic mixer or a frequency extension module, before the analyser’s own IF chain ever sees it. That down-conversion step is where dynamic range quietly erodes: Conversion loss at the mixer reduces the effective signal level reaching the analyser, which pushes the usable noise floor upward. High-order harmonic mixing — common in low-cost mixer solutions that multiply the LO many times to reach mmWave frequencies — generates its own set of internal spurious responses, which directly eats into SFDR. LO leakage and image response can appear as false signals that look like real ones on the trace, unless the extension hardware and the analyser’s own harmonic identification are both well matched. The result: a spectrum analyser with an excellent on-paper dynamic range at its native frequency can look dramatically worse once you’ve bolted an inexpensive harmonic mixer onto the front to reach 100 or 200 GHz. The instrument didn’t change — the front end did, and the front end is doing most of the work.   Reading the Trace, Not Just the Spec Sheet This is why an experienced test engineer looks at a mmWave trace differently to a microwave one. A raised noise floor, a cluster of unexplained tones at regular frequency intervals, or a signal that seems to compress earlier than expected are rarely faults in the device under test — they’re usually dynamic range limitations in the measurement chain itself, surfacing as artefacts on the display. Before trusting a low-level mmWave measurement, it’s worth asking: What mixing scheme is generating the down-conversion — fundamental, sub-harmonic, or high-order harmonic? What’s the actual conversion loss at the frequency being measured, not just at the extender’s centre frequency? Has the noise floor been checked and, ideally, calibrated out for this specific configuration?   Getting Dynamic Range Back This is exactly the problem Farran Technology’s frequency extension systems are engineered to solve. Rather than relying on high-order harmonic mixers that trade simplicity for spurious content, Farran’s mmWave extension technology is built around low-conversion-loss mixing to preserve as much of the host analyser’s native sensitivity as possible as frequency increases. Farran’s SAE modules interface with an existing microwave signal or spectrum analyser and extend its frequency coverage up to 170 GHz, using proprietary mixer technology to achieve low conversion loss and noise levels for exceptional measurement sensitivity. Depending on configuration, these systems are capable of dynamic range performance from 100 dB up to 135 dB typical, with magnitude trace stability as tight as ±0.1 dB. That’s the difference between a mmWave measurement you can act on and one you have to second-guess. The same principle carries through Farran’s wider extension portfolio, including its FEV frequency extension heads for vector network analysers, which extend S-parameter measurement into the millimetre-wave domain up to 500 GHz while leveraging the host VNA’s native performance.     The Takeaway Dynamic range at mmWave isn’t a spec you check once and forget — it’s a live property of your entire measurement chain, and it moves with every mixer, cable, and extension module you add in front of the analyser. Understanding what’s really limiting it, and choosing extension hardware engineered to protect it, is what separates a mmWave measurement you can trust from one that just looks like data.     Looking to extend your spectrum or signal analyser into the mmWave region without sacrificing sensitivity? Explore Farran’s SAE Frequency Extension Systems or get in touch with our engineering team to discuss your application.