The spectrum analyzer is the most widely used tool in electromagnetic interference (EMI) testing, diagnosis, and troubleshooting. This article will focus on the versatility of the spectrum analyzer as a diagnostic test instrument in a wide range of EMI applications. A spectrum analyzer is as important to an electromagnetic compatibility (EMC) engineer as a logic analyzer in the hands of a digital circuit design engineer. The spectrum analyzer's wide frequency range, bandwidth selectability, and wide-range scanning CRT display make it useful in almost every EMC test application. With the emergence of digitally modulated data transmission and ultra-wideband (UWB) transmission methods, coupled with the increasing frequency of unintended radiation devices in the form of high-speed digital clocks, the current EMI specifications can no longer fully address all existing interference types and their effects on communications. The impact of the system.
Bursts. Short- and high-frequency interference that occasionally occurs in consumer electronics and communications is becoming more and more common. For example, pattern-dependent spread spectrum clocks used in computers and noisy regular hard disk access cycles in embedded system designs. Hard drive. These complex digital devices are increasingly approaching wireless communication systems that operate in a frequency agile and packet-based mode.
As the interference characteristics of communication systems change, so do test equipment. The functions previously implemented by analog circuits can now be implemented digitally, and the measurement speed continues to increase, and we can obtain measurement results faster. The real-time spectrum analyzer introduced by Tektronix can instantly view a very wide spectrum span without losing information in the frequency band, so that it can find, capture and measure transient peaks that are extremely challenging to traditional technologies.
Diagnostics, pre-conformance and conformance testing
In the field of electromagnetic compatibility (EMC), different equipment and technologies are used in different stages of design and inspection. In the early stages of development, the combination of EMC design technology and diagnostics resulted in lower EMI characteristics and low sensitivity to external and internal interference. A general-purpose spectrum analyzer with corresponding filters and detectors is usually used to determine the impact of design optimization on EMC. Usually complete the detection directly on the circuit board, or use E-field and H-field probes to determine the influence and shielding effect of design optimization.
Of course, diagnosis has no limitation in ensuring excellent EMC performance; it usually requires comprehensive diagnosis and debugging of system integration to ensure that all RF subsystems reach the required performance level and will not be degraded by other parts of the integrated system. Perform pre-compliance testing after system integration to identify problem areas in the design.
Meeting international standards does not require pre-conformance testing. The goal of pre-conformance testing is to discover potential problems and reduce the risk of failures during the conformance testing phase. The equipment used can be non-standard equipment. If sufficient margin is added to the test results, its accuracy and dynamic range can be lower than standard receivers. Pre-conformance testing can be done in a certified laboratory using rapid measurement techniques, which are designed to "quickly see" problem areas; pre-conformance testing can also be done by engineering designers in temporary locations. Pre-certification usually uses general-purpose spectrum analyzers that include corresponding filters and detectors because they provide fast measurement tools that are usually already used in the design process and do not require additional capital expenditure. If problems are found at this stage, then further diagnosis and design changes are required.
In addition to diagnosis, the functions provided on the RSA6100A can also perform some pre-consistency measurements. Figure 5 is an example of pre-consistent scanning, which combines the trajectory of the CISPR QP detection with the antenna factor table and the spurious signal search function. In this example, the trajectory is an "environmental scan", which examines the background signal that exists when there is no device under test.
Conformance testing requires methods, equipment and measurement locations in compliance with international standards. Conformance testing is usually done as part of the design inspection before equipment production. Conformance testing is exhaustive testing and takes a long time. EMC failures at this stage of product development may cause expensive redesigns and delay product launches.
Filters, detectors and averaging
A filter is a device used to eliminate interference noise, which filters the input or output to obtain pure direct current. The circuit that effectively filters out the frequency point of a specific frequency or the frequency other than the frequency point is a filter, and its function is to obtain a specific frequency or eliminate a specific frequency. The filter is a network composed of inductors and capacitors, which can separate mixed AC and DC currents. In the power rectifier, the network is used to filter the ripples in the pulsating DC to obtain a relatively pure DC output.
A device that can detect changes in the components and their amounts effluent from the chromatographic column. Refers to mechanical, electronic or chemical devices used to distinguish, record or indicate changes in a certain variable in the environment, such as temperature, pressure, charge, electromagnetic radiation, nuclear radiation, particles or molecules, etc. For example, an ultraviolet detector is a device that converts the change of light intensity after passing through the substance to be measured into an electrical signal. This type of signal converter is also called a transducer or an identifier in English. It is an instrument that detects the physical or chemical properties or content changes of chromatographic separation components (in most cases, they are converted into corresponding voltage and current signals). It is a key component in the chromatographic system and the eye of the chromatographic separation process.
In many commercial EMI measurements, these measurement units are stipulated by Comite International Special des Perturbations Radioelectriques (CISPR), which is a technical agency under the International Standards Agency-International Electrotechnical Commission (IEC). Other standards and certification bodies, such as Japan's TELEC, also have requirements for measurement methods and certification techniques. The US Department of Defense has developed the MIL-STD 461E standard, which places special requirements on military equipment.
The measurement bandwidth is determined by the receiver bandwidth shape or the resolution bandwidth (RBW) filter of the spectrum analyzer. The test bandwidth is usually a frequency band where interference may exist in the frequency spectrum, and the test bandwidth varies with the frequency.
The detector is used to calculate a single point that represents the signal at a certain point in time. The detection method can calculate the positive or negative peak value, the RMS or average value of the voltage, or in many EMI measurements, calculate the quasi-peak value (QP).
During the measurement, the averaging method is used on the detected signals. The averaging algorithm defined by the CISPR standard aims to reproduce the effect of using a voltmeter with a specified response time to read the signal value. By applying a specified bandwidth to the detected output, a "video filter" can also be used for averaging. For EMI testing, video filtering is specified in the TELEC standard.
Although many EMI measurements can be done using a simple peak detector, the EMI measurement standard specifies a dedicated measurement method, the quasi-peak (QP) detector. The QP detector is used to detect the weighted peak value (quasi-peak value) of the signal envelope. It weights multiple signals based on signal duration and repetition rate. The QP detector is characterized by fast response, slow attenuation, and includes a time constant that represents the critical damping table. For a signal with a higher frequency, the QP measurement value is higher than that of an occasional pulse.
Quasi-peak detectors have traditionally been used in analog designs, as shown in Figure 1.
To help view the response of the quasi-peak detector and the related instrument combination, Figure 2 shows the input response (repetitive pulses, indicated in blue) and the obtained quasi-peak detector response (with the characteristics of fast response and slow attenuation, indicated in green) And the integrated response of the detector and the meter (indicated in red) is separated.
For the constant index on the receiver with QP detector, Figure 3 shows the relationship between the amplitude and the repetition frequency described in the CISPR 16-1-1 standard.
It is an example of peak detection and QP detection. Here, the signal with 8μs pulse width and 10ms repetition rate was viewed in peak detection and QP detection. The obtained QP value is 10.1dB lower than the peak value. When measuring the EMI of the device under test, the peak value is usually measured first to find the problem area that exceeds or is close to the specified limit. Then only perform slower quasi-peak measurements on signals that are approaching or exceeding the limit. A spectrum analyzer with a standard peak detector is usually used to quickly assess any problem areas.
The influence of peak and quasi-peak detection on the signal with 8μs pulse width and 10ms repetition rate, the quasi-peak value is 10.1dB lower than the peak value
Averaging and video filters
In addition to the QP detector, the real-time spectrum analyzer also supports the peak and average detectors specified in the CISPR specification. The peak detector detects the peak value of the signal envelope, and the average detector calculates the average value of the envelope. The real-time spectrum analyzer can simultaneously measure QP, peak value and average value from the same input signal, and understand the signal characteristics of the DUT in a unique way. Some EMI measurements specify a video filter, which is the earliest method used in spectrum analyzers to reduce the impact of measurement noise changes. The term video filter originates from the earliest implementation, that is, a low-pass filter is placed between the detected output and the Y-axis analog drive input of the spectrum analyzer CRT. Real-time spectrum analyzers and some modern spectrum analyzers use digital technology to smooth the noise on the signal.
In most EMI measurements, the video filter is specified to be off, or the video filter is specified to be at least three times higher than the specified RBW of the measurement (see Table 1).
The purpose of specifying that the video filter is turned off (or not less than 3 times RBW) is to eliminate the influence of the video filter on the detected signal. Figure 4 shows the effect of video bandwidth when the ratio of video bandwidth (VBW) to RBW changes. When VBW≥3*RBW or 10*RBW (or failure), the noise standard deviation remains at 5.4dB. When VBW=RBW, for example, in the TELEC specification section, the noise change is reduced to about 4.7dB.
Digital implementation of EMI filter, detector and averaging algorithm
The standard EMI filter is usually a low-pass filter circuit composed of a series reactor and a parallel capacitor. Its function is to allow the frequency signal of the equipment to enter the equipment during normal operation, and it has a greater blocking effect on the high-frequency interference signal. The power cord is the main way to interfere with incoming and outgoing equipment. Through the power cord, the interference from the power grid can enter the equipment and interfere with the normal operation of the equipment. The interference generated by the equipment may also be transmitted to the power grid through the power cord and interfere with other equipment. The normal operation of the equipment.
The detector is a device that detects some kind of useful information in the fluctuating signal. A device used to identify the presence or change of waves, oscillations, or signals. The detector is usually used to extract the information carried. The detector is divided into envelope detector and synchronous detector. The output signal of the former corresponds to the envelope of the input signal, and is mainly used for demodulation of standard amplitude modulation signals. The latter is actually an analog multiplier. In order to obtain the demodulation effect, an oscillating signal (coherent signal) that is exactly the same as the carrier of the input signal needs to be added. Synchronous detectors are mainly used for the demodulation of single-sideband AM signals or the demodulation of vestigial sideband AM signals.
For spectrum analyzers based on Discrete Fourier Transform (DFT) technology, filtering can be performed digitally by applying a window function to discrete collected data. The sampling rate depends on the bandwidth of the required filter. When the sampling frequency is the same, more samples are required to achieve a smaller filtering bandwidth.
The real-time spectrum analyzer uses the Kaiser window to simulate the EMI filter. The frequency response amplitude of the window function determines the shape of the IF filter, which must meet the bandpass selection restrictions specified in CISPR 16-1-1.
In a real-time spectrum analyzer, the quasi-peak detector is implemented using a digital filter. Digital filters, such as infinite impulse response (IIR) filters, can be used to simulate the RC charging and discharging circuits used in traditional EMI receivers. This critical damping table can also be modeled as a second-order digital IIR filter. The maximum value displayed on the meter is taken as the value of the quasi-peak detector.
On the real-time spectrum analyzer, the video filter is implemented using averaging technology. The average number used depends on the selected video bandwidth and the RBW used in the measurement. When using VBW, the measurement analysis length obtained depends on the selected VBW. If RBW is used when there is no video bandwidth, it will be longer. The real-time spectrum analyzer selects the average number to achieve a good correlation with the noise change on the VBW/RBW curve, as shown in Figure 5.
Measurement speed and real-time spectrum analyzer
The measurement speed of QP and averaging has always been a challenge faced by measurement receivers and spectrum analyzers. QP detectors and meters have long response times, so it is not practical to scan one frequency at a time in a wide range of frequencies. To solve this problem, the measurement is completed with a peak detector, which can quickly determine the highest EMI peak in the device under test. Then use a single frequency measurement in all problem areas and perform the measurement repeatedly.
Recently, receivers and real-time spectrum analyzers that can handle large signal spans have appeared on the market. The speed of applying QP detection and averaging functions is several orders of magnitude higher than that of single frequency measurement technology. This method of calculating all frequency points in the bandwidth has a clear speed advantage. Compared with the scanning technology, there is another advantage: the transient signal in the frequency band can be viewed with a much higher probability of listening. This is particularly important in the current design environment, because the signal changes and moves with time, and single-frequency measurements cannot represent these dynamically changing signals.
Find current EMI problems
Although the standards-based measurement methods described above are indispensable for statutory compliance testing, they usually cannot solve or even detect the problems faced by EMI designs in current systems.
Fortunately, the development of measurement technology has been able to meet these needs. The above describes how to use a real-time spectrum analyzer to find transient signals and EMI hidden dangers. In the following example, a single transient signal will produce a series of transient signals, each of which only lasts a short time. In this example, the device is an embedded system that caused transient EMI when buffering data to the hard disk. After a simple check with the peak detector of a swept frequency analyzer (yellow trace, as shown in Figure 6) , It seems that there is only one continuous signal; keeping the instrument at Max-hold for a few minutes while cycling the DUT working mode will indicate the problem (blue trace). But performing a fast scan in the peak detection mode will result in a yellow trace, and no problem is detected.
Use Digital Phosphor Processing (DPX) technology to investigate the EMI characteristics of the DUT and find the problem immediately. The unique DPX spectrum display technology of the Tektronix real-time spectrum analyzer can process more than 48,000 spectrums per second, ensuring that signals with a duration of more than tens of microseconds can be captured and displayed instantaneously. In Figure 6, signals with a higher frequency of occurrence are shown in red, and signals with a lower frequency of occurrence are shown in blue and green. In this way, you can immediately see which signals are continuous and which are transient. Transient signals occasionally appear, but their levels are higher than continuous signals.
After using DPX to discover potential problems, trigger and capture signals for further analysis. By defining the frequency template trigger based on the continuous signal curve, and then capturing the occasional transient signal in the spectrum, you can easily trigger and capture the signal. Any signal with a duration longer than 10.3μs and higher than the frequency mask threshold will cause a trigger, and the pre-trigger and
The signal after the trigger is stored to the 4 acquisition results triggered by the transient signal.
Now you can fully analyze the signal. The mark in Figure 8 shows that the repetition rate of the transient signal is 1.0s, but the length of the transient signal is not always the same, but varies from 752μs to 200μs during the 5 acquisitions. This repetition frequency and varying pulse width provide important clues to determine the source of the transient signal in the circuit, in this case the disk cache operation, which only occurs under the special operating conditions of the device under test.