RF safety for radar and high-power transmitters
Professional guide to pulsed and high-power RF sources: peak vs average mistakes, time-averaging, near-field complications, and practical exclusion-zone controls.
Offline pages
Saved on this device only. Connect at least once every 30 days to confirm Premium. Sign-out or a change of account removes saved copies.
On this page
You stopped at:
ContinueA pulsed source has at least two time questions: what happens while a pulse is present, and how much of the assessment interval contains those pulses. A scanning beam adds another question about how long it illuminates one location. Keep those time scales separate before interpreting a large peak or a small average.
Use time averaging, field-region selection and instrument response. This lesson models specified time records; the adopted framework still determines every required average, brief-exposure criterion and method.
Describe the operating state and accessible location
Record pulse shape, width, repetition rate, scan behaviour, antenna pattern and credible operating modes. Include stationary-beam or maintenance states when they are possible. High-power broadcast sources may be continuous rather than strongly pulsed; high output power by itself does not imply a large peak-to-average ratio.
Identify occupied routes, ladders, platforms and touch points near conductive structures. Large apertures and reflecting structures can invalidate a simple plane-wave conversion, while contact-current hazards require their own assessment. Separate an environmental field measurement from a permission to enter an antenna-adjacent work position.
Derive each duty factor from its time record
For identical rectangular pulses of duration τ and repetition rate fₚ with no overlap, pulse duty is dₚ = τ fₚ. If a power-like quantity is Sₒₙ during each pulse and zero between pulses, its complete-period average during illumination is dₚSₒₙ.
A scan factor dₛ can be included only if it describes the fraction of the assessment time that the same point is illuminated, the pulse pattern within that illumination is known, and any off-beam contribution is accounted for. In the simplified zero-off-beam case, S bar = dₛdₚSₒₙ. Do not apply a scan factor again to a measured duty or average that already includes scanning. For an irregular record, integrate S(t) directly over the required window.
Guided practice 1
Track pulses within a scanning beam
Problem
At a fixed far-field-like point, each rectangular radar pulse gives incident power density 600 W/m² for 2.00 µs. Pulses repeat at 1500 s⁻¹ while the point is illuminated. The beam illuminates it for 4.00 s in each 20.0 s scan, with zero contribution outside that interval. Consider a complete 100 s record containing five scans, and neglect edge effects at the pulse scale.
Find the pulse-averaged power density during illumination, the complete-record average, and the incident energy density during one 4.00 s illumination. Explain why the complete-record average does not establish compliance with every shorter-window or brief-exposure criterion.
Make the decisions, then compare your working
Hints
Hint 1
Find pulse duty first
Use τ fₚ, with pulse width converted from µs to s. That duty describes the pulse train while the beam is present.
Hint 2
Then account for scan occupancy
The same point is illuminated for 4.00/20.0 of complete scans. A window lying inside an illumination has a different average.
Show solution step by step
Average pulses during illumination
Method
Multiply the pulse on level by the pulse duty.
Reason
The constant pulse amplitude and complete-period assumptions justify the multiplication.
Working
Answer
The pulse width is 2.00 × 10⁻⁶ s. Multiplying by 1500 s⁻¹ gives dₚ = 0.00300. During illumination, multiply 600 W/m² by this duty to obtain a pulse-period average of 1.80 W/m².
Average the complete scan record
Method
Weight the illuminated density by the illuminated-time fraction.
Reason
Five identical complete scans give 20.0 s of illumination in 100 s.
Working
Answer
dₛ = 4.00/20.0 = 0.200, so S bar = 1.80(0.200) = 0.360 W/m². The full-record incident energy density is 36.0 J/m².
Keep the shorter interval in view
Method
Integrate over a single illumination and identify its comparison scope.
Reason
An average over the full record cannot replace every shorter exposure condition.
Working
Answer
The incident energy density in 4.00 s is 1.80(4.00) = 7.20 J/m². A shorter window spanning many pulse periods within an illumination has an average near 1.80 W/m², rather than 0.360 W/m². A window within a single pulse can have a much higher average. Consult the required window placement and brief-exposure criteria; no exposure limit was supplied here. Compare that reasoning yourself; the checker verifies the three requested quantities only.
Change the scan state before reusing the result
Check your understanding 1 A service mode changes the time record
In the same system, a service mode stops beam rotation and leaves the assessed point continuously illuminated for 100 s. Pulse width, repetition rate and on-pulse density remain unchanged. The access permit still quotes the previous 0.360 W/m² complete-scan average.
Evaluate the quoted result for the new state and identify the control and measurement decisions that need revision.
Show answer
Scan duty is now one, so the 100 s average is 1.80 W/m², five times the earlier result. Incident energy density is 1.80(100) = 180 J/m². Pulse duty remains 0.00300; rotation stopping does not turn the pulse train into a continuous 600 W/m² signal.
The original result describes the scanning state and cannot authorise access under the stationary-beam state. Reassess the relevant accessible locations, required averages and brief criteria, and verify the actual inhibit, reduction or shutdown conditions required for the task. Retain an enforceable boundary and stop work if the verified state changes. Compare your answer for the changed scan duty, unchanged pulse duty and the operating-state dependency of the controls.
Protect the measurement before interpreting its average
A probe may display a low average yet overload or respond incorrectly during the pulses. Verify its bandwidth, pulse response, linear range, damage rating, detector and any averaging already applied. Peak, peak-hold, RMS and average indications need their instrument definitions; a screenshot cannot reconstruct missing duration.
Search for candidate maxima using an appropriate method, then fix measurement height and stand-off and reduce operator perturbation. Preserve the time record and operating-state evidence. ICNIRP 2020 includes additional conditions for brief local exposure; a long-window average alone is not the whole assessment.
Tie access controls to verified conditions
Use barriers, guards and transmit-inhibit arrangements suited to the assessed geometry. Specify the state required before access, who verifies it, how it is maintained during the task and what triggers stopping. A request to reduce RF is different from evidence that the required reduction occurred. Include other active transmitters and changes in configuration.
Review peaks, averages and control evidence
Check your understanding 2
A source produces constant Sₒₙ = 100 W/m² during pulses and zero between pulses. The duty factor over complete periods is D = 0.02. Estimate the time-averaged S bar.
Show answer
S bar ≈ D Sₒₙ = 0.02 × 100 = 2 W/m².
Check your understanding 3
Why peak screenshots mislead. If a peak-hold display captures a rare alignment event that lasts a fraction of the averaging window, what extra information do you need to compare to a time-averaged criterion?
Show answer
You need the time behavior (duty/occupancy/scan pattern), the framework-defined averaging window (and whether sliding-window max applies), and a method to estimate the worst-case averaged value over that window.
Check your understanding 4
Why is a peak-hold number often not directly comparable to a guideline value?
Show answer
Because many guideline values are defined with time averaging (and sometimes sliding-window maxima). Peak-hold captures a maximum event, not the defined averaged quantity.
Check your understanding 5
What does “defensible” require you to record for a pulsed source measurement?
Show answer
Detector mode, bandwidth/integration method, averaging window/time constant, operating state assumptions (scan/duty), geometry (height/stand-off), and overload headroom.
Check your understanding 6
Propose two controls for antenna-adjacent maintenance at a high-power RF site and explain their purpose.
Show answer
One good pair is a physical access boundary marked against the assessed geometry, to prevent entry into unassessed exposure locations, and a task permit requiring verified RF reduction or shutdown before antenna-adjacent work. Trained workers need stop-work authority if the verified state or boundary changes.
Keep the time scales and the access state visible
Write pulse duty, scan occupancy, window placement and any off-state contribution separately. Match each required criterion to its own quantity and method. When the credible operating state changes, reconsider the assessment and the conditions supporting access.
Spotted an error? Report a correction
Syllabus and review details
No official syllabus alignment is listed for this lesson.
Last reviewed: