Glossary term
Spectral Mask
Engineering definition of spectral mask covering emission limits, out-of-band power, ACLR, measurement bandwidth, margin and validation evidence.
Definition
metricA spectral mask is a frequency-dependent limit that specifies the maximum permitted transmitted power or emission level at offsets from a carrier or inside defined measurement bands.
Spectral masks are used to verify that transmitters fit within regulatory, standard-specific or coexistence emission limits. A mask may constrain in-band shape, occupied bandwidth, adjacent-channel power, out-of-band emissions and spurious regions. It is not the same as ACLR, which reduces adjacent leakage to one or more ratios, and it is not the same as receiver adjacent-channel rejection. A useful mask statement must include frequency offsets, power reference, measurement bandwidth, detector settings, averaging, operating mode and uncertainty.
A spectral mask defines how much transmitted power is allowed at each frequency offset from an assigned channel or carrier. It is a compliance and coexistence tool: the transmitter may deliver the correct wanted power and still fail if its skirts, shoulders, spurs or spectral regrowth exceed the allowed envelope.
Spectral masks are used in wireless standards, radio licensing, telemetry systems, radar-like emitters, EMC reviews and production transmitter tests. They force the design to answer not only “how much power is transmitted?” but also “where does that power go in frequency?”
Mask Function
A mask can be treated as a limit function:
where f is frequency or offset from the carrier. The measured spectrum is:
The pass condition is:
for every frequency point or measurement band covered by the requirement.
Mask margin at a point can be written as:
Positive margin passes. Negative margin fails. The worst-case mask margin is:
Measurement Boundary
A spectral mask is only meaningful with its measurement boundary. Requirements may be stated in absolute power, relative carrier power, conducted power at an RF connector, radiated field strength, power spectral density or integrated power over a measurement bandwidth.
Resolution bandwidth matters:
Changing RBW, detector mode, averaging or sweep time can change the displayed trace. A valid result states those settings instead of treating a screenshot as self-explanatory.
Worked Example
A transmitter is tested against a conducted emission mask. At three frequency offsets, the mask limits and measured powers are:
| Offset | Mask limit | Measured power |
|---|---|---|
0.5 MHz | -24 dBm/100 kHz | -21 dBm/100 kHz |
1.0 MHz | -34 dBm/100 kHz | -36 dBm/100 kHz |
2.0 MHz | -45 dBm/100 kHz | -48 dBm/100 kHz |
The margins are:
The worst margin is:
The transmitter fails because the close-in shoulder exceeds the mask, even though wider offsets pass.
Corrected Screen
After amplifier backoff and output filtering, the measured powers become:
| Offset | Mask limit | Corrected power |
|---|---|---|
0.5 MHz | -24 dBm/100 kHz | -27 dBm/100 kHz |
1.0 MHz | -34 dBm/100 kHz | -39 dBm/100 kHz |
2.0 MHz | -45 dBm/100 kHz | -50 dBm/100 kHz |
The new margins are:
The mask now passes, but the backoff may reduce EIRP or coverage. A spectral fix must still preserve service margin.
Difference From ACLR
Adjacent-channel leakage ratio integrates power in one or more adjacent measurement channels and compares it with main-channel power. A spectral mask can be more detailed: it may define multiple offset regions, close-in shoulders, far-out emissions, spurious bands, absolute limits and detector settings. Passing ACLR does not automatically prove every point of a mask, and passing a few mask points does not replace the required ACLR calculation when a standard asks for it.
Spectral mask also differs from occupied bandwidth. Occupied bandwidth describes where a chosen percentage or defined part of signal power resides. A mask states what emissions are allowed at frequency offsets or bands.
Engineering Use
Engineers use spectral masks during transmitter design, power-amplifier backoff selection, digital predistortion release, filter selection, compliance testing, production screening, site coexistence review and failure investigation. Excess mask emissions may come from clipping, compression, modulator error, DAC images, phase noise, poor filtering, impedance mismatch, thermal drift, firmware misconfiguration or measurement setup error.
Corrective actions depend on the failing offset. Close-in shoulders often point to amplifier linearity, pulse shaping or predistortion. Far-out lines may point to clocks, mixers, spurs or shielding. A single “fail” result without offset detail is not enough to guide a fix.
Validation Evidence
A defensible spectral-mask record includes waveform, modulation, symbol rate, output power, duty cycle, center frequency, frequency offsets, limit table, RBW, VBW if used, detector mode, averaging, sweep time, reference level, attenuation, cable loss, calibration state, conducted or radiated boundary, uncertainty and environmental condition.
For field coexistence, the record should also say whether the measured transmitter configuration matches the installed antenna, EIRP, duty cycle and channel plan.
Common Mistakes
Common mistakes include using the wrong measurement bandwidth, comparing peak trace values with integrated limits, ignoring the worse side of the spectrum, testing at reduced power, leaving predistortion or gain state different from the field configuration, quoting EIRP without emission evidence, and assuming that bandwidth alone proves mask compliance.
The practical rule is to test the real transmit mode against the exact mask table, keep the measurement settings with the result, and check that any fix does not consume link margin or create a new coexistence problem.