Download slide presentation here.

Overview:
Measurement accuracy is a critical requirement for EMC Compliance testing. This webcast will discuss tools that can be used to overcome frequency and amplitude measurement errors associated with measuring emissions over broad frequency ranges. It opens with a discussion of the measurement errors associated with making prescan measurements used to identify suspect emissions prior to final measurements. These errors are due to the characteristics of both the emissions and the measuring receivers. It then discusses in detail two powerful tools that help EMC professionals improve the accuracy and quality of their final compliance measurements.

Who Should Attend?
Engineers or technicians involved in EMC precompliance or full compliance conducted and radiated emissions testing to both commercial and MIL standards and design engineers who need to understand the conducted and radiated emissions performance of the device they are designing.

Speaker: Mark Terrien
Mark Terrien has over 20 years of product development experience with Hewlett Packard and Keysight (formerly Agilent) Technologies, with a focus on EMC receivers, spectrum analyzers and microwave test equipment. He holds an MBA from Golden Gate University in San Francisco and an MSEE in Electromagnetic Wave Theory from the University of Wisconsin-Madison.

The following are questions presented to the speaker by the attendees during the webinar, along with answers to each.

“Hello Mark. The prescanning activity can be used as a justification for the suspect frequecy selection, and for the decision of performing a reduced set of measurements for compliance? Thank you.
Answer: Hello – thank you for your question.  Yes, the prescanning activity with a peak detector can be used to identify signals for final measurement.  Peak detector values will always be greater than or equal to Quasi-Peak, EMI-Average or RMS-Average detector values.  For precompliance work, please ensure to have sufficient margin to account for the uncertainties associated with your precompliance test site.  This margin will help to minimize missing a signal due to site contributions to the measurement.

Is it possible to find the highest peak with broader bandwidth than 120 kHz?
Answer: Thank you for your question. It is recommended to scan with the required CISPR bandwidth, but not mandatory for pre-scanning. If you prescan with a broader resolution bandwidth and measure a broadband signal, the signal you will measure will be larger than the signal measured with the 120kHz bandwidth. Narrower bandwidths will report a lower amplitude (assuming that the signal is still broadband relative to the narrower bandwidth). The reported amplitude for narrowband signals will not change with bandwidth.

Can you comment about the prescan analysis tools and real time spectrum analyzer? Thanks.
Answer: Thank you for your question. The prescan analysis tools mentioned in my talk use traditional frequency scan, Time domain scan or FFT technologies. Scans can be made with peak detectors or traditional weighted CISPR detectors. While peak detector scans can be fast, it can still be difficult to fully capture very fast broadband transients. Real Time analysis offers the opportunity to capture very fast signals and to see signals buried underneath other signals (using a density display), but RTSA measurements typically are constrained in the number of resolution bandwidths available (not CISPR). RTSA would be a good diagnostic tool to use in concert with prescan analysis.

When using for prescan, why not just add signals within 6dB of limit line line to the suspect list? That is just accept error and home in later?
Answer: Thank you for your question.  I highly recommend adding a margin to all precompliance prescanning because the scanning is typically not done in a CISPR or MIL-qualified facility. However, CISPR does require that you make your measurement at the frequency of the maximum emission.  If you don’t do the analysis prior to final measurement, you could easily miss the maximum emission.  If I understand your comment correctly, you could do the analysis later (prior to final measurement), as long as the analysis is done to ensure that you are making your final measurement at the correct frequency.

What methods are used to determine signal level accuracy probability using the step method?
Answer: Thank you for your question.  When stepping during prescan, the accuracy of the peak measurement is a function of the alignment of the narrowband signal relative to the center of the resolution bandwidth, and the modulation characteristics of the signal in question.  You can minimize the amplitude errors by using 4 measurement points per resolution bandwidth, but this adds measurement time.  Regardless, to meet commercial compliance requirements you do need to check at least the QP value of each signal to ensure that you’ve made a measurement at the correct frequency and to ensure that the QP value doesn’t change with time.