Automated EMC measurement performed with suitable test and controlmethods can improve efficiency in the laboratory and eliminate the timespent in making repeated EMI measurements.
An integrated test system helps prevent errors caused by manuallyreading and recording results, and the software ensures the conformanceof the instrument settings, enabling the production of highlyrepeatable test results.
There is no difference between the accuracy of emission valuesmeasured manually and measured under software control. In both cases,the measurement uncertainty is based on the accuracy specifications ofthe equipment used for the test.
However, differences in the settings of instruments controlled bysoftware may lead to varying test results. This is solved by usingautomated EMI measurement standards,specifically the IEC CISPR16-2 standard. There are also correctionfactors and automatic monitoring functions built into test systems toensure the validity of the test results.
Another advantage of automated EMC measurement is the ease andefficiency of test result and data management that the softwareprovides. Using the software, it is very easy to establish a databaseand save and review test results.
But these advantages also have accompanying challenges that can leadto wrong results including problems during frequency sweep, measuredtime setting, key data reduction and choice, height of antenna,calibration and correction.
This article describes how to establish an automated EMI testsystem, with an emphasis on problems encountered, solutions used andhow to configure the system and parameters correctly. It also stressesthe importance of normalization of the system in order to avoiddifferences between test results from the laboratory and fromoperations.
|Figure1. If step size is equal to IF bandwidth, test results of somefrequency points will be 6dB less than correct values (a) and if stepsize is equal to 1/2 IF bandwidth, the error will be less than 1dB (b).|
The following section discusses two scanning rules for EMI measurement:
1) Frequency scan stepsize. During EMI measurement, there is a concern that somefrequency points may be ignored. In receiver mode, the test frequencyis sweeping point by point in step mode.
It is important to know how to set the step size to ensure thatpoints are not ignored. In spectrum analyzer mode, the test frequencyis sweeping continuously, but the number of points in the displayedspectrum is limited. In this mode, knowing how to locate the realfrequency of the peak value is the most important.
In receiver mode, how should engineers set the step size? As a rule,the step size should be less than half of the IF bandwidth; but usuallyhalf of IF bandwidth is used. EMC standards define the IF bandwidth ofthe receiver as -6dB bandwidth. There will be some peak values lost orthere will be wrong results if an unsuitable step size is used.
For example, if the step size is equal to IF bandwidth, the testresults of some frequency points (in the middle of two adjacent testpoints) will be 6dB less than the correct values. If the step size isless than half the IF bandwidth, the error will be less than 1dB(Figure 1 above).
According to EMC standards, we can use the spectrum analyzer with-6dB IF bandwidth and preselector. In analyzer mode, the number ofdisplayed points will influence the test results. To know the exactfrequency value on the result curve, a partial scan may be necessary(zoom in the frequency axis). During partial scan, the setting shouldbe: SPAN/(number of points) <0.5 RBW.
2) Measurement time and scan rate. Measurement times and scan ratesof measuring and scanning receivers shall be set to measure the maximumemission as listed in Table 1 below .
|Table1. Minimum scan times for CISPR ranges.|
Dwell time for each frequency point is calculated using the formula:
The minimum scan and dwell time is usually used during prescan. Forthe final measurement at peak points, the dwell time shall be longenough at each frequency to measure the signal peak.
The generic procedure for an automated EMI test, with the correspondingsoftware configuration, is outlined in Figure2 below.
|Figure2. The generic procedure of an automated EMI test and the correspondingsoftware configuration are shown.|
The first step in automated EMI measurement, as it is in a manualtest, is the pre-scan: perform a sweep in the target frequency range tolocate any emissions from the equipment under test (EUT).
According to the EMC standard, the critical limits are given byquasi-peak detector, but the use of the quasi-peak detector during thetest in the whole range of interest leads to excessive test times.
The frequency points of interest should be limited to thefrequencies at which the measured peak amplitude of the emission isabove or near the emission limit – only the emissions at these pointswill be maximized and measured. For the pre-scan step, note thefollowing:
1) Conducted emissions. Pre-scan may be performed on a representative lead, for example lead”L” of the power line, using peak and average detection and the fastestscan time possible. Two limits, for quasi-peak and average detector,may be called out. Disturbance power—Pre-scan may be performed with theabsorbing clamp close to the EUT. Peak and average detection with theirlimits should be used.
2) Radiated emissions. In the frequency range from 9kHz to 30MHz, both the loop antenna andthe EUT need to be rotated to find the maximum field strength while thereceiver is scanning the emission spectrum. In the frequency range from30MHz to 1,000MHz, the antenna height may be preset to fixed heightsgiven in Table 2 below.
|Table2. Recommended antenna heights for pre-scan.|
Results from the pre-scan are put through a data reduction step.This is important to decrease the number of signals collected duringpre-scan and to reduce overall measurement time. The data reductionmethod consists of an acceptance analysis and a sub-range maximasearch.
For the acceptance analysis, you may select a limit line for eachdetector which will be used for the level evaluation in the finalmeasurements. In addition to this, the acceptance offset has to bedefined. For sub-range maxima search you can define some frequencysub-range in the whole range, and search the peaks in each sub-range.
Prior to testing the points remaining after data reduction, adjustthe settings of the accessories (including antenna, turntable, LineImpedance Stabilization Network and absorbing clamp) to catch maximumemissions.
Then test the points with the defined detectors according tostandards. The measurement time must be long enough at each frequencyto measure the signal peak. One advantage of automated measurement isthe automatic correction of test values. Calibration work for everysignal path and accessory is necessary. Test results should be givenwith the calibration data.
Also, obtaining a test report is often the objective of the test.The results of the automated EMI test will be presented in tables andgraphics in the test report. Moreover, information about the testsystem itself – including transducers and correction used, instrumentsconfiguration, and documentation of the EUT setup as required by theproduct standard – will also be part of the report.
Most EMC laboratories already perform automated EMI measurements.Still, it is very important to ensure that the system and software areconfigured correctly. IEC standard CISPR16 provides the guidelines forautomated EMI measurement which are being followed by operators andsystem integration engineers. The guidelines ensure that an automatedEMI test will produce the same results for a given EUT wherever andwhenever the test is done.