CISPR Detector
Understanding CISPR Detectors
Electromagnetic interference signals are rarely continuous sine waves; they are typically pulsed, modulated, or broadband noise with varying amplitude, duty cycle, and repetition rate. Different detector types weight these signals differently, producing different numerical readings from the same physical emission. The choice of detector fundamentally affects whether a product passes or fails EMC compliance. A pulsed emission might read 60 dBμV on the peak detector but only 40 dBμV on the average detector, with the QP reading somewhere between.
The quasi-peak detector was originally designed in the 1930s to simulate the annoyance level of radio interference to human ears. Its asymmetric time constants (fast charge, slow discharge) mean that frequently repeated pulses produce readings close to the peak value (high annoyance), while infrequent pulses produce readings much lower than peak (low annoyance). For a continuous wave (CW) signal, QP, peak, and average all read the same. For a pulsed signal with 1% duty cycle and 100 Hz repetition rate, peak might read 20 dB higher than average, with QP about 10 dB above average. Modern EMI receivers implement all three detectors digitally using FFT-based processing, enabling simultaneous QP, peak, and average measurements in a single frequency sweep, dramatically reducing test time compared to analog detector implementations.
CISPR Detector Time Constants
QP charge: 1 ms ; QP discharge: 160 ms ; RBW: 9 kHz
Band C/D (30 MHz to 1 GHz):
QP charge: 1 ms ; QP discharge: 550 ms ; RBW: 120 kHz
QP vs Peak for Pulsed Signal:
VQP / Vpeak ≈ frep · τdischarge [for low duty cycle]
Where frep = pulse repetition frequency, τdischarge = QP discharge time constant. For CW: VQP = Vpeak = Vavg. Above 1 GHz: average detector with 1 MHz RBW replaces QP.
Detector Type Comparison
| Detector | Measures | CW Reading | Pulsed Reading | Primary Use |
|---|---|---|---|---|
| Peak | Maximum amplitude | = signal level | Highest | Pre-scan, MIL-STD-461 |
| Quasi-Peak | Rep-rate weighted | = signal level | Between peak & avg | CISPR emissions < 1 GHz |
| Average | Linear mean | = signal level | Lowest | CISPR > 1 GHz, conducted |
| RMS | Root-mean-square | = signal level | Between avg & QP | Military, some IEC |
Frequently Asked Questions
How does the quasi-peak detector work?
QP uses fast charge (1 ms) and slow discharge (160 to 550 ms). High-rep-rate signals stay near peak (barely discharge between pulses). Low-rep-rate signals discharge significantly, reading much lower. This weights signals by perceived annoyance: frequent interference is more annoying than occasional clicks. For CW, QP = peak = average.
Why do different bands use different time constants?
Band A (9 to 150 kHz): charge 45 ms, discharge 500 ms. Band B (150 kHz to 30 MHz): 1 ms / 160 ms. Band C/D (30 MHz to 1 GHz): 1 ms / 550 ms. Constants are tuned to the radio services being protected and the characteristics of typical interference sources in each band.
When is average used instead of QP?
Average is primary above 1 GHz (CISPR 32) and supplementary for conducted emissions. For CW, average = QP. For pulsed signals, the difference can be 10 to 20 dB. Some standards require both QP and average compliance. Above 1 GHz, average with 1 MHz RBW is standard because QP weighting is less meaningful at those frequencies.