Simple and effective computer-aided EMC diagnosis method for electronic products

In the EMC (electromagnetic compatibility) compliance process of electronic products, reducing the number of cycles or simplifying the EMC test method can effectively reduce the cost of EMC compliance and shorten the compliance cycle. Based on this idea, this paper proposes a computer-aided EMC design and diagnosis method. A "golden standard method" is proposed and described in detail as the basis for computer-aided EMC diagnosis, and a brief application case of computer-aided EMC diagnosis is introduced.

In the EMC (electromagnetic compatibility) compliance process of electronic products, reducing the number of cycles or simplifying the EMC test method can effectively reduce the cost of EMC compliance and shorten the compliance cycle. Based on this idea, this paper proposes a computer-aided EMC design and diagnosis method. A "golden standard method" is proposed and described in detail as the basis for computer-aided EMC diagnosis, and a brief application case of computer-aided EMC diagnosis is introduced.

Electronic products must meet EMC (electromagnetic compatibility) standards before they can be sold in developed countries such as Europe and the United States. The China CCC certification, which was enforced on August 1, 2003, also includes EMC as a mandatory inspection item. EMC testing usually includes electromagnetic conductivity and radiation emission testing and anti-interference characteristics testing of electronic products. These tests are becoming increasingly complex and can only be performed in professional EMC laboratories, consuming a lot of manpower and material resources, resulting in high testing costs. If the tested sample fails the test of the EMC laboratory, the tested sample can only be rectified based on the experience of the design engineer, and then sent to a professional EMC laboratory for testing to verify the effect of the modification. This cycle repeats itself, making the product's EMC testing and improvement costs uncontrollable and the development cycle unpredictable (Figure 1). EMC compliance of electronic products occurs not only in new product development, but also throughout the product life cycle. Any changes to the product during the product sales period will lead to re-certification of the product's EMC. Therefore, finding a low-cost way to meet EMC standards has become a top priority.

Computer-aided EMC design

As shown in Figure 1, reducing the number of cycles in the product EMC compliance process is one of the effective ways to resolve the above contradictions. Computer-aided EMC design can improve the EMC status of electronic products in advance and reduce the number of times electronic products are tested in EMC professional laboratories (Figure 2), thereby reducing the cost of EMC compliance and shortening the EMC compliance cycle. Therefore, it has received increasing attention from scientific research institutions in various countries. At present, the more typical computer-aided EMC design software includes HFSS electromagnetic field simulation software from ANSOFT in the United States, which uses the finite element method to solve from the perspective of the field and is suitable for antenna simulation; and HyperLynx software from INNOVEDA, which uses the circuit method to perform crosstalk and EMI (electromagnetic interference) analysis on printed circuit boards. Among them, field analysis software is complex to use, has a narrow application range, and is suitable for qualitative analysis; circuit analysis software is limited to the EMC evaluation of local segments (nodes) of printed circuit boards, and it is difficult to fully reflect the status after signal superposition. The high price of EMC auxiliary design software also limits their wide application.

Computer-aided EMC diagnosis

As shown in Figure 1, another effective way to reduce the cost of EMC compliance and shorten the EMC compliance cycle is to simplify EMC testing and reduce testing costs. Computer-aided EMC diagnosis is such a method (Figure 3). As shown in Figure 3, after the prototype is modified in the EMC compliance process, it is no longer sent to the EMC professional laboratory for testing. Instead, computer-aided EMC diagnosis is applied at the user's site at any time to verify the modification effect, thereby accelerating the compliance process; since computer-aided EMC diagnosis is relatively inexpensive compared to professional EMC laboratory testing, it reduces the cost of compliance.

Computer-aided EMC diagnosis usually consists of two parts: physical testing of the object being tested and software analysis of the test results. The golden standard method is the basis for the correct implementation of computer-aided EMC diagnosis.

1. Golden Standard Method

The golden standard method is an effective means to assist EMC low-cost testing to obtain reasonable and repeatable results. Its implementation steps are as follows:

* Obtain at least one standard test result as a reference point. First, test the sample to be tested in a professional EMC laboratory using standard methods to obtain standard test results. Standard test results are used to indicate the frequency of the sample to be tested that exceeds the standard. The interference of these frequencies will be eliminated or reduced by applying the golden standard method through computer-aided EMC diagnosis.

* Copy the standard laboratory layout to the user's site. Copy the layout of the tested samples in the professional EMC laboratory to the user's site as intact as possible. If it is not possible to copy, mark the wiring and placement of the tested samples before leaving the laboratory, and bundle various cables so that they can be accurately restored at the user's site.

* Determine a test site at the user's site. This test site is a dedicated place for users to test samples using the gold standard method, and its electromagnetic environment should be kept as constant as possible.

* Construct a golden standard. A golden standard consists of a representative sample of the object being tested, with a fixed set of settings (including software versions), cables, and peripherals. Ideally, this would be a wooden board the size of a typical test lab bench, to which the object being tested and its cables are glued or otherwise permanently affixed. This permanently fixed, unchanging combination is the golden standard. In some companies, a representative sample of the golden standard and its cables are painted bright pink, superglued to a base plate, and then locked in an out-of-the-way location with only a lock, which may be the only way to prevent someone from selling a part of the golden standard to a customer or removing parts from it. Always make sure to use the same cables in the golden standard, not just the same specifications, even if they are from the same supplier.

* Set up the instruments and auxiliary equipment required for the test. To eliminate the interference transmitted from the power supply, the connection between the tested sample and the external power supply should be isolated through the LISN. The test equipment used in the first test, including the LISN and the test cables and probes of the test instruments, and their settings and operation methods can be adjusted at will, but cannot be changed after the first test.

* Perform the first test on the sample to be tested. Evenly select several points on the sample to be tested as test points. Test points should be numbered. If using an oscilloscope, the waveform of each measurement point should be stored in two different file formats, image and data, with different file names. The file names and test point numbers should correspond to each other and be recorded (preferably in a spreadsheet). Take a photo of the placement of the gold standard, LISN, and test instrument so that the first test site can be accurately restored the next time the measurement is made. All records should be detailed enough to enable others to repeat the test on the gold standard many years later.

* Save and utilize the original golden standard and measurement result template. The prototype used for the first measurement of the golden standard is carefully preserved as the original golden standard, and its measurement results will be used as the measurement result template as a reference point for future measurements. In the process of applying the golden standard method, if the test is not performed frequently, the first thing to do for each measurement is to test the original golden standard. If the measurement results match the template, it means that the measurement method and system settings are correct, regardless of whether the prototype of the golden standard in the next test has a PCB replaced, a microprocessor or IC added, or other improvements made based on the last measurement results, the measurement can be performed immediately. If the test is performed every day, the measurement of the original golden standard can be done once a week to see if there is any "drift" in the measurement method, or damage to the antenna, cable, probe or other equipment. If an antenna is changed, or a cable is broken and repaired, re-measure the original golden standard to ensure repeatability.

* Improve the design of the sample to be tested based on the gold standard. Install another sample to be tested on the base of the original gold standard as the object of improvement. Modify the object of improvement based on the results of the previous test and analysis using the computer EMC auxiliary diagnosis system, then test the test points of the improved object and analyze the test results again using the computer EMC auxiliary diagnosis system. Compare each analysis result with the previous result to ensure that the improvement is moving in the right direction until the intensity of the excessive interference frequency indicated by the standard test results is reduced to below the expected level.

2. Data Collection and Analysis

The following factors should be considered in data collection:

* It can reflect the electromagnetic interference distribution of the tested sample, so that users can find the interference source;

* The test method is simple and easy to use with low cost;

* Testing can be performed at the user's site at any time.

The data analysis method should mainly consider the following factors:

* Ability to find the physical location of the interference source by analyzing the collected data;

* Easy to operate and can interface with common electronic product computer-aided design software;

* By comparing with the last analysis results, evaluate the effect of modifications to the tested prototype to guide the user to modify the tested prototype in the right direction.

At present, there is an EMC-Scanner system (electromagnetic radiation and thermal radiation scanning system) that uses a mechanical scanning method to scan the object under test and uses a PC to display a two-dimensional or three-dimensional image of electromagnetic radiation and thermal radiation. However, the product under test is limited by the size of the scanner frame and is mainly suitable for the PCB board of the product.

Another computer-aided EMC diagnostic system under development uses a digital oscilloscope commonly owned by electronic product manufacturers to record the voltage waveform of the tested product at the user's site according to the golden standard method and uses the newly developed EMCExplorer software to transform the tested signal into a time-frequency domain signal, indicating the specific location of the interference in the tested prototype and the location in the time domain, and can quantitatively compare the results of prototype modification to make the prototype modification go in the right direction. EMCExplorer software can also interface with existing printed circuit board EDA software (such as PROTEL) to directly analyze the simulation results of EDA software and realize EMC computer-aided diagnosis in the electronic product design stage.

Application Examples

The EMC test results of a typical DC/DC converter show that EMI exceeds the standard at frequencies of 4MHz, 10MHz and 13MHz. Some test points are evenly selected on the printed circuit of the converter and numbered (in this example, we select 19 points, numbered from 00 to 18), as shown in Figure 4. The selected test points are tested with a digital oscilloscope (in this example, we only measure the PCB board, but in actual applications, samples can be taken from any part of the tested product), and the test waveform of each point is saved as a corresponding data file (SC1.001-SC1.019) for EMCExplorer analysis. EMCExplorer analyzes the 13MHz interference frequency of each test point and obtains the results in Figure 5, among which the interference amplitude of data files SC1.019 (point 18) and SC1.004 (point 03) is the largest. The same conclusion is obtained from the analysis of the 10MHz interference frequency of each test point. The analysis of the 4MHz interference frequency of each test point shows that point 18 has the largest interference, followed by point 9. Analyzing the above test results, we found that point 3 is the grounding point, but the interference value is abnormally high. Considering that there is a T1 oscillator near point 3, the interference is likely to be caused by this. Further observation revealed that the T1 grounding welding was poor. After re-welding and testing again, the 13MHz interference at the point 3 grounding point was indeed reduced. Considering that point 18 is the concentration point of interference, we welded a 300uf capacitor between point 18 and point 0, and the EMI of 13MHz and 10MHz frequencies was further reduced below the standard curve.

We then do a time-frequency analysis of the data at point 9 (Figure 6). The upper figure in Figure 6 is the original waveform, the middle figure is the time-frequency analysis of the waveform, and the lower figure is the Fourier transform of the waveform. The middle figure shows that the most serious 4 MHz interference occurs at times 0.012ms, 0.022ms, and 0.042ms, although the waveform amplitude at these moments does not appear to be the largest in the upper figure. Based on the phase relationship of the waveforms at these moments, we found that these waveforms were generated by the opening or closing of different switching diodes. By replacing the original devices with switching diodes with slower opening or closing speeds, we effectively reduced the 4MHz interference at these moments.

summary

Computer-aided EMC diagnosis can simplify the EMC test in the EMC compliance process of electronic products and can be carried out at any time at the user's site, thereby reducing the cost of electronic product EMC compliance and shortening the compliance cycle. Computer-aided EMC diagnosis consists of two parts: data collection and data analysis, which are correctly implemented through the golden standard method. The data analysis software can be interfaced with the printed circuit board EDA software to realize EMC computer-aided diagnosis in the electronic product design stage.

Author: Wu Wei

Wenzhou Galileo Technology Co., Ltd.

Technical Director

Email: w.wu@ieee.org