Chinese
Spanish
Russian
Portuguese
French
German
Indonesian
Thai
Vietnamese
Arabic
Italian
Turkish
Polish
Japanese
An FMCW radar level meter measures distance using a frequency sweep and the reflected signal. Its specification may list operating frequency, range, accuracy, resolution and response time, but those terms describe different properties. Confusing them can produce a demanding purchase specification that does not actually test what the process needs.
Read the FMCW radar level meter specification as separate requirements
The transmitter sweeps its frequency, receives a delayed reflection and processes the difference between the signals to estimate distance. The radar measurement principle guide explains the subsequent distance-to-level calculation. To establish usable range and accuracy for an installation, review the manufacturer's stated conditions and the actual target.
For METRAVON MV-RD14, the published product identity establishes an 80 GHz liquid radar with a 44 mm lens antenna. Do not infer an unlisted sweep bandwidth from the 80 GHz designation, or substitute a radar-chip laboratory result for the complete instrument's specification.
| Item | Meaning | Ask the supplier |
|---|---|---|
| Band | The operating-frequency region, not the width of the frequency sweep. | Which antenna and beam information apply to the ordered version? |
| Bandwidth | The frequency span swept during the measurement waveform. | Is it specified, and what practical echo-separation performance is documented? |
| Accuracy | The stated closeness of the measured distance or level to a reference, under defined conditions. | What range, target, temperature and reference conditions support the value? |
| Repeat test | The agreement of repeated readings under the stated repeat conditions. | Were the target and settings unchanged, and what statistic is reported? |
| Response | How the reported value follows a real change, including processing and configured filtering. | Which settings and output path were used for the stated response? |
Target separation is not single-target accuracy
For an ideal FMCW range measurement, the familiar separation scale is approximately c / (2B), where c is propagation speed and B is sweep bandwidth. As a hypothetical example, a 1 GHz sweep gives about 0.15 m using c = 300,000,000 m/s. This describes a nominal ability to separate returns in range; it does not mean the distance error must be 0.15 m.
A processor can estimate the location of one suitable target more finely than the spacing at which two returns can be distinguished. Conversely, a fine displayed increment cannot guarantee that nearby echoes will be separated. Real echo strength, processing and installation affect the result. The example is not a bandwidth or performance claim for MV-RD14.
In a tank, this distinction matters when the desired surface return is close in distance to a nozzle or internal fitting. Ask whether the proposed installation can distinguish and track the liquid return through the required interval. Asking only for another decimal place on the display does not address that problem.
A stable reading can still have an offset
Repeatability describes how closely readings agree with one another under the stated conditions. It does not establish agreement with the correct reference. A consistently wrong sensor reference height can produce a very stable level error even while the underlying distance readings repeat well.
Keep three values distinct during commissioning: measured distance, calculated level and the value received by the PLC or data logger. Check the antenna reference plane and level offset before adjusting output scaling. If the local level is correct but the receiving system is wrong, investigate the signal conversion instead of compensating the sensor to hide the mismatch.
For analog outputs, the receiving input and scaling contribute to the displayed result. For digital outputs, verify the selected data item, units and invalid-value handling. More display digits, HART access or Modbus RTU communication do not independently establish better measurement accuracy.
Match filtering to the filling event
Filtering can make a fluctuating indication easier to use, but it can also delay the response to a genuine level change. A calm display during a fast fill is not necessarily a better control signal. Record the settings used during a response test and assess the entire path through the transmitter, receiving device and display or control logic.
Define the observation that matters: how promptly a real rise becomes visible, whether a brief disturbance creates a false excursion, and what happens during a lost echo. Test under representative operating conditions. Do not infer a suitable alarm response merely from the sensor's nominal update rate.
The radar acceptance test guide provides a practical way to record references and operating states. Agree tolerances and test conditions before ordering, rather than selecting them after seeing the most favourable results.
Questions about FMCW specifications
Q1. Does 80 GHz mean an 80 GHz sweep bandwidth?
No. The operating band and the sweep span are different quantities. Use the actual bandwidth if a documented analysis requires it. If the product information does not provide that value, request clarification rather than calculating resolution from the operating-band label.
Q2. Can a millimetre display prove millimetre accuracy?
No. A display increment is a formatting or output property. Verify the complete instrument against a suitable reference over the required interval and conditions. A reading can be shown to a millimetre while retaining a much larger offset or application error.
Q3. Should a specification request accuracy or repeatability?
It may need both, but for different reasons. Accuracy matters when the absolute level must agree with a reference; repeatability matters when repeated states must give consistent readings. State the operating conditions and test method for each instead of using the terms interchangeably.
Q4. Can a stronger filter solve a noisy measurement?
It can smooth reported variation, but it does not remove an obstruction, wrong echo or unstable mounting. Diagnose the source of the variation first. If filtering is appropriate, demonstrate that the resulting delay remains acceptable for the process.
Q5. What should an enquiry say instead of just "high accuracy"?
State the required level interval, acceptable error, independent reference, filling rate or relevant response requirement, and receiving output. Include the tank geometry and process conditions. Request an MV-RD14 application review with the proposed configuration and test conditions, so the response addresses an observable requirement rather than an undefined adjective.
Summary
Evaluate bandwidth, target separation, accuracy, repeatability and response as distinct properties. Use documented instrument specifications and a representative installation test to connect them to the process. Neither an operating-frequency label nor a finely divided display can replace that evidence.
