As of 2026-06-21, a DirectIndustry industrial-manufacturer index lists 34 vendors and 69 distinct ultrasonic level transmitter models still in active catalog [S1], while Chinese OEM catalogs such as Shanghai GN Instruments continue to release paired 26 GHz radar and small-blind-area ultrasonic lines side-by-side [S4] — proof that both technologies are still in parallel production for liquids and bulk solids.
Process engineers typically narrow the choice using five field variables: dielectric constant of the medium, presence of foam, vapor or dust, process temperature and pressure, required CIP/SIP compatibility, and the physical stand-off (nozzle height / blocking distance). On those axes, the two non-contact families — radar level meters and ultrasonic level meters — overlap heavily only on clean, atmospheric liquids; everything else (steam, vacuum, agitated slurries, hot reactors) trends to radar, while the lowest-cost atmospheric-water and open-channel duty cycles still trend to ultrasonic.
Operating principle and frequency bands in 2026 catalogs
Ultrasonic level meters drive a piezo transducer at 20–200 kHz, time-of-flight the acoustic echo through air, and convert distance to level using a temperature-compensated sound velocity. They cannot operate in a vacuum, and accuracy degrades when the air column is heated unevenly, saturated with vapor, or loaded with dust [S3].
Radar level meters emit FMCW or pulsed microwave energy in two commercial bands: 6 GHz (C-band, long legacy range up to 35 m on solids) and 26 GHz (the dominant 2026 catalog frequency per CX-RLM-080 and similar 26 GHz radar lines) [S4]. Newer 80 GHz FMCW devices are now common for small nozzles and turbulent surfaces, but the 26 GHz family still anchors most price lists. For guided-wave / TDR variants, see guided wave radar and TDR level meter pages for the contact-probe branch of the same technology tree.
Decision criteria: dielectric, vapor, foam, dust, temperature, blind zone
Microwaves at 26–80 GHz couple strongly into media with dielectric constant εr ≥ 1.8–2.0, so radar works on hydrocarbons and most organic liquids; ultrasonic sound reflects from the gas/liquid interface regardless of εr, but is killed by turbulence, foam layers thicker than a few cm, heavy vapor, and any condensing mist on the transducer face [S3].
On the radar side, heavy foam still attenuates 26 GHz, and very low-εr media (LPG, LNG at εr ≈ 1.5) demand 80 GHz or guided-wave probes — see radar level meter for the frequency trade-off. On the ultrasonic side, the so-called "blind zone" or blocking distance (typically 0.2–0.6 m on compact CX-ULM-T and CX-ULM-TII small-blind-area models) sets a hard lower measurement limit, so the nozzle must sit above that distance from the highest expected level [S4]. For tanks where the level can reach the flange, this is a non-negotiable reason to step up to radar.
Comparison: radar vs ultrasonic across four engineering axes
Across the four axes that drive most spec rejections, the 2026 catalog position is qualitatively clear: [S1]
• Beam angle / nozzle fit: 26 GHz radar focuses to ~8° in a 100 mm stilling well, 80 GHz to ~3–4°; ultrasonic transducers are physically large (often 40–80 mm dia) and diverge at 6–10° at 1 m range, so narrow nozzles and long stand-offs favor radar.<br/>• Vapor / dust tolerance: radar microwaves pass through steam, mild dust, and most non-conductive foam; ultrasonic echoes collapse in saturated vapor, condensing atmospheres, or foaming reactors [S3].<br/>• Process temperature: standard ultrasonic sensors are typically limited to ~80 °C at the transducer face; radar can sit behind a high-temperature lens up to 450 °C in standard process builds.<br/>• Cost & commissioning: ultrasonic transmitters remain the lowest-cost non-contact option for open channel, wastewater, and clean atmospheric water — for example, 10–99-piece ultrasonic level sensors cataloged at US$399/100+ pieces, undercutting most radar offerings [S7]. Where the ultrasonic level meter fits the duty cycle, it is the budget default.
Who the ultrasonic path is FOR — and who it is NOT for
Ultrasonic fits four duty cycles that the 2026 market still lists heavily: open-channel flow measurement in flumes/weirs, atmospheric wastewater sumps and lift stations, small chemical tanks at ambient conditions, and bulk-solid silos with low dust (e.g. plastic pellet bins) [S1][S3]. For the deeper, sound level meter instrument family used in environmental noise, the working principle is the same time-of-flight acoustic stack.
Do not specify ultrasonic for: vacuum vessels, high-temperature reactors, steam-saturated headspace, agitated or foaming process liquids, media where condensate can form on the transducer face, or any tank where the highest credible level can climb within the transducer's blind zone [S3][S4]. On all of these, radar — or, in conductive liquids, guided wave radar — will perform; ultrasonic will read erratically or fail to read at all.
Who the radar path is FOR — and what it costs
Radar fits the bulk of 2026 process work: hydrocarbon storage, chemical reactors with vapor, high-temperature/pressure separators, agitated process vessels, hygienic CIP/SIP tanks in food & pharma (where the smooth, crevice-free antenna matters), and most bulk-solid silos with dust. 26 GHz is the workhorse; 80 GHz FMCW is the modern pick for narrow nozzles and small tanks; C-band (6 GHz) remains the cheapest radar option for tall solids bins. [S2]
Cost discipline matters: a standard 26 GHz radar still commands a 3–8× price premium over a comparable ultrasonic transmitter in Chinese OEM catalogs (radar lines such as CX-RLM-070/080 sit alongside CX-ULM series ultrasonic lines from the same supplier, allowing direct price comparison) [S4][S8]. For greenfield work where neither vapor nor foam is present, the lowest-cost option that meets the spec is still the right option — radar is not automatically "better."
Limits, failure modes, and standards to keep in mind
Ultrasonic failures cluster around transducer fouling, condensate bridging, and air-temperature gradients. Radar failures cluster around low-εr media, heavy buildup on the horn antenna, and foam thicker than a few cm at 26 GHz. Both families need an unobstructed launch zone, and both can be defeated by internal agitator blades or cage ladders unless those are mapped in the configuration software. [S3]
For hazardous-area installs, both technologies ship in Ex variants (CX-ULM-T "ex" small-blind-area ultrasonic and the matching radar line are listed with CE / ISO 9001:2008 process-quality certification in the same 2026 catalog) [S4]. Hygienic / 3-A and EHEDG radar antennas are available from major brands; ultrasonic hygienic variants are rarer and almost always need a separate PTFE-faced transducer. Where redundant level is required for safety function, combining a non-contact radar with a secondary automatic level switch family — tuning fork, RF admittance, or float — is now standard practice in 2026 vendor literature.
Field checklist before locking the spec
Before stamping the purchase order, force the supplier to answer six questions: (1) what is the dielectric constant of the medium at operating temperature; (2) is there a vapor, foam, or dust layer at the maximum credible process condition; (3) what is the highest temperature at the instrument flange; (4) what is the nozzle ID and stand-off, and does the chosen beam angle clear internal obstructions; (5) what is the clean-in-place regime (chemistry, temperature, spray-ball coverage); (6) is Ex rating required, and which zone/class. With those answers on paper, the radar-vs-ultrasonic call is mechanical: if any of (1)–(5) push toward radar, do not try to "save cost" with ultrasonic; if all six are clean, ultrasonic is the right pick. [S4]
For adjacent selection problems on the same 2026 spec bench — pressure transmitters vs gauges, pH meters vs online analyzers, or thermal mass flowmeters — see the in-house comparison articles: Pressure Gauge vs Differential Pressure Transmitter: 2026 Selection Criteria and pH Meter vs Online pH Analyzer: Range, Output, Selection. The next trackable signal in this segment is the 80 GHz FMCW price line: as more Chinese OEMs release sub-$1,000 80 GHz units, the 26 GHz "comfortable default" will be squeezed from below.