Troubleshooting of R&S EPL1000

Time domain scanning configuration and performance failures

2.1 Spectrum leakage or low amplitude in time-domain scanning results

Fault phenomenon:

When using FFT based time-domain scanning to measure CISPR Band B (150kHz~30MHz), compared with traditional stepped scan results, the amplitude of certain frequency points is more than 2dB lower or non-existent false signals appear.

Principle Review:

The time-domain scanning of EPL1000 adopts parallel FFT processing to collect the entire CISPR frequency band at once, fully meeting the requirements of CISPR 16-1-1 for measurement time and detector response. However, FFT processing is sensitive to the signal-to-noise ratio and aliasing effects of the input signal. If the pre selection filter is not correctly enabled, it may lead to spectral leakage.

Solution:

Check the status of the pre selection filter: EPL1000 has built-in pre selection filters for CISPR Band A (9kHz~150kHz) and Band B (150kHz~30MHz). Under the 'FREQ' menu, confirm that 'Preselection' is set to 'Auto' or 'Band B'. If set to 'Off', strong signals can cause ADC saturation and harmonic distortion.

Adjust attenuator settings: Press the [AMPT] key to set the RF attenuation to "Auto". The pulse protection input of EPL1000 can withstand high instantaneous power, but excessive attenuation will reduce sensitivity. The automatic attenuation function optimizes the dynamic range based on the signal level.

Check overload indicator: If a red warning of "OVLD" appears in the status bar at the top of the screen, it indicates that the input signal exceeds the linear range of the mixer. At this point, it is necessary to increase the attenuation (manually set to 10dB or higher), or enable a 20dB preamplifier (only used in weak signals).

Measurement time setting: The measurement time for time-domain scanning should be at least 1 second to ensure that the CISPR quasi peak detector reaches steady state. If the setting is too short (such as 10ms), the quasi peak response will be insufficient and the amplitude will be lower. Recommended settings: Band B measurement time ≥ 1s, Band A measurement time ≥ 5s.

Verification method:

Comparative test: Under the same signal source (such as a comb wave generator), time-domain scanning and step scanning measurements should be used separately, and the deviation between the two should be within the ± 1.5dB range allowed by CISPR 16-1-1. If exceeded, contact the R&S service center for calibration.

2.2 Intermittent interference cannot be captured

Fault phenomenon:

There is occasional transient interference in the device (such as pulses generated by relay jumps), but multiple repeated scans have not shown this interference, resulting in missed reports in the pre compatibility test.

Reason:

The default Preview Scan uses a peak detector for fast scanning, but the scanning time is short (such as 100ms), which may miss transient events with low probability of occurrence.

Solution - Long term gapless measurement:

The time-domain scanning of EPL1000 supports a continuous measurement time of up to 100 seconds. The operation steps are as follows:

Enter the "Scan Table" menu and define the frequency range to be monitored (e.g. 150kHz to 30MHz).

Set the 'Measurement Time' to 10 seconds or higher (maximum 100 seconds). Note: The longer the measurement time, the larger the amount of data in a single scan, but it can ensure the capture of all transient signals.

Enable the 'Max Hold' tracking feature to overlay the maximum hold curve after multiple scans.

If you need to accurately locate the time point where interference occurs, use the Real time Spectrogram function - this mode ensures seamless acquisition of the entire frequency band and displays amplitude changes over time in color coded format.

case

When testing the laptop power adapter, it was found that some frequency points occasionally exceeded the CISPR 22 Class B limit. After setting the measurement time to 10 seconds and repeating the scan 5 times, the maximum retention curve clearly shows that the peak appears at 22.3 MHz. Further use of real-time spectrograms identified interference occurring every 2 seconds for approximately 50ms, ultimately confirming it as switch noise from the power management IC.

Troubleshooting of detector and limit line faults

3.1 Significant deviation between peak measurement results and certified laboratory

Fault phenomenon:

Under the same device and LISN, the quasi peak value (QP) measured by EPL1000 is more than 3dB lower than that of a third-party certified laboratory, but the peak detector results are consistent.

Root cause:

The charging and discharging time constants of the quasi peak detector strictly follow CISPR 16-1-1. If the measurement time (dwell time per point) of EPL1000 is insufficient, or the pulse repetition frequency of the measured signal is extremely low, the QP detector cannot reach a fully charged state.

Standard operation:

Confirm measurement time: For QP detectors, CISPR requires that the minimum measurement time for each frequency point depends on the intermediate frequency bandwidth of the receiver. For a bandwidth of 9kHz (Band B), the minimum dwell time is 1s; for a bandwidth of 200Hz (Band A), the minimum is 5s. In EPL1000, enter the "Detector" setting, set "Meas Time" to "Auto" or manually enter the above value.