1. Why is electromagnetic compatibility (EMC) design required for products?
Answer: To meet product functional requirements, reduce debugging time, ensure products comply with EMC standards, and prevent products from generating electromagnetic interference to other equipment in the system.
2. From which aspects can EMC design be carried out for products?
Answer: Circuit design (including component selection), software design, PCB design, shielding structure, signal/power line filtering, and circuit grounding design.
3. Why is the decibel (dB) unit always used in the EMC field?
Answer: Because the amplitude and frequency ranges to be described are very wide, which are easier to represent on a logarithmic coordinate in graphs, and dB is the unit used for logarithmic representation.
4. I have little knowledge of EMC, but the data transmission rate in circuit design is getting faster and faster now. I have encountered some PCB EMC problems when making PCB boards, but my understanding is superficial. I want to learn this knowledge well, not just follow the trend. I truly realize that EMC will become increasingly important in future circuit design. As I said before, I don't have a deep understanding and don't know where to start. I want to ask what basic knowledge is needed to excel in EMC, which basic courses should be learned, and what is a good way to learn? I know it's not easy to master any knowledge and don't expect to understand it in a short time. I just hope for some suggestions to avoid detours as much as possible.
Answer: For EMC, you first need to understand EMC standards such as EN55022 (GB9254) and EN55024, as well as basic test principles. In addition, you need to learn the application of EMI components such as capacitors, magnetic beads, differential-mode inductors, and common-mode inductors. At the PCB level, you need to understand PCB layout, stack-up structure, the impact of high-speed routing on EMC, and relevant rules.
Another point is to master the ideas for analyzing and solving EMC problems. These are the basic knowledge that a hardware engineer must master in the future!
5. As a novice in PCB design, I want to ask what knowledge I should master to do PCB design well? In addition, where can I generally find the safety regulation knowledge encountered in PCB design? I look forward to your guidance and would be very grateful!
Answer: For PCB design, you should master:
① Proficiency in relevant PCB design software such as POWERPCB/CADENCE;
② A clear understanding of the specific architecture of the designed product, as well as schematic circuit knowledge including digital and analog knowledge;
③ Mastery of PCB processing flow, technology, and maintainable processing requirements;
④ Knowledge of high-speed signal integrity, electromagnetic compatibility (EMI and EMS), SI, and PI simulation design of PCB boards;
⑤ RF knowledge if the relevant work involves radio frequency;
⑥ For safety regulation knowledge in PCB design, mainly refer to GB4943 or UL60950; general insulation spacing requirements can be obtained by looking up tables!
6. Basic principles of EMC design
Answer: Electronic circuit design criteria: Electronic circuit designers often only consider product functions without integrating functions and electromagnetic compatibility. Therefore, while completing their functions, products also generate a large number of functional disturbances and other interferences, and fail to meet sensitivity requirements. The EMC design of electronic circuits should consider the following aspects:
Component selection: In most cases, the degree to which the basic components of a circuit meet electromagnetic characteristics determines the degree to which functional units and the final equipment meet EMC requirements. The main criteria for selecting suitable electromagnetic components include out-of-band characteristics and circuit assembly technology, because the realization of EMC is often determined by the component response characteristics far from the fundamental frequency. In many cases, circuit assembly determines the out-of-band response (e.g., lead length) and the degree of coupling between different circuit components.
Specific rules:
① At high frequencies, compared with lead-type capacitors, feedthrough capacitors or chassis capacitors with low lead inductance should be preferred for filtering.
② When lead-type capacitors have to be used, the impact of lead inductance on filtering efficiency should be considered.
③ Aluminum electrolytic capacitors may experience temporary dielectric breakdown of several microseconds, so solid capacitors should be used in circuits with large ripples or transient voltages.
④ Use resistors with low parasitic inductance and capacitance; chip resistors can be used in the ultra-high frequency band.
⑤ Large inductors have high parasitic capacitance. To improve the insertion loss in the low-frequency part, do not use single-section filters, but multi-section filters composed of several small inductors.
⑥ Pay attention to the saturation characteristics when using magnetic core inductors, especially that high-level pulses will reduce the inductance of magnetic core inductors and the insertion loss in filter circuits.
⑦ Use shielded relays as much as possible and ground the shield housing.
⑧ Select input transformers with effective shielding and isolation.
⑨ Power transformers used in sensitive circuits should have electrostatic shielding, and both the shield housing and transformer housing should be grounded.
⑩ Interconnection signal lines inside the equipment must use shielded wires to prevent disturbance coupling between them.
⑪ Select connectors with enough pins to connect each shield to its respective pin.
7. Problems with square wave pulse-driven inductive sensors
Answer: ① Conduct signal testing in a shielded environment as much as possible; if not convenient, at least shield the sensor and the front stage.
② Use differential probes for testing as much as possible, or at least shorten the length of the probe ground wire as much as possible to reduce test errors.
③ The actual operating frequency of your circuit is not too high; ringing can be reduced through routing. For better noise performance, the suppression of common-mode signals should be considered, and a common-mode reactor can be inserted if necessary. At the same time, pay attention to the switching power supply noise in the entire working environment and avoid power supply coupling.
④ If allowed by the sensor, current amplification mode can be used, which helps improve speed and reduce noise. Place the analog switch after the preamplifier as much as possible; although an additional preamplifier is added, the performance is greatly improved and the debugging difficulty is reduced.
⑤ If the waveform is a major concern, consider additional frequency compensation; if only digital detection is required, the operating frequency should be reduced. In short, use low frequencies if possible and block DC if possible.
⑥ Pay attention to anti-aliasing filtering before AD conversion and software filtering to improve data stability.
8. Performance of GPS electromagnetic interference: Especially for GPS applied in PMP products, which are handheld and vehicle-mounted GPS terminal products with MP4, MP3, FM radio + GPS navigation functions, an internal GPS Antenna is a must. In this way, the GPS Antenna is prone to generate EMI/EMC electromagnetic interference with components such as MCU, SDRAM, and crystal oscillator on the GPS terminal product, resulting in a significant drop in the satellite acquisition capability of the GPS Antenna and almost making normal positioning impossible. What methods can be taken to solve such EMI/EMC electromagnetic interference?
Answer: An ESD Filter can be added, which has both anti-static and anti-electromagnetic interference functions. Our mobile phone customers with GPS functions use this method. Manufacturers of such products include Tyco (Raychem), Chipbond, South Korea ICT, and many others.
9. Almost all important signal lines on the board are designed as differential pairs to enhance the anti-interference ability of signals. I have always had many confusions:
① Are differential signals only defined for analog signals or digital signals, or both?
② In actual circuit diagrams, how to analyze the frequency response of networks such as filters on differential pairs? Is it the same as the method for analyzing general two-port networks?
③ How to convert differential signals carried on differential pairs into general signals? What are the signal waveforms on differential pairs and their mutual relationship?
Answer: ① A differential signal is a circuit that transmits one signal using two signal lines and makes judgments based on the voltage difference between the signals. It can be either an analog signal or a digital signal. All actual signals are analog signals, and digital signals are only the sampling results of analog signals quantized by threshold levels. Therefore, differential signals can be defined for both digital and analog signals.
② Good question about the frequency response of differential signals. An actual differential port is a four-port network with two analysis methods: differential-mode and common-mode.
When analyzing the frequency response, a common-mode sweep source with the same polarity and a differential-mode sweep source with opposite polarities should be added respectively, and the corresponding terminal needs to set the common-mode voltage test point Vcm＝(V1＋V2)/2 and the differential-mode voltage test point Vdm=V1-V2 accordingly. There are many articles on the Internet about the impedance calculation and principle of differential signals for detailed understanding.
③ Differential signals usually enter a differential drive circuit and are amplified to obtain differential signals. The simplest one is the differential common-emitter mirror amplifier circuit, which is introduced in general analog circuit textbooks.
They are generally required to be completely out of phase with a sufficient voltage difference greater than the differential-mode voltage threshold. Of course, the signal inevitably contains common-mode components, so a very important indicator of the differential amplifier is the common-mode rejection ratio Kcmr=Adm/Acm.
10. I designed a speed control circuit for the company's DC magnetic steel motor. The power supply end was connected with 0.33uF + Sharp TV inductor + 0.33uF, but the effect was not ideal. Later, 4 inductors were connected in series at the PCB power supply end, but it exceeded 12dB in the 30~50MHz range. How to handle it?
Answer: Generally speaking, LC or PI-type filter circuits have better effects than single capacitor filtering or inductor filtering. It is unclear what you mean by the unsatisfactory effect of using 0.33uF + Sharp TV inductor + 0.33uF at the power supply end—was it radiation exceeding the standard? In which frequency band? I guess the feedback noise amplitude in the power supply loop of the DC magnetic steel motor is large and the frequency is low. It is necessary to use an inductor with a larger inductance for filtering and adopt multi-stage capacitor filtering at the same time, which will achieve better results.
11. I recently want to make a broadband amplifier with 0--150M bandwidth and gain not less than 80dB! What problems should I pay attention to in terms of EMC?
Answer ①: Special attention should be paid to low noise issues when designing a broadband amplifier, such as ensuring a sufficiently stable power supply.
Answer ②: ⑴ Pay attention to the impedance matching of input and output, such as common-base input and emitter-follower output;
⑵ Decoupling of each stage, including high and low frequency ripples;
⑶ Deep negative feedback, as well as prevention of self-oscillation and loop self-oscillation;
⑷ Design of band-pass filters.
Answer ③: It's really hard to answer without seeing the actual design. All suggestions are still the usual ones: pay attention to the three elements of EMC, the conduction and radiation paths, and the power distribution and ground bounce noise.
150MHz is the analog signal bandwidth—how fast is the rising edge of the digital signal? If the corner frequency is also below 150MHz, I personally think that conduction coupling and power plane radiation will be the main considerations. First, design the power distribution, division and decoupling circuits well.
With an 80dB gain, it is quite high. Do a good job in the isolation and protection of the front-stage small signals and their reference power supply and ground, and minimize the power supply impedance of this part.
12. Ask for EMC methods and considerations in the design of low-power DC permanent magnet motors. A 90W DC permanent magnet motor (110~120V, 2000 rpm) was produced, but its EMC has been exceeding the standard. After production, the 16 slots were first changed to 24 slots, and then shaft insulation was done, but it still failed to meet the standard! Now it is necessary to design and produce a 125W motor. How to handle it?
Answer: EMC problems in the design of DC permanent magnet motors are mainly caused by the back electromotive force generated during motor rotation and arcing during commutation. For specific analysis, RMxpert can be used to design and optimize motor parameters, and Maxwell2D can be used to simulate the actual EMI radiation.
13. Can it be set by impedance boundary? Or use similar layered impedance RLC? Or use ADS to design circuits and cooperate with HFSS/EMPro?
Answer: Lumped resistors can be realized by RLC boundaries; if it is a thin-film resistor, it can be realized by surface impedance or impedance editing.
14. I am currently doing an electrostatic test on a machine with a circle of metal decorative parts on the shell. During the test, I encountered the problem that the 32k crystal oscillator is normal at 4kV contact discharge, but stops oscillating at 8kV air discharge. How to handle it?
Answer: With metal parts, the effects of air discharge and contact discharge are almost the same. It is recommended to spray insulating paint on the metal bracket and try.
15. It is very troublesome for us to measure PCB electromagnetic radiation now. We use a spectrum analyzer plus a self-made near-field probe. Not to mention the accuracy problem, we are troubled by high-voltage points for fear of damaging the spectrum analyzer. I wonder if it can be solved by simulation.
Answer: First of all, EMI testing includes near-field probe and far-field radiation testing, and no simulation tool can replace actual testing. Secondly, Ansys's SIwave (PCB single board noise and radiation simulation tool) and HFSS (high-frequency structure simulator for arbitrary 3D structures) can simulate the near-field and far-field radiation of single boards and systems respectively, as well as EMI radiation in a limited shielding environment.
The effectiveness of simulation depends on your consideration of the EMI problems in your design and the corresponding software settings, such as differential-mode or common-mode radiation on the single board, current source or voltage source radiation, etc. Based on our practice and experience, the vast majority of EMI problems can be solved through simulation analysis, and the effect is very good compared with actual testing.
16. I heard that Ansys's EMC tools generally simulate signals above 1GHz. The highest frequency clock line on our board is only 133MHz from the main chip to SDRAM, and most of the other frequencies are at the KHz level. We mainly use ADS/Hyperlynx for SI/PI design, which is relatively simple to operate, but the EMC of the entire board still exceeds the standard, affecting the picture quality. In addition, do your tools have an interface with Mentor PADS?
Answer: Ansys's tools can simulate signals from DC to dozens of GHz and above, but the lossy transmission line model above 1GHz is more accurate compared with other tools.
As far as I know, ADS/HyperLynx is mainly used for SI and crosstalk simulation, as well as a little EMI radiation analysis of single signal lines, and currently has no PI analysis function.
There are many reasons affecting the EMC of a single board. Solving signal integrity and crosstalk is only one aspect of solving EMC. Power plane noise, decoupling strategies, shielding methods, current distribution paths, etc., will all affect the EMC indicators, which can be investigated through simulation in Ansys's SIwave tool. Additional note: Ansys's tools have an interface with Mentor PADS.
17. Please explain when to use bottom layer division to reduce interference and when to use ground layer zoning to reduce interference.
Answer: I have never heard of bottom layer division—what does it mean? Can you give an example? Ground layer division is mainly to improve the isolation between the interference source and the interfered object, such as the isolation between digital and analog circuits.
Of course, division will also bring signal integrity problems such as cross-division. Ansys's SIwave can be used to easily check the isolation between any points.
Of course, there are other ways to improve isolation, such as layering, decoupling, and single-point connection. The effect of specific applications can be simulated by software.
18. A capacitor is connected across two different power copper foil partitions as a return path for high-frequency signals. As we all know, a capacitor blocks DC and passes AC, and the higher the frequency, the smoother the current. My confusion is that the power supply connected to the PCB now is mostly filtered to remove AC, so what does the capacitor pass through as mentioned above? "AC signals"?
Answer ①: This question is quite mysterious, and I have never seen a convincing explanation. For AC, ideally, the power supply and ground are "short-circuited", but in reality, the impedance between them cannot be really 0. The capacitance of the capacitor you mentioned should not be too large to reflect the principle of "less grounding at low frequencies and more grounding at high frequencies". This is probably the value of the existence of this capacitor. It is often encountered that after two components with their own power supplies are connected, inexplicable interference occurs, and the interference disappears when a ceramic capacitor is connected across the two power supplies.
Answer ②: This capacitor is used for voltage stabilization and EMI, and it passes AC signals. It is true that "the power supply connected to the PCB now is mostly filtered to remove AC", but don't forget that digital circuits themselves generate AC signals and interfere with the power supply. When a large number of switching tubes act at the same time, the fluctuation to the power supply is very large.
However, in practice, such capacitors mainly play an auxiliary role to improve the performance of the system. They can be completely omitted if other parts are well designed.
Answer ③: AC is change. For the so-called DC level, such as the power supply, due to the impedance of the wiring, when its load changes, the demand for the power supply will change, either larger or smaller.
In this case, the "series" wiring impedance will generate a larger or smaller voltage drop, so there will be an AC signal on the DC power supply. The frequency of this signal is related to the frequency of the load change. The function of the capacitor is to store a certain amount of charge energy nearby, so that the energy required for this change can be directly obtained from the capacitor.
Approximately, there seems to be an AC loop between the capacitor (which can be regarded as a power supply at this time) and the load. The capacitor acts as an AC loop, roughly like this...
19. The company has made a new mobile phone, which failed a radiation index in the 3C certification at 50-60MHz, exceeding 5dB. It should be caused by the charger. We only added a few capacitors (1uF, 100uF) and nothing else. Are there any good solutions (modifying only the mobile phone circuit without changing the charger)? Can adding capacitors at the charger input end of the mobile phone board solve the problem?
Answer ①: Increase the capacitance of large capacitors and reduce that of small ones, and connect a BIT in series. However, the possibility of the battery being the cause is not very high.
Answer ②: Try short-circuiting and shielding the shell of the variable frequency inductor to the ground.
20. How to avoid high-frequency interference in PCB design?
Answer: The basic idea to avoid high-frequency interference is to minimize the interference of the electromagnetic field of high-frequency signals, which is the so-called crosstalk.
The distance between high-speed signals and analog signals can be increased, or ground guard/shunt traces can be added next to analog signals.
Attention should also be paid to the noise interference of digital ground to analog ground.
21. How to solve the conflict between high-speed routing and EMI in PCB design?
Answer: Resistors, capacitors or ferrite beads added for EMI must not cause some electrical characteristics of the signal to fail to meet the specifications.
Therefore, it is best to first use routing and PCB stack-up techniques to solve or reduce EMI problems, such as routing high-speed signals on inner layers. Only use resistors, capacitors or ferrite beads as a last resort to reduce damage to the signal.
22. How to connect the ground wires between each board when a system consists of several PCBs?
Answer: When signals or power supplies between interconnected PCBs act (e.g., power or signals are sent from board A to board B), an equal amount of current will definitely flow back to board A from the ground layer (this is Kirchhoff's current law).
The current on the ground layer will find the path with the smallest impedance to flow back. Therefore, the number of pins allocated to the ground layer at each interface where power or signals are interconnected cannot be too small to reduce impedance and thus the noise on the ground layer.
In addition, the entire current loop can be analyzed, especially the part with large current, and the connection method of the ground layer or ground wire can be adjusted to control the current path (e.g., create a low impedance somewhere to let most of the current flow from there), reducing the impact on other more sensitive signals.
23. Can a ground wire be added between differential signal lines in PCB design?
Answer: Generally, a ground wire cannot be added between differential signals. Because one of the most important aspects of the application principle of differential signals is to utilize the benefits brought by the mutual coupling between differential signals, such as flux cancellation and noise immunity. Adding a ground wire in the middle will destroy the coupling effect.
24. What is the principle for selecting the appropriate grounding points between the PCB and the shell?
Answer: The principle for selecting the grounding points between the PCB and the shell is to use the chassis ground to provide a low-impedance path for the returning current and control the path of this returning current.
For example, the ground layer of the PCB can usually be connected to the chassis ground with fixing screws near high-frequency devices or clock generators to minimize the area of the entire current loop, thereby reducing electromagnetic radiation.
25. When the size of the circuit board is fixed, if the design needs to accommodate more functions, it is often necessary to increase the routing density of the PCB, but this may lead to increased mutual interference between traces, and overly thin traces also make it impossible to reduce impedance. Please introduce the skills in high-speed (>100MHz) and high-density PCB design?
Answer: Crosstalk interference is indeed a key concern when designing high-speed and high-density PCBs, because it has a great impact on timing and signal integrity.
The following are some points to note:
① Control the continuity and matching of the characteristic impedance of traces.
② The spacing between traces: the commonly seen spacing is twice the line width. Simulation can be used to understand the impact of trace spacing on timing and signal integrity and find the tolerable minimum spacing—the results may be different for different chip signals.
③ Select the appropriate termination method.
④ Avoid the same routing direction of adjacent upper and lower layers, and even traces overlapping exactly up and down, because this kind of crosstalk is larger than that of adjacent traces on the same layer.
⑤ Use blind/buried vias to increase routing area, but this will increase the manufacturing cost of the PCB board. It is indeed difficult to achieve complete parallelism and equal length in actual implementation, but it should be done as much as possible.
In addition, differential termination and common-mode termination can be reserved to mitigate the impact on timing and signal integrity.
26. LC circuits are often used for filtering at analog power supplies in PCB design, but why is the effect of LC filtering worse than RC filtering sometimes?
Answer: The comparison of LC and RC filtering effects must consider whether the frequency band to be filtered and the selection of inductance value are appropriate, because the reactance of the inductor is related to the inductance value and frequency.
If the noise frequency of the power supply is low and the inductance value is not large enough, the filtering effect may not be as good as RC at this time. However, the cost of using RC filtering is that the resistor itself consumes energy with low efficiency, and attention should be paid to the power that the selected resistor can withstand.
27. What is the method for selecting inductance and capacitance values for filtering in PCB design?
Answer: The selection of inductance value not only considers the noise frequency to be filtered, but also the response capability of the instantaneous current.
If the output end of the LC may need to output a large current instantaneously, an excessively large inductance value will hinder the speed of this large current flowing through the inductor and increase the ripple noise. The capacitance value is related to the allowable ripple noise specification value—the smaller the required ripple noise value, the larger the capacitance value.
The ESR/ESL of the capacitor also has an impact. In addition, if this LC is placed at the output end of a switching regulation power supply, attention should also be paid to the impact of the pole/zero generated by this LC on the stability of the negative feedback control loop.
28. EMI problems and signal integrity problems are interrelated. How to balance the two in the process of defining standards?
Answer: Although signal integrity and EMC are closely related, there has been no particularly mandatory standard issued for signal integrity. As for how to balance signal integrity and EMI, this is not a matter of test specifications. To achieve a balance between the two, it is best to reduce the signal rise and fall edges while meeting the requirements of signal integrity.
29. How to meet EMC requirements as much as possible in PCB design without causing too much cost pressure?
Answer: The cost increase of PCB boards due to EMC is usually due to the increase in the number of ground layers to enhance the shielding effect and the addition of devices such as ferrite beads and chokes to suppress high-frequency harmonics.
In addition, it is usually necessary to match other structural shielding structures to make the entire system meet EMC requirements. The following only provides several PCB design techniques to reduce the electromagnetic radiation effect generated by the circuit:
① Select devices with slower signal slew rate as much as possible to reduce the high-frequency components generated by the signal.
② Pay attention to the placement position of high-frequency devices and keep them away from external connectors.
③ Pay attention to the impedance matching of high-speed signals, the routing layer and their return current path to reduce high-frequency reflection and radiation.
④ Place sufficient and appropriate decoupling capacitors at the power pins of each device to mitigate the noise on the power layer and ground layer. Pay special attention to whether the frequency response and temperature characteristics of the capacitor meet the design requirements.
⑤ The ground near the external connector can be properly divided from the ground layer, and the ground of the connector can be connected to the chassis ground nearby.
⑥ Ground guard/shunt traces can be appropriately used next to some particularly high-speed signals, but attention should be paid to the impact of guard/shunt traces on the characteristic impedance of the traces.
⑦ The power layer is indented by 20H from the ground layer, where H is the distance between the power layer and the ground layer.
30. When a PCB board has multiple digital/analog functional blocks in PCB design, the conventional practice is to separate the digital/analog grounds. Why?
Answer: The reason for separating digital/analog grounds is that digital circuits generate noise on the power supply and ground when switching between high and low potentials, and the magnitude of the noise is related to the signal speed and current magnitude.
If the ground plane is not divided and the noise generated by the digital area circuit is large while the analog area circuit is very close, the analog signal will still be interfered by ground noise even if the digital and analog signals do not cross.
That is to say, the method of not dividing digital and analog grounds can only be used when the analog circuit area is far away from the digital circuit area that generates large noise.
31. In high-speed PCB design, what aspects should designers consider for EMC and EMI rules?
Answer: Generally, EMI/EMC design needs to consider both radiated and conducted aspects—the former belongs to the higher frequency part (>30MHz) and the latter to the lower frequency part (<30MHz), so high frequencies cannot be focused on while ignoring low frequencies. A good EMI/EMC design must consider the position of devices, the arrangement of PCB stack-up, the routing of important connections, the selection of devices, etc. from the beginning of the layout. If these are not properly arranged in advance, solving them afterwards will be twice the effort with half the result and increase costs.
For example, the position of the clock generator should be kept away from external connectors as much as possible; high-speed signals should be routed on inner layers as much as possible with attention to characteristic impedance matching and the continuity of the reference layer to reduce reflection; the slew rate of the signal driven by the device should be as small as possible to reduce high-frequency components; when selecting decoupling/bypass capacitors, attention should be paid to whether their frequency response meets the requirements to reduce power layer noise.
In addition, pay attention to the return path of high-frequency signal current to minimize the loop area (that is, minimize the loop impedance) to reduce radiation. The ground layer can also be divided to control the range of high-frequency noise. Finally, select the grounding points between the PCB and the shell appropriately.
32. How to reduce EMI problems by arranging stack-up in PCB design?
Answer: First of all, EMI should be considered from the system level, and PCB alone cannot solve the problem. For EMI, stack-up mainly provides the shortest return path for signals, reduces the coupling area, and suppresses differential-mode interference. In addition, the tight coupling between the ground layer and the power layer, with the ground layer appropriately extended beyond the power layer, is beneficial to suppressing common-mode interference.
33. Why is copper pouring required in PCB design?
Answer: Generally, copper pouring is required for the following reasons:
① EMC: Large-area ground or power copper pouring plays a shielding role, and some special grounds such as PGND play a protective role.
② PCB process requirements: Generally, to ensure electroplating effect or prevent lamination deformation, copper pouring is performed on PCB layers with few traces.
③ Signal integrity requirements: Provide a complete return path for high-frequency digital signals and reduce the routing of DC networks.
Of course, there are also reasons such as heat dissipation and copper pouring required for the installation of special components.
34. Safety regulation question: What are the specific meanings of FCC and EMC?
Answer: FCC: Federal Communication Commission; EMC: Electromagnetic Compatibility. FCC is a standard-setting organization, and EMC is a standard. The promulgation of standards has corresponding reasons, as well as standard and test methods.
35. When making a PCB board, should the ground wire form a closed loop to reduce interference?
Answer: When making a PCB board, generally the loop area should be minimized to reduce interference. When routing the ground wire, it should not be arranged in a closed form, but a dendritic shape is better. In addition, the ground area should be increased as much as possible.
36. How to avoid crosstalk in PCB design?
Answer: A changing signal (e.g., a step signal) propagates along the transmission line from A to B, and a coupled signal is generated on the transmission line C-D. Once the changing signal ends (i.e., the signal returns to a stable DC level), the coupled signal no longer exists. Therefore, crosstalk only occurs during signal transition, and the faster the change (slew rate) of the signal edge, the greater the crosstalk generated.
The coupled electromagnetic field in space can be extracted as a collection of numerous coupled capacitors and coupled inductors. Among them, the crosstalk signal generated by the coupled capacitor on the victim network can be divided into forward crosstalk and reverse crosstalk Sc, which have the same polarity; the crosstalk signal generated by the coupled inductor is also divided into forward crosstalk and reverse crosstalk SL, which have opposite polarities.
The forward and reverse crosstalk generated by coupled inductors and capacitors exist at the same time and are almost equal in magnitude. Thus, the forward crosstalk signals on the victim network cancel each other out due to opposite polarities, and the reverse crosstalk signals superpose and enhance due to the same polarity.
Crosstalk analysis modes usually include default mode, three-state mode and worst-case mode analysis:
- Default mode: Similar to the actual crosstalk test method, that is, the aggressor network driver is driven by a switching signal, the victim network driver remains in the initial state (high or low level), and then the crosstalk value is calculated. This method is more effective for crosstalk analysis of unidirectional signals.
- Three-state mode: The aggressor network driver is driven by a switching signal, and the three-state terminal of the victim network is set to a high-impedance state to detect the crosstalk magnitude. This method is more effective for bidirectional or complex topology networks.
- Worst-case analysis: The driver of the victim network remains in the initial state, and the simulator calculates the sum of crosstalk from all default aggressor networks to each victim network. This method is generally only used for the analysis of individual key networks because there are too many combinations to calculate and the simulation speed is relatively slow.
37. Severe harmonic exceeding of clock signals was found in EMC testing, and only decoupling capacitors are connected to the power supply pins. What aspects need to be paid attention to in PCB design to suppress electromagnetic radiation?
Answer: The three elements of EMC are radiation source, propagation path and victim. The propagation path is divided into space radiation propagation and cable conduction. Therefore, to suppress harmonics, first check its propagation path. Power supply decoupling solves the propagation by conduction; in addition, necessary matching and shielding are also required.
38. In PCB design, the ground wire is usually divided into protective ground and signal ground; the power ground is divided into digital ground and analog ground. Why is the ground wire divided?
Answer: The main purpose of dividing the ground is for EMC considerations, worrying that the noise on the power supply and ground of the digital part will interfere with other signals, especially analog signals, through conduction.
As for the division of signal and protective ground, it is due to the consideration of ESD electrostatic discharge in EMC, similar to the role of lightning rod grounding in our daily life. No matter how it is divided, there is only one ultimate earth, only with different noise discharge paths.
39. In PCB design, is it necessary to add ground wire shielding on both sides when routing the clock?
Answer: Whether to add shielded ground wires depends on the crosstalk/EMI situation on the board, and improper handling of the shielded ground wires may even make the situation worse.
40. Are near-end crosstalk and far-end crosstalk related to the signal frequency and signal rise time? Do they change with them? If there is a relationship, can a formula explain the relationship between them?
Answer: It should be said that the crosstalk caused by the aggressor network to the victim network is related to the signal transition edge—the faster the change, the greater the crosstalk (V=L*di/dt).
The impact of crosstalk on the judgment of digital signals on the victim network is related to the signal frequency—the faster the frequency, the greater the impact.
41. When designing a PCB board, there are the following two stack-up schemes:
Stack-up 1: Signal → Ground → Signal → Power +1.5V → Signal → Power +2.5V → Signal → Power +1.25V → Power +1.2V → Signal → Power +3.3V → Signal → Power +1.8V → Signal → Ground → Signal
Stack-up 2: Signal → Ground → Signal → Power +1.5V → Signal → Ground → Signal → Power +1.25V +1.8V → Power +2.5V +1.2V → Signal → Ground → Signal → Power +3.3V → Signal → Ground → Signal
Which stack-up order is more preferred? For Stack-up 2, will the two divided power layers in the middle have an impact on the adjacent signal layers? These two signal layers already have a ground plane as the return path for signals.
Answer: Both stack-ups have their own advantages. The first one ensures the integrity of the plane layers, and the second one increases the number of ground layers, effectively reducing the impedance of the power plane, which is beneficial to suppressing system EMI.
Theoretically, the power plane and the ground plane are equivalent for AC signals, but in practice, the ground plane has a better AC impedance than the power plane, and the signal prefers the ground plane as the return plane.
However, due to the influence of stack-up thickness factors (e.g., the dielectric thickness between the signal and power layer is smaller than that between the signal and ground layer), the signals crossing the division in the second stack-up also have the problem of incomplete signal return at the power division.
42. When using Protel 99se software to design a PCB, the processor is 80C51, the crystal oscillator is 12MHz, and the system also has a 40KHZ ultrasonic signal and an 800Hz audio signal. How to design the PCB to provide high anti-interference ability at this time? For single-chip microcomputers such as 89C51, what magnitude of signal can affect the normal operation of 89C51? In addition to increasing the distance between them, are there any other techniques to improve the anti-interference ability of the system?
Answer: To provide high anti-interference ability in PCB design, it is certainly necessary to minimize the signal transition edge rate of the interference source signal. The specific frequency of the interference signal depends on the type of level of the interference signal and the length of the PCB routing.
In addition to increasing the spacing, solving the reflection, overshoot and other problems of the interference signal through matching or topology can also effectively reduce signal interference.
43. Is it necessary to pay attention to the power distribution and routing in PCB routing as much as grounding? What problems will be caused if not paid attention to? Will it increase interference?
Answer: If the power supply is treated as a plane layer, the method should be similar to that of the ground layer. Of course, to reduce the common-mode radiation of the power supply, it is recommended to indent the power layer by 20 times the height from the ground layer.
If routing is used, a dendritic structure is recommended, and attention should be paid to avoiding power supply loop problems—power supply closed loops will cause large common-mode radiation.
44. I made a TFT LCD display. During EMC testing by others, interference signals were transmitted through space, causing the image displayed on the screen to shake significantly. Can anyone give some guidance on how to handle it! It is to add interference pulse groups on several signal lines. I don't know the specific name. The interference signals are radiated through the signal lines.
Answer: A separate LCD can hardly pass the pulse group test in EMC testing, especially when using a coupling clamp. If the LCD is used in an instrument, it is not difficult to solve, such as decoupling processing of signal lines, appropriately reducing the impedance of the LCD entrance with conductive paste, and adding a shielding conductive wire mesh on the screen surface.
45. Some time ago, in the EMC test, the GSM fixed wireless phone had radiation spurious phenomena in the 100MHz-300MHz range. Later, the company sent me two shielded shell phones sprayed with electrostatic paint. The laboratory did not allow the replacement of the entire phone, so I replaced the shell of the phone to be modified and tested with the shell sprayed with electrostatic paint of ferromagnetic material. The test results showed that the previous spurious phenomena disappeared, but there was a problem with the main frequency. The main operating frequency of the phone is 902MHz, but several frequencies appeared in the 905-910MHz range. This is the basic situation. During the modification, I only replaced the shell, and no modifications were made to the circuit board and other hardware.
Answer: Phones can be understood as wireless mobile phones, cordless phones, etc. It is necessary to clarify: the type of phone, the operating frequency range of the host, and the type of electrostatic spraying material of the case, such as ferromagnetic or non-ferromagnetic conductive materials and their conductivity.
46. What are the side effects of selecting "pour over all same net objects" when using Protel Dxp for solid copper pouring? Will it cause interference signals to scurry around the entire board, thus affecting performance? I made a low-frequency data acquisition card, so this problem may not need to be worried about, but I still want to figure it out.
Answer ①: For analog-digital mixed PCB boards, it is recommended to separate the analog, digital and ground, and then ground them at the same point, such as connecting with magnetic beads or 0-ohm resistors. High-speed data lines are best to run in parallel with two ground wires to reduce interference.
Answer ②: "Pour over all same net objects" has no impact on signal performance, only on the soldering of some pads with faster heat dissipation. This is beneficial to EMI and increases the contact area between the pad and copper.
Answer ③: There is no side effect of selecting "pour over all same net objects" for solid copper pouring. Solder mask defined pads should be selected instead of solid pads, because solid pads have fast heat dissipation and may cause tombstoning during reflow soldering.
47. What is a magnetic bead and what is its use? What are magnetic bead connection, inductor connection or 0-ohm resistor connection?
Answer: Magnetic beads are specially used to suppress high-frequency noise and spike interference on signal lines and power lines, and also have the ability to absorb electrostatic pulses.
Magnetic beads are used to absorb ultra-high frequency signals. Some RF circuits, PLLs, oscillation circuits, and circuits containing ultra-high frequency memory circuits (DDR SDRAM, RAMBUS, etc.) all need to add magnetic beads at the power input part. Inductors are energy storage components used in LC oscillation circuits, low and medium frequency filter circuits, etc., and their application frequency range rarely exceeds 50MHz.
The main function of magnetic beads is to eliminate RF noise existing in the transmission line structure (circuit). RF energy is an AC sine wave component superimposed on the DC transmission level—the DC component is the useful signal needed, while the RF energy is useless electromagnetic interference transmitted and radiated along the line (EMI).
To eliminate these unwanted signal energies, chip magnetic beads are used as high-frequency resistors (attenuators), which allow DC signals to pass through and filter out AC signals. Usually, high-frequency signals are above 30MHz; however, low-frequency signals are also affected by chip magnetic beads.
To select magnetic beads correctly, the following points must be noted:
- What is the frequency range of the unwanted signal?
- Who is the noise source?
- How much noise attenuation is needed?
- What are the environmental conditions (temperature, DC voltage, structural strength)?
- What are the circuit and load impedances?
- Is there space to place magnetic beads on the PCB board?
The first three points can be judged by observing the impedance-frequency curve provided by the manufacturer. Three curves in the impedance curve are very important, namely resistance, inductive reactance and total impedance. The total impedance is described by Z=√(R²+(2πfL)²).
Through this curve, select the magnetic bead model with the maximum impedance in the frequency range where noise attenuation is desired and the minimum signal attenuation at low frequencies and DC.
The impedance characteristics of chip magnetic beads will be affected under excessively high DC voltage. In addition, if the working temperature rise is too high or the external magnetic field is too large, the impedance of the magnetic beads will be adversely affected.
Reasons for using chip magnetic beads and chip inductors: The choice between chip magnetic beads and chip inductors mainly depends on the application. Chip inductors are needed in resonant circuits, while chip magnetic beads are the best choice when it is necessary to eliminate unwanted EMI noise.
48. I am engaged in hardware design work. Please ask how to determine the capacitance value to eliminate crosstalk between wires.
Answer: In PCB routing, attention should be paid to avoiding too long parallel routing, especially for high-speed or high-swing signals. If it is unavoidable, keep a sufficient distance or add a ground wire for isolation. Parts limited by volume and with high anti-interference requirements can be isolated with metal shielding enclosures.
49. In actual product development, I found a very troublesome problem. When the developed prototype is placed in a car with severe interference, to solve the freewheeling problem, a small battery is connected in parallel to the car's power supply (a diode is added to prevent the voltage of the small battery from being pulled down). However, it is found that the terminal will be interfered once connected to the car's chassis ground. Any good suggestions?
Answer: This is an obvious EMC problem. The electric spark interference in the car causes interference to your terminal equipment, which may be radiated or conducted to your terminal.
There are many reasons for this problem:
- Grounding problem: the routing of the ground wire on the terminal motherboard and the copper pouring situation;
- Shell shielding problem: it is better to use a metal shell, and seal the non-metal part of the shell with tin foil, you can try it;
- PCB layout: the power supply part and the CPU part should be separated as much as possible; the routing of the power supply part should be as thick and short as possible—routing rules are very important;
- The number of PCB layers is important: generally, the motherboard of electronic products on cars is best to have at least 4 layers; 2-layer boards may have poor anti-interference ability;
- Add a magnetic ring: you can consider putting a magnetic ring on the power line during the test.
Of course, there may be many other solutions, and the specific situation may be different. I hope it will be helpful to you.
50. Why are resistors connected in series to SCL, SDA and AS in the circuit? What impact does the resistance value have on the circuit?
Answer: Pull-up resistors increase anti-interference ability, and the general value is Vcc/1mA~10K; series resistors are used for damping, and the general value is 33ohm~470ohm. That is, when the pulse frequency on the signal line is high, it will be reflected from one end of the line to the other, which may affect data and cause EMI. Adding a series resistor in the middle of the line can effectively control this reflection.
51. If the radiation of a product is high at 200MHz and exceeds the acceptable range during CE/FCC testing, how to eliminate it and how to select magnetic beads? In addition, how to eliminate the radiation of the crystal oscillator frequency multiplication part?
Answer: The problem you mentioned is too simple to give you a very accurate answer, but based on my personal experience, here are some thinking methods.
If you are sure it is frequency multiplication, focus on processing the components that generate frequency multiplication—this should be targeted. During processing, you can directly try to simply shield the components that generate frequency multiplication (only a cola can is needed to make a shield cover, the key is to pay attention to grounding), and then test to see if the radiation value is reduced. If it is reduced, the radiation source is identified, and then special shielding processing is performed on it. If there is no change, focus on the exposed transmission lines—if the transmission lines can be grounded, they must be grounded, and it is best to try using shielded wires to see if there is any change to confirm whether it is related to the transmission lines.
Finally, there is the shielding problem of the box itself, which is relatively complex and costly, and is only considered as a solution when there is no other way. After trying these methods, the radiation value should be reduced.
52. I have recently written software for a 2KW vacuum cleaner. The function is realized, but it cannot pass EMC. Please give some guidance on which algorithm can be used in the software to pass EMC!
Function brief:
① Soft start and soft speed regulation functions (the so-called soft start means the motor accelerates slowly and the speed does not change suddenly);
② The motor speed can be adjusted;
③ The motor is controlled by a thyristor with sine wave chopping as the control method.
In terms of hardware, the circuit is very simple, and the only hardware for EMC processing is a 0.1uF safety capacitor.
Answer: Communicate with the hardware team and make more efforts—EMC can hardly be solved by software alone.
53. The DA conversion frequency in the DECODER is radiated from the chip along the power supply and ground at 166M. I connected 1nF, 630pF, or 30pF in parallel on the power supply, but none of them can filter it out. It is a 2-layer board with a short power supply loop. Please give some suggestions and analyze the reasons for the failure of filtering.
Answer ①: Poor power supply quality (load capacity)—the DA should use an independent power supply.
Answer ②: First check whether the output end is well grounded, and then try connecting a BEAD in series at the signal output port.
Answer ③: I think you can use a 100M magnetic bead to attenuate the 166M high frequency.
54. I want to do multi-channel temperature acquisition using K-type thermocouples. The power supply uses a charge pump conversion module, and the signal conditioning part wants to use AD620 and OP07 for two-stage amplification. Now I am not sure about several points:
① Power supply: I now use a 12v battery for power supply and convert it to +/-12v with a charge pump. Such a voltage has a certain ripple, which is unfavorable for signal acquisition. Should I directly use the battery voltage to make a single power supply?
② Can the two signal ends of the thermocouple be directly input to the input end of AD620 as in the example in the AD620 datasheet? I see there is also an EMI FILTER part in the datasheet—how should this part be added for measuring thermocouples? Should the cold end of the thermocouple be grounded or connected to a stable voltage?
③ Because the temperature I require involves below zero, the output of AD620 needs to be sent to the A/D port after non-inverting amplification and inverting amplification respectively. I plan to make a two-stage filter with OP07, the first stage is an infinite gain filter circuit, and the second stage is a filter circuit with 2 times non-inverting amplification and 2 times inverting amplification. I wonder if this is okay?
Answer: If the cold end of your thermocouple is grounded (one end of the thermocouple in many devices is already grounded) and the temperature measurement is below zero, it is best to use a +/- power supply, which is the usual practice.
The ripple of the power supply should be low, but it is not necessarily positive and negative symmetric—you can add a stable LDO to achieve this. Low-frequency filtering has a great impact on the results, but one-stage filtering should be sufficient; the EMI part depends on your application environment.
For multi-channel temperature measurement, you can place the multiplexer before amplification to reduce costs. The multiplexer should have differential input, and the thermocouple input wires should also be thermocouple-type, which is quite expensive.
55. Basic EMC questions: Common EMC problems encountered in certification.
Answer: The following is a summary of some problems for electronic products:
The most common problems of general electronic products are: RE (Radiated Emission), CE (Conducted Emission), ESD (Electrostatic Discharge).
Communication electronic products include not only the above three items (RE, CE, ESD), but also Surge (surge, lightning strike).
The most common problems of medical devices are: ESD (Electrostatic Discharge), EFT (Electrical Fast Transient/Burst), CS (Conducted Susceptibility), RS (Radiated Susceptibility).
For the dry northern regions, the ESD requirements for products are very high.
For regions with many thunderstorms such as Sichuan and some southwest areas, the EFT lightning protection requirements are very high.
56. How to remove electromagnetic interference in ICs?
Answer: The electromagnetic interference suffered by ICs mainly comes from electrostatic discharge (ESD). To solve the problem of ICs being immune to ESD interference, on the one hand, ESD (and EMI) issues should be considered during board layout, and on the other hand, components should be added for ESD protection.
There are currently two types of components: Varistor and Transient Voltage Suppressor (TVS).
The former is made of zinc oxide with relatively slow response speed and poor voltage suppression, and it ages after each ESD impact until it fails.
TVS is made of semiconductor with fast response speed and good voltage suppression, and can be used infinitely. From a cost perspective, the cost of varistors is lower than that of TVS.
57. Performance of electromagnetic interference: Especially for GPS applied in PMP products, which are handheld and vehicle-mounted GPS terminal products with MP4, MP3, FM radio + GPS navigation functions, an internal GPS Antenna is a must. In this way, the GPS Antenna is prone to generate electromagnetic interference with components such as MCU, SDRAM, and crystal oscillator on the GPS terminal product, resulting in a significant drop in the satellite acquisition capability of the GPS Antenna and almost making normal positioning impossible. I wonder if any GPS designers and developers have encountered such electromagnetic interference and adopted effective methods to solve it—what solutions?
Answer ①: I think this problem is mainly in the circuit design, mostly due to poor circuit protection and shielding. My current customers no longer have this confusion. They now have two types of electromagnetic interference phenomena, but they have basically been solved: 1. Bluetooth electromagnetic interference, 2. Remote control electromagnetic interference. Solutions: I haven't found the answer for the first item yet; for the second item, increase the effective distance of the remote control to 5M.
Answer ②: The distribution of each functional module on the PCB is very important. Before PCB Layer design, reasonable planning should be carried out according to the current size and the crystal frequency of each part, and then the grounding of each part is very important to solve the interference of common power supply and ground.
According to actual measurements, the spatial distance between the main oscillation sources has a great impact on radiation—a slight increase in distance can significantly reduce interference. If space is not allowed, it is necessary to perform local shielding on them, but the premise is that they are in the same grounding area on the PCB. Then decouple the power supply inlet and outlet; magnetic beads and capacitors are good choices, and printed board inductors can be used for Bluetooth and GPS.
The selection of the conversion frequency of the power supply DC/DC is also very important—do not let the frequency multiplication (multiple harmonics) coincide with the frequency of other circuits (especially the receiving circuit). Some DC/DC frequencies are fixed, and a simple filter circuit can be added. Same-frequency suppression is the main reason for the decrease in the receiving sensitivity of GPS and remote control.
In addition, the local oscillator amplitude of the receiving circuit should be adjusted as small as possible, otherwise it will become a continuous interference source. We have integrated Bluetooth, GPS receiving, another 2.4GHz transceiver, and 433M remote control receiving into one box with good results and high GPS receiving sensitivity.
58. Encounter a single-chip microcomputer system:
① Main control chip: Motorola MC908JL3
② 8M ceramic resonator
③ Power supply connected by wires
Now the conducted voltage in EMI exceeds the standard by 0.8dB at a single point of 24M. Please give some guidance on good methods to suppress the over-standard, such as adding magnetic rings, adding Y2 capacitors, etc. In addition, is this frequency in the conduction range or radiation range?
Answer: It is unclear whether the 24M exceeds the standard in the EMI test or in the conduction test—if it is the former, it is radiation over-standard; if it is the latter, it is conduction over-standard.
59. A bidirectional thyristor is used to control the speed of a DC motor, but the motor interferes with the power supply and affects the zero-crossing detection, resulting in loss of control or speed change. Please give some guidance!
Answer ①: The possible causes of this phenomenon are: 1. The motor is a non-resistive load, so phase shift occurs in the circuit, leading to inaccurate control—capacitors can be added for filtering; 2. Generally, bidirectional thyristors control high-power or large-current loads with zero-crossing conduction instead of phase modulation, which can reduce the impact of EMC.
Answer ②: Current phase-shift speed regulation is very commonly used. If the hardware part of zero-crossing detection is no problem, the software processing method should be carefully improved. The conduction of the thyristor should be processed twice in one cycle (50Hz 20mS), and the delay output time after detecting zero-crossing determines your phase shift angle.
60. Has anyone done EMC design for V.35, E1, G.703 (64K), and relay interfaces? Can you give some suggestions?
Main standards to comply with:
GB/T 17626.12 (IEC61000-4-12) Electromagnetic compatibility - Testing and measurement techniques - Oscillatory wave immunity test
GB/T17626.2 (IEC61000-4-2) Electromagnetic compatibility - Testing and measurement techniques - Electrostatic discharge immunity test
GB/T 17626.3 (IEC61000-4-3) Electromagnetic compatibility - Testing and measurement techniques - Radiated, radio-frequency, electromagnetic field immunity test
GB/T 17626.4 (IEC61000-4-4) Electromagnetic compatibility - Testing and measurement techniques - Electrical fast transient/burst immunity test
GB/T 17626.5 (IEC61000-4-5) Electromagnetic compatibility - Testing and measurement techniques - Surge immunity test
GB/T 17626.6 (IEC61000-4-6) Electromagnetic compatibility - Testing and measurement techniques - Immunity to conducted disturbances, induced by radio-frequency fields
Answer: These standards are all basic EMC test standards, and specific indicators need to be determined in combination with your product. These interfaces of yours are communication interfaces, and there are generally standard circuits.
When the filter design of the single board schematic and the correct layout and routing design of the PCB are done well, the test can generally be passed. In other cases, it is necessary to add EMC filtering and transient suppression components, which needs to be analyzed in combination with specific interfaces.
61. Routing cannot cross the gap between divided power supplies. Can anyone give a detailed explanation?
Answer: If a power layer is divided into several different power supply parts (e.g., 3.3V, 5V power supplies), the signal lines should not appear on different power planes at the same time, that is, the routing cannot cross the gap between the divided power supplies, otherwise unnecessary EMC problems will occur. The same applies to the ground—routing cannot cross the gap between divided grounds.
62. Currently, a single-chip microcomputer controls the pull-in of a 12V relay through a Darlington transistor and an optocoupler, and then controls the pull-in of an AC contactor. The single-chip microcomputer often resets at the moment of pull-in. An effective reset signal is detected at the reset pin through an oscilloscope (a three-pin reset IC is used). The single-chip microcomputer is powered by 5V, and 1000uF capacitors are connected before and after the 5V zener diode, and no power supply fluctuation is detected through an oscilloscope. In addition, no reset phenomenon is found if the relay is no-load (not connected to the AC contactor). Please ask how to solve this problem?
Answer ①: A resistance-capacitance absorption circuit with a resistor and a capacitor connected in series can be connected in parallel at both ends of the AC contactor coil. The capacitance is between 0.01UF and 0.47UF, and the withstand voltage is preferably 2-3 times the rated voltage of the coil. Let's see if this works?
Answer ②: This should be caused by EMC interference generated when the AC contactor acts. The resistance-capacitance absorption method from the above friend is a good solution. At the same time, you can also consider connecting a 100P to 47P high-voltage capacitor in parallel at the output contact of the 12V relay and try.
Answer ③: Adding RC absorption to the AC contactor is effective, but you also have to check your power supply loop to see if the CPU power supply routing is too long. Try to connect decoupling capacitors in parallel at the power pins of the chip, and the voltage stabilization part can also add an LC absorption loop to absorb interference from the power supply as much as possible.
Answer ④: First, check whether the same phenomenon occurs without a load and judge and eliminate the problem step by step. You can first disconnect the optocoupler, then the relay. If the reset still occurs when the optocoupler is disconnected, check whether there is a short circuit between the hardware output port and the reset. If there is no reset, connect the optocoupler but not the relay.
If the reset still occurs, the possible situation is that the ground wire is too thin, the ground of the reset pin is too close to the optocoupler and far from the power supply, and the current-limiting resistor of the optocoupler is too small, leading to an instantaneous rise in ground potential. During routing, the CPU should be kept away from high-current devices, and the ground wire should adopt star single-point grounding. If the reset still occurs, it is the electromagnetic interference caused by the relay coil, contact arcing or large load changes.
Shielding and methods to eliminate contact arcing can be adopted to solve it. In most cases, the power supply is not handled well, and the ground wire or +5V wire is too long and thin, or the CPU position is unreasonable.
63. Can AC filters and DC filters be used interchangeably? Generally speaking, AC line filters can be used in DC occasions, but DC line filters must not be used in AC occasions. Why?
Answer: The bypass capacitor used in the DC filter is a DC capacitor, which may overheat and be damaged when used in AC conditions; if the withstand voltage of the DC capacitor is low, it will also be broken down and damaged.
Even if these two situations do not occur, the capacitance of the common-mode bypass capacitor in general DC filters is large, and using it in AC occasions will cause excessive leakage current, violating the provisions of safety standards.
64. In a box-type device such as an Ethernet switch or a PC, there are chassis ground and circuit working ground. I found that some devices connect the two grounds with capacitors, some with 0-ohm resistors, and some with ferrites. Which one is correct?
Answer: We generally use 102 high-voltage ceramic capacitors.
65. What does "mechanical protection" refer to? Is it the protection of the case?
Answer: Yes, the case should be as tight as possible, use less or no conductive materials, and be grounded as much as possible.
66. What is the great impact of using all metal as the shell (such as aluminum, stainless steel and other materials) on the ESD protection of the product? How to handle it better?
Answer: If a product uses an all-metal shell, poor grounding is certainly not conducive to ESD protection, but there will be no problem as long as the grounding is done well. As for how to ground, it depends on the specific situation of the equipment. If it is a large-scale equipment, it can be directly grounded through the equipment, and the effect is certainly very ideal.
67. Why can a spectrum analyzer not observe transient interferences such as electrostatic discharge?
Answer: Because a spectrum analyzer is a narrow-band sweep receiver that only receives energy in a certain frequency range at a certain moment.
Transient interferences such as electrostatic discharge are pulse interferences with a wide frequency spectrum but short duration. In this way, what the spectrum analyzer observes when the transient interference occurs is only a small part of its total energy, which cannot reflect the actual interference situation.
68. In on-site diagnosis of electromagnetic interference problems, a near-field probe and a spectrum analyzer are often needed. How to make a simple near-field probe with a coaxial cable?
Answer: Strip off the outer layer (shielding layer) of the coaxial cable to expose the core wire, wind the core wire into a small ring with a diameter of 1~2 cm (1~3 turns), and weld it to the outer layer.
69. Measuring the biomagnetic information of the human body is a new medical diagnosis method. The measurement of this biomagnetism must be carried out in a magnetic field shielding room, which must be able to shield the alternating electromagnetic field from the static magnetic field to 1GHz. Please propose a design scheme for this shielding room.
Answer: First, consider the selection of shielding materials. Since it is necessary to shield magnetic fields with very low frequencies, materials with high magnetic permeability such as permalloy should be used. Since the magnetic permeability of permalloy will decrease after processing, heat treatment must be carried out.
Therefore, the shielding room should be made of assembled type, composed of plates. Each plate is processed according to the design in advance, then heat-treated, transported to the site, and installed very carefully. The joints of each plate should be overlapped to form a continuous magnetic path.
The shielding room constructed in this way can have a good shielding effect on low-frequency magnetic fields, but the gaps will cause high-frequency leakage. To make up for this deficiency, a second layer of shielding is welded with aluminum plates on the outer layer of the permalloy shielding room to shield the high-frequency electromagnetic field.
70. What factors are considered when selecting shielding materials when designing a shielded chassis?
Answer: From the perspective of electromagnetic shielding, the main consideration is the type of electric field wave to be shielded. For electric field waves, plane waves or magnetic field waves with higher frequencies, general metals can meet the requirements; for magnetic field waves with low frequencies, materials with higher magnetic permeability should be used.
About Maxipcb
Maxipcb empowers innovators to turn cutting-edge technologies into reality.
We offer one-stop solutions for design, simulation, testing, PCB manufacturing, component procurement and SMT assembly, enabling efficient development, rapid deployment and risk control across the full product lifecycle.Serving the world in communications, industrial automation, aerospace, automotive, semiconductor and beyond, we build a safer, more connected future together.
