Introduction to Electromagnetic Compatibility Design Rules for Electronic Circuits

1.1 Electronic Circuit Design Guidelines

Electronic circuit designers often only consider the function of the product, but do not consider the function and electromagnetic compatibility. Therefore, the product also generates a large amount of functional harassment and other harassment while completing its functions. Moreover, the sensitivity requirements cannot be met.

The electromagnetic compatibility design of electronic circuits should be considered from the following aspects:

1.1.1 Component selection

In most cases, the extent to which the basic components of the circuit meet the electromagnetic characteristics will determine the extent to which the functional unit and the final device meet electromagnetic compatibility. The main criteria for choosing the right electromagnetic component include out-of-band characteristics and circuit assembly techniques. Because the realization of electromagnetic compatibility is often determined by the component response characteristics away from the fundamental frequency. In many cases, the circuit assembly determines the degree of out-of-band response (such as lead length) and the coupling of different circuit components. The specific rules are:

(1) At high frequencies, it is preferable to use a feedthrough capacitor or a pedestal capacitor with a small lead inductance for filtering in comparison with a lead type capacitor.

(2) When a leaded capacitor must be used, the influence of the lead inductance on the filtering efficiency should be considered.

(3) Aluminum electrolytic capacitors may undergo temporary dielectric breakdown for a few microseconds, so solid capacitors should be used in circuits with large ripple or transient voltages.

(4) A resistor with a parasitic inductance and a small capacitance is used. Chip resistors can be used in ultra-high frequency bands.

(5) The large inductance parasitic capacitance is large. In order to increase the insertion loss of the low frequency part, do not use a single-section filter, but a multi-section filter composed of several small inductances should be used.

(6) Pay attention to the saturation characteristics when using the core inductor. Pay special attention to the high-level pulse to reduce the inductance of the core inductor and the insertion loss in the filter circuit.

(7) Use shielded relays as much as possible and ground the shield case.

(8) Select an input transformer that is effectively shielded and isolated.

(9) The power transformer used for sensitive circuits should be electrostatically shielded, and both the shielded case and the transformer case should be grounded.

(10) Interconnected signal lines inside the device must use shielded wires to prevent disturbance coupling between them.

(11) In order to connect each shield to its own pin, a plug with a sufficient number of pins should be used.

1.1.2 Electrical design

Each unit can be described as receiving an input signal, processing the input signal, and then outputting the processed signal at the output. Consideration must be given to unwanted signals that may be present at the input, as well as unwanted signals entering through other paths than the input. It is best to process these unwanted signals at the input point. Power supply

The EMI coupling of the device power supply involves sensitivity to conducted emissions on the power supply line (mains harmonics, differential mode or common mode transients, narrowband signals from the radio transmitter) and emissions conducted to the supply line. The power supply in the device is widely connected to other functions. On the one hand, the unwanted signals generated in the power supply can be easily coupled to the functional units. On the other hand, the unwanted signals in one unit may be coupled through the common impedance of the power supply. Go to other units. Therefore, from the viewpoint of electromagnetic compatibility, we must first care about the power supply.

(1) Powering each functional unit separately under possible conditions.

(2) All circuits using the public power source are as close as possible to each other.

(3) All circuits using a common power supply must be compatible with each other.

(4) The power supply filter should be used on the AC and DC mains to prevent external disturbances from entering the equipment through the power supply, preventing switching transients and other signals generated inside the equipment from entering the primary power supply.

(5) The input and output lines of the power supply and the input and output lines of the filter are effectively isolated.

(6) Effective electromagnetic field shielding of the power supply, especially the switching power supply.

(7) Switching power supplies can cause high-frequency radiation and conducted disturbances, but it also has the advantage of repelling power line transients (typical regulators cannot).

(8) The rectifier diode should operate at the lowest current density (proportional to the maximum rated current).

(9) The power supply should maintain a low output impedance for all circuit function states. Even in the RF range, the output capacitor should exhibit low impedance.

(10) Ensure that the regulator has a fast enough response time to suppress high frequency ripple and transient loading.

(11) Provide sufficient RF bypass for the Zener diode.

(12) Reasonably shield and carefully isolate the high voltage power supply from the sensitive circuit.

(13) The power transformer should be symmetrically balanced and should not be power balanced.

(14) For the core material used in the transformer, the lower limit of the saturation magnetic induction Bm should be taken. In any case, it must be ensured that the core is not driven to saturation.

(15) The core structure of the transformer should be preferably D type and C type, and E type is the second most.

(16) The electrostatic shielded power transformer suppresses common mode disturbance on the power line, and the multiple shielded isolation transformer (super-isolated) has better performance. Control unit

(1) The control unit and the device body are often far apart, so the grounding and shielding methods must be properly applied to prevent the formation of ground loops and coupling unwanted signals.

(2) The main unwanted sources in the control unit are those that can suddenly disconnect the control signal path. Such as switches, relays, thyristors, switching diodes, etc.

(3) When various switches that generate unwanted signals are operated together with inductive loads, serious transients occur.

(4) Minimize the steep wavefront transient process and limit the inrush current through the switch when switching on and off.

(5) If necessary, use an RC network or diode to suppress switching transients.

(6) If necessary, use a buffer or damper to reduce the vibration of the relay contacts. Amplifier

Due to their wide range of applications, which can affect the generation and coupling of unwanted signals, strict electromagnetic compatibility design requirements must be imposed on the amplifier.

(1) The layout of the amplifier should be designed to transmit a low level signal at the shortest distance, otherwise it is easy to introduce disturbance.

(2) The amplifier's occupied bandwidth should match the wanted signal. The out-of-band response of the amplifier must be controlled. An excessive bandwidth can easily amplify unwanted signals or create parasitic oscillations.

(3) Pay attention to the decoupling between the stages of the multistage amplifier.

(4) Decouple the inputs of all amplifiers and only let the useful signals enter the amplifier.

(5) For amplifiers with a working frequency lower than 1MHz, it is better to use a balanced input type (especially an audio amplifier).

(6) The noise of the operational amplifier is higher than the noise level of the transistor, which is 21/2 times or more.

(7) The power supply of the instantaneous high current load should be separated from the power supply of the operational amplifier to prevent the instantaneous undervoltage state of the operational amplifier power line.

(8) The input transformer of the isolation amplifier should be effectively shielded from isolation between the primary and secondary.

(9) Use an input transformer to disconnect any ground loop to the far end audio input circuit.

(10) The audio input transformer should be magnetically shielded to avoid picking up power supply magnetic field disturbances.

(11) The audio amplifier should use a balanced input type and use a shielded twisted pair as the input signal line.

(12) The audio gain (volume) control should be after the high gain preamplifier, otherwise the noise and disturbance pickup levels on its trace will become a significant portion of the low level input signal.

(13) If the audio amplifier uses a switching power supply, it must use a switching speed of 20KHz or higher. Digital Circuit

Digital and analog devices have different transmit and sensitivity characteristics, and generally cannot be used to implement analog circuit electromagnetic compatibility by filtering digital signals. For example, narrowband harassment is often produced and is often sensitive to continuous wave disturbances; digital circuits often generate broadband nuisance and are sensitive to spike nuisance. The extent and extent of shielding and filtering used to control the transmission and sensitivity of digital circuits is determined by the performance of the digital circuit unit and the rate of the circuit components.

Most of the important reasons for the malfunction of the digital system are caused by the fluctuation of the potential of the casing ground and the signal ground. When the potential of the 0V terminal of the integrated circuit changes, its working state is unstable, which affects the condition of the input of the next stage, and the next stage is also unstable. The change in the 0V line potential is caused by the inductance of the ground line itself and the DC resistance.

(1) The slowest rise and fall times allowed by the circuit function must be selected to limit the generation of unnecessary high frequency components.

(2) Avoid generating and using unnecessary high logic levels. Do not use the 12V level if you can use the 5V level.

(3) The clock frequency should be the lowest selected under the conditions allowed by the work.

(4) To prevent the data pulse from being coupled to the DC power bus through the filter and the secondary regulated power supply.

(5) The input and output lines of the digital circuit should not be close to the electromagnetic hot wire such as the clock or the oscillator line or the power line, and should not be close to the fragile signal lines such as the reset line, the interrupt line and the control line.

(6) Whenever possible, connect the input and output of the digital circuit at a low impedance point, or use an impedance conversion buffer stage.

(7) Strictly limit the spikes, overshoots, and damped oscillations of the pulse waveform.

(8) If a pulse transformer is used, it should be shielded.

(9) The power and control lines must be decoupled to prevent external disturbances from entering.

(10) Do not use long, unshielded signal lines. The length of the trace is about 5 cm per ns rise time to consider matching terminations.

(11) It is noted that the opto-isolator has an inhibitory effect on the differential mode disturbance, but has no significant effect on the common mode disturbance.

(12) The inductance component of the printed conductor plays a leading role in generating common impedance coupling. The power cord, especially the lines should be as thick and short as possible.

(13) For devices with transient steep supply currents and devices susceptible to power supply noise, decouple the capacitors with high frequency characteristics near them.

(14) Install an LCL-shaped T-type filter at the power inlet of each printed board to prevent shock input from the power supply.

(15) Improve the anti-harassing performance of the flat cable with a shielded mesh (woven tape) and a ferrite clamp.

(16) It is easy to increase the emission and immunity performance by 10 times from a 2-layer printed circuit board to a multilayer printed circuit board.

(17) The “five-five” rules can help you make decisions. That is, the clock frequency is greater than 5MHz or the pulse rise time is less than 5ns, which is suitable for selecting a multi-layer circuit board.

(18) Use manual cloth key lines (clock, high-speed repeat control signal, reset line, trunk line, I/O line, etc.). If automatic routing is used, the EMI control must be carefully checked and modified. Other

(1) Decoupling

The measure to eliminate the harmful effects of common impedance coupling is decoupling. The key components of the decoupling filter are high frequency capacitors with as short a lead as possible.

(2) Isolation

1 Note that the ground loop forms common mode disturbance.

2 Use an isolation transformer to cut the ground loop, which is most suitable when the signal does not contain DC components. Broadband signals should not be used. In the industrial field, a signal containing a direct current component is modulated into an alternating current signal, which is sent to a receiving end via a voltage or current transformer for demodulation. A non-ideal transformer has a distributed capacitance between the primary and secondary, which allows the disturbance to be coupled via the transformer, and thus the size of the distributed capacitance directly affects its high frequency isolation performance. That is to say, the distributed capacitance provides a channel for the signal to enter the power grid. Therefore, when selecting a transformer, the size of the distributed capacitor must be considered. When using the transformer, it is necessary to add electrostatic shielding (Faraday shielding) and grounding, which can reduce the distribution parameters, because the electrostatic shielding destroys the direct coupling of the primary and secondary questions, and the conduction disturbance can be reduced.

In order to better reduce the distributed capacitance and improve the common mode rejection performance of the switching transformer, a three-layer shielding can be used: the first layer shield is connected to the primary potential end; the second layer shield is connected to the secondary low potential end, the central Faraday shield Connect to the outer casing of the transformer and securely.

3 optocoupler isolation method.

Due to the poor linear relationship between input and output, it should not be used directly for analog signals, but it is best for transmitting digital signals. With the optical pulse width modulation method, an analog signal containing a direct current component can be transmitted, and an excellent linear effect is obtained.

(3) Ways to improve the ability to resist common mode disturbance

It is sometimes difficult to cut the ground loop with an isolation device, for example, two devices must be connected in a direct current. At this time, only measures can be taken to minimize the effects of common mode disturbance caused by the ground loop.

1 differential amplifier

DC to high frequency, good linearity, suitable for analog signals. Common mode rejection is good when symmetrically balanced. In the case of imbalance, common mode disturbance is converted into differential mode, and the degree of influence is related to the degree of imbalance.

2 series common mode choke (neutralization transformer or longitudinal choke)

1.2 Printed Circuit Design Guidelines

In the design of printed circuit boards, product designers tend to focus only on increasing density, reducing space, making simple, or pursuing aesthetics, uniform layout, ignoring the influence of line layout on electromagnetic compatibility, and allowing a large amount of signals to radiate into space. Form harassment.

When designing printed circuit boards, the following points should be noted:

(1) From the perspective of reducing radiation disturbance, the multi-layer board should be selected as much as possible. The inner layer is used as the power layer and the ground layer respectively to reduce the impedance of the power supply line, suppress the common impedance noise, and form a uniform ground plane for the signal line. The distributed capacitance between the large signal line and the ground plane suppresses its ability to radiate into space.

(2) The power line, ground line, and printed board trace should maintain low impedance to high frequency signals. At very high frequencies, power lines, ground lines, or printed circuit board traces can become small antennas for receiving and transmitting disturbances. In addition to adding filter capacitors, the method of reducing such disturbances is more important to reduce the high-frequency impedance of the power lines, ground lines, and other printed circuit board traces themselves. Therefore, the traces of various printed boards should be short and thick, and the lines should be uniform.

(3) The power line, ground wire and printed conductor should be arranged on the printed board properly, so as to be short and straight as possible to reduce the loop area formed between the signal line and the return line.

(4) The layout of circuit components and signal paths must minimize the mutual coupling of unwanted signals.

1 The low electronic signal path cannot be close to the high level signal path and the unfiltered power line, including circuits that generate transients.

2 Separate the low-level analog circuit from the digital circuit to avoid common impedance coupling between the analog circuit, the digital circuit, and the power supply common return line.

3 high, medium and low speed logic circuits use different areas on the PCB.

4 When arranging the circuit, the signal line length should be minimized.

5 Ensure that there are no long parallel lines between adjacent plates.

The 6EMI filter should be placed as close as possible to the EMI source and placed on the same board.

The 7DC/DC converter, switching components and rectifiers should be placed as close as possible to the transformer to minimize the length of the conductor.

8 Place the voltage regulator and filter capacitor as close as possible to the rectifier diode.

1.3 Equipment internal routing guidelines

Some products, especially those that are not very formal, have very messy internal wiring, and various wirings are bundled together without any shielding, filtering or grounding measures. This internal routing processing method not only causes mutual interference between the wires transmitting the high and low level signals, but also brings inconvenience to the later remedial measures such as screen filtering. Proper wiring is also an electromagnetic compatibility design measure that greatly reduces harassment and does not require additional procedures, but can achieve satisfactory results. Therefore, when wiring, you should:

(1) All bare traces in the chassis should be as short as possible.

(2) The wires that transmit different electronic signals are bundled, and the digital signal lines and analog signal lines should also be bundled and kept at an appropriate distance to reduce the mutual influence between the wires.

(3) For flat strip lines, which are often used to transmit signals, the ground--signal-ground-signal-ground arrangement should be used, which not only can effectively suppress the disturbance, but also significantly improve the immunity.

(4) The low frequency incoming and return wires are twisted together to form a twisted pair, so that the disturbance currents existing between the two wires are almost equal in magnitude, and the opposite directions, the disturbance fields can cancel each other in space, thereby reducing the disturbance.

(5) Shield the wires that can be determined and have large radiation disturbances.

(6) Separation of functional units and circuits within the device limits unwanted signals to a limited range to effectively decouple unwanted signals and potentially sensitive circuits and wires.

1 Use a modular structure (functional unit with a shielded enclosure) where possible.

2 In particular, the power line filter, high level signal circuit, and low level signal circuit should be placed in different shielding compartments.

3 Shields, such as plates or partitions, are used inside the device to separate the high-level source and the sensitive receiver.

4 provides effective electrical and magnetic field shielding for the power supply. Especially for switching power supplies.

5 Reasonably shield the high voltage power supply and isolate it from sensitive circuits.

6 Use magnetic shielding around the entire audio-sensitive circuit to reduce coupling to the power line. This method can be used to effectively reduce the 400 Hz/50 Hz hum. The input circuit is differential, and the input signal is twisted pair.

For example, in the display, the socket of the AC power cord is generally on the rear panel, and the power switch is often on the front panel, so that the power supply wiring in the machine is very long, and many manufacturers do not take corresponding measures for this situation. Such as shielded wire or twisted pair. This will cause the internal wiring to receive the working signal and conduct it through the power line.

For example, in a microcomputer, although the power supply is shielded, the DC output line of the power supply is outside the shield. If the DC output line is too long, it is easy to receive the disturbance on the main board and transmit it to the AC power line. Therefore, the length of the DC output line should be minimized during design. Alternatively, a magnetic bead or a ferrite bead can be added to the DC output line.

1. 4 floor and casing design guidelines

The structural design of the base plate and the casing, ie the structural materials and assembly techniques, often determines whether EMC can be achieved with the work environment.

The backplane and chassis are the most efficient way to provide shielding for unwanted signal paths in control devices or functional units. The degree of shielding depends on both the choice of structural materials and the design techniques used in the assembly. Designed shields are limited only by the designer's knowledge and skill in designing seams, openings, penetrations, and lap joints to the backplane and chassis.

1.4.1 Shielding

Shielding is a matter of dealing with the field. The nature of the field is different in areas where the distance from the field source is different. A critical distance is d.


(1) Field division

Roughly divided: the region of d<λ/2π is the near-field region

The region of d>λ/2π is the far field region

Strictly divided: d<d0/3, that is, the region of d<λ/20 is the near-field region (but can be extended to d<λ/1.2π)

The area where d>3 d0=λ/2 is the far field area

The region of λ/20<d<λ/2 is the transition zone

(2) Field properties

The near field is the induction field—the electrostatic field and the static magnetic field, and does not radiate energy to the outside. (Faraday shielding treatment)

The far field is the radiation field—the E and H vectors radiate energy outward in phase with each other in time. (Chassis shielding processing, especially the handling of electrical continuity problems)

The transition zone is an inductive electromagnetic field—the nature of the field is complex.

(3) The problem is mainly near-field inside the equipment.

It is complicated and not practical to use the field theory to solve Maxwell's equations, so it is treated by approximation circuit theory. That is, the coupling of the electric field is investigated by using the lumped parameter capacitance, and the coupling caused by the magnetic field is investigated by the parameter of the mutual inductance concentration.

1 If the voltage of the wave source is high and the current is small, the effect of the electric field is more obvious than that of the magnetic field, and electric field shielding can be used.

2 If the voltage of the wave source is low and the current is large, the magnetic field plays a leading role and magnetic field shielding should be used.

Electric field shielding - Faraday shielding to eliminate the effects of the electric field.

The magnetic field shield distorts the magnetic flux of the loop 1 or directs it to the other side, avoiding the interconnection with the loop 2 to eliminate the magnetic field coupling.

(4) Shielding of electromagnetic fields - electrical continuous closing of the metal box against the far field and near field.

1.4.2 structural materials

(1) Most materials suitable for the base plate and the casing are good conductors that can shield electric fields such as aluminum and copper. The primary shielding mechanism is the reflected signal rather than the absorption.

(2) The shielding of the magnetic field requires ferromagnetic materials such as high permeability alloys and iron. The main shielding mechanism is absorption rather than reflection.

(3) In a strong electromagnetic environment, the material is required to shield both the electric field and the magnetic field, and thus a structurally intact ferromagnetic material is required. Shielding effectiveness is directly affected by material thickness and the quality of the lap and grounding methods.

(4) The plastic casing may be coated with a shielding layer on its inner wall or metal fiber during injection molding.

1.4.3 gap

The electrical discontinuity of the structure must be minimized in order to control the leakage of radiation through the bottom plate and the casing. Structural measures to improve the effectiveness of the gap shielding include increasing the depth of the gap, reducing the length of the gap, adding a conductive gasket to the joint surface, applying a conductive coating to the joint, and shortening the screw pitch.

(1) As far as possible, each joint and discontinuity of the bottom plate and the casing should be overlapped as much as possible. The worst electrical lap joints play a decisive role in reducing the shielding effectiveness of the housing.

(2) Ensure metal-to-metal contact at the joint to prevent leakage and radiation of electromagnetic energy.

(3) Where possible, the joints shall be welded so that the joint faces are continuous. In the case of limited conditions, spot welding, small pitch riveting and screw connections can be used.

(4) When no conductive pad is added, the screw pitch should generally be less than 1% of the maximum operating frequency, at least not greater than 1/20 wavelength.

(5) When lap jointing with screws or riveting, first overlap the middle part of the joint and then gradually extend to both ends to prevent the metal surface from bending.

(6) Ensure that the fastening method has sufficient pressure to maintain surface contact in the presence of deformation stress, shock, and vibration.

(7) Where the seam is not flat, a conductive gasket or finger spring material must be used at the movable panel or the like.

(8) Select a liner with high conductivity and elasticity. When selecting a gasket, consider the frequency at which the joint is used.

(9) Select a liner made of a hard and tough material.

(10) Ensure that the metal surface mated with the gasket has no non-conductive protective layer.

(11) When active contact is required, use a finger-shaped compression spring (without a mesh pad) and pay attention to maintaining the pressure of the elastic finger spring.

(12) When the conductive rubber gasket is used on the surface of aluminum metal, attention should be paid to galvanic corrosion. Rubber or Monel wire liners in sterling silver will exhibit severe electrochemical corrosion. The conductive rubber of the silver-plated aluminum filler is the best gasket material for the aluminum metal mating surface in the fog salt environment.

1.4.4 penetration and opening

(1) Pay attention to the extent to which the overall shielding effectiveness is reduced due to the passage of the cable through the casing. The shielding effectiveness is reduced by more than 30 dB when a typical unfiltered wire passes through the shield.

(2) When the power cord enters the casing, all should pass through the filter box. Preferably, the input end of the filter can be worn out of the shield case;

If the filter structure is not suitable for wearing the enclosure, a compartment should be provided for the filter at the power cord into the enclosure.

(3) Signal line, when the control line enters/passes through the casing, it must pass the appropriate filter. A multi-core connector (socket) with filter pins is suitable for this application.

(4) The metal control shaft that passes through the shielded housing should be grounded with a metal contact, grounding nut or RF pad. It is also possible to use a non-grounded metal shaft and use other insulating shafts to pass through a circular tube whose waveguide cutoff frequency is higher than the operating frequency as a control shaft.

(5) It must be noted that the penetration of the metal shaft or wire in the cut-off waveguide hole will seriously reduce the shielding effectiveness.

(6) When it is required to use a metal-controlled shaft that is insulated from the ground, a short recessive control shaft can be used, and if it is not adjusted, it can be covered with a nut or a metal gasket elastic mounting cap.

(7) Add a metal cap to the fuse, jack, etc.

(8) Use a conductive gasket and a washer, a nut, etc. to achieve a leak-proof installation of the toggle switch.

(9) Shield the vents with honeycomb panels when the shielding, ventilation and strength requirements are not critical. It is best to keep the connection by soldering to prevent leakage.

(10) Apply shielding to the indicator and display as much as possible, and filter all leads with a feedthrough capacitor.

(11) When the indicator/display and the lead are not filtered from the rear, use a metal mesh or conductive glass that is continuously connected to the chassis to shield the front of the indicator/display. The shielding glass for the clamped wire can achieve a shielding performance of 50~110dB for 30~1000MHz under the condition of maintaining reasonable transmittance. A transparent conductive film is plated on transparent plastic or glass, and the shielding effect is generally not more than 20 dB. However, the latter eliminates the accumulation of static electricity on the viewing window and is commonly used on instruments.

The shield should have the following three elements to be shielded:

a. The shield is a complete electrical continuum;

b. There are perfect filtering measures;

c. There must be good grounding for electrical shielding.

Taking microcomputer products as an example, due to its special structure, it is really difficult to obtain good shielding performance. The main factors for the unsatisfactory shielding effectiveness of microcomputer products are:

1 The power device, the switching device and the signal line with sudden change of current generated in the microcomputer system are not filtered and shielded, so that the internal disturbance field of the casing is large.

2 Many microcomputers are plastic casings, the surface is not coated with conductive materials, or coated, but the coating performance is not good, the shielding effectiveness is very low.

3 Microcomputer casings have many holes and holes due to ventilation holes, installation switches and other components. Since there is no special treatment between the upper and lower covers and the side plates, the contact is not very good, resulting in the chassis itself not being an electrical continuum, thus affecting Shield performance.

4 improper filtering of the incoming and outgoing lines of the power supply is also a factor affecting the shielding effectiveness.

The factors affecting the shielding effectiveness are not incapable of elimination, but efforts should be made, such as improving the performance of conductive coatings, rationally arranging the positions of holes and slits, and opening directions, installing filter connectors, shielding copper meshes and conductive gaskets to improve the assembly process. Level. In short, solving this problem requires the attention of the company and the efforts of the designers.

1.4.5 lap

Lap is the process of mechanically connecting certain metal parts together, in order to achieve low-resistance electrical contact, to ensure the stability of the system's electrical performance, and to help achieve suppression of radio frequency disturbance.

(1) Use the same metal as much as possible.

(2) Ensure that the DC resistance of the lap is not more than 25 milliohms. Do not use an ohmmeter to evaluate RF straps or RF washers.

(3) Pay attention to the relative positions of various metals in the electrochemical sequence table by lapping different metals. The potential difference should be as small as possible and have appropriate anti-corrosion measures.

(4) Trimming the lap surface to obtain the maximum contact area. Apply a protective layer immediately after lap jointing.

(5) Clean all mating surfaces before lap joints. To prevent oxidation, the mating surface is overlapped after the protective layer is removed.

(6) For permanent lap joints, all joints should be joined by welding or brazing or soldering. RF lap joints should preferably be permanently lapped.

(7) It is not allowed to use the thread of bolt or screw to complete the RF bonding.

(8) Conductive paint is not allowed to be used for electrical or RF bonding. The conductive adhesive joint must provide a pressure of approximately 700 g/cm2 to ensure high electrical conductivity at the conductive coating. The conductivity of the conductive paste is required to be approximately 2 to 5 mΩ/cm.

(9) Compact all RF pads.

1. 5 wiring design guidelines

Wiring is the guideline and cable arrangement. Wiring actually includes a series of separation, isolation, classification and cable placement.

1.5.1 Cable Connector

The connection of the cable can degrade the performance of the electronic/electrical subsystem. Not only because external harassment signals can interact with or couple into the connection cable in the system/subsystem, it poses a serious threat to sensitive equipment; it can also cause problems due to improper design, classification (isolation), strapping, and routing.

(1) Cables should be avoided in the field as much as possible; replacement cables tested or inspected by the production unit should be used.

(2) The connecting cable inside the equipment compartment is difficult to replace. To this end, an appropriate safety margin should be determined to allow for the life of the system.

The performance of the connecting cable has deteriorated.

(3) Special attention should be paid to connectors for low-level signals and low-impedance circuits, as well as connectors that can cause errors due to increased impedance.

(4) The design of the connecting cables and connectors between the subsystems should be consistent. (For example, one end cannot be required to have all of its shields separated from each other, while the other end only has one pin for one connector for the shield to terminate. No shield wire can be used to control the disturbance radiation, and the other end is non-conductive. Layer connector.)

(5) Do not let the main power and signal wires pass through the same connector.

(6) Try not to let the input and output signal lines pass through the same connector.

(7) According to the wire classification, the connector shield is terminated correctly.

1.5.2 Wire classification and bundle

A major aspect of EMI control is the division of wires and cables into levels similar to those of processing power electronics. The classification table grouped by 30dB power level is shown in Table 2:

Table 2 cable bundle classification

Category Power Range Features

A 》40dBm high power DC, AC and RF (EMI) source

B 10~40dBm low power DC, AC and RF (EMI) sources

C -20~10dBm pulse and digital circuit source Video output circuit (audio, video source)

D -50~-20dBm Audio and sensor sensitive circuit Video input circuit (audio sensitive circuit)

E -80~-50dBm RF, IF input circuit, safety circuit (RF sensitive circuit)

F "-80dBm antenna and RF circuit (RF sensitive circuit)

The benefits of this classification are:

1 EMI source and receiver are respectively classified by power

2 In the same harness or wire tie, the power levels of adjacent wires do not differ by more than 30 dB.

1.5.3 Wire marking for laying cables

1 Mark the mark at each end of the wire from the joint or the connected device no more than 15 cm, and the mark spacing on each line is 40 cm.

2 When the actual strapping, the wires with the same mark can be bundled in the same harness. It is not possible to bundle different marked wires in the same harness without the approval of the person in charge of EMI control.

1.5.4 Shield termination

(1) Shielded wire

1 Shielded wires are used to prevent unwanted radiation or to protect the wires from stray fields.

2 Isolation of the shield to prevent unnecessary grounding.

3 Do not use the shield for signal return.

4 twisted pair has similar electromagnetic shielding effect.

(2) Protection of sensitive circuits

1 The shield used to protect the audio-sensitive circuit is grounded only at one end. Never use the shield as a return line for audio-sensitive circuits.

2 Both ends of the shielding layer used for the RF sensitive circuit should be grounded.

3 For circuits that are both audio sensitive and RF sensitive, use a tight shield pair. The shorter the distance between the twists, the better the shielding effect. Both ends of the shield should be grounded.

1.6 Grounding Design Guidelines

In the product design, the grounding is considered from the safety point of view or from the function, and the design according to the ground is less from the viewpoint of suppressing the disturbance. Therefore, when the grounding method, the grounding point, and the grounding wire are selected, some can be avoided. mistake. In addition, a good grounding design must be supported by a good assembly process to achieve the intended purpose.

1.6.1 When designing the grounding, select the grounding method and grounding point according to the actual situation.

For example, the frequency of the computer's radiated disturbance exceeding the limit value is concentrated in the range of 30 to 200 MHz. Therefore, the units and shielded cables inside the microcomputer should be grounded at a multiple point to the chassis. Using a single point grounding will increase the length of the grounding wire. If the grounding wire length is close to or equal to l/4 of the wavelength of the disturbance signal, its radiation capacity will be greatly increased, and the grounding wire will become the antenna. In general, the length of the ground wire should be less than 2.5cm. The grounding of the shielded cable is shown in Figure 1.

1.6.2 Selection of grounding wire

It is often seen that the internal grounding wire is a very thin single-strand wire. When the high-frequency current is passed inside, the grounding effect can be imagined due to the high-frequency impedance. Therefore, in consideration of the skin effect, the grounding wire needs to use a ribbon braided wire. If the grounding requirements are very high, silver can also be plated on the surface, which is mainly to reduce the surface resistivity of the wire, thereby achieving the purpose of reducing the high-frequency impedance of the grounding wire.

1.6.3 The grounding wire should be well connected to the grounding surface.

Generally, the standard stipulates that the DC lap impedance of the grounding wire and the grounding surface should be less than 2.5mW. For high quality grounding, the grounding surface should be surface treated to avoid oxidation and corrosion. There should be no lock washers or gaskets between the ground wire and the ground plane, and gaskets, bolts, and nuts should not be used as part of the ground loop.

1.6.4 Three grounding methods: floating ground, single point grounding and multi-point grounding

The purpose of the floating ground is to isolate the circuit or equipment from a common ground line or a public line that may cause circulation.

Disadvantages: Since the device is not directly connected to the earth, static electricity accumulation is likely to occur, and after a certain degree, breakdown will occur, which is a highly destructive source of disturbance. The compromise method is to connect a large bleeder resistor between the floating ground and the earth to eliminate the influence of static electricity accumulation. The way to achieve floating: transformer isolation, charging isolation. In addition to "floating" the ground line, the floating ground also solves the trouble caused by the inconsistent potential in the single-ground system.

Single-point grounding means that only one physical point is defined as a ground reference point, and other points that need to be grounded are directly connected to this point. If the system operating frequency is very high, the grounding wire length can be compared with the working frequency (wavelength of the signal), and the single-point grounding method can no longer be used (the grounding effect is not ideal), but multi-point grounding is required. The concept of it.

Multi-point grounding means that each grounding point in a system is directly connected to the ground plane closest to it, so that the length of the grounding wire is the shortest. The grounding point can be the bottom plate of the device, the ground wire that runs through the entire system, or the structural frame of the device. The advantage of multi-point grounding is that the circuit structure is simpler than a single point grounding. Since multiple grounding is used, many grounding loops are formed, so it is important to improve the quality of the grounding system, requiring frequent maintenance and maintaining good electrical conductivity.

Hybrid grounding: Use multi-point grounding only for places where high-frequency grounding is required, and grounding with a single point. The grounding length is measured in terms of 0.05λ~0.15λ. If this value is exceeded, multi-point grounding should be used.

In addition, in the case of a large current sudden change such as a relay, separate grounding is used to reduce transient coupling to other circuits.

Direct grounding of the load is not appropriate. Excellent shielding performance is also achieved with tightly wound twisted pairs.

When the shielded cable transmits high-frequency signals, the outer shield of the cable should be grounded at multiple points. The typical demarcation point is 100KHz. Above this value, multi-point grounding is used. Below this value, single-point grounding is used. There is a grounding point every 0.05λ~0.1λ.

The shield grounding should not be grounded in a braided shape, but the shield should be wrapped around the core and the shield should be grounded 360 degrees.

1.6.5 Ground Design Guidelines

A single point grounding can be used when the circuit size is less than 0.05λ, and multiple points can be grounded when it is greater than 0.15λ.

For systems with a wide operating frequency, use a hybrid ground.

In the event of a ground loop problem, floating ground isolation (eg transformer, optoelectronics) is available.

All grounding wires should be short.

The grounding wire should be well conductive and avoid high resistance.

There must be separate grounding systems for signal lines, signal loops, power system loops, and backplanes or enclosures, which can then be connected to a reference point.

For circuits that will experience large current abrupt changes, separate grounding systems or separate ground return lines are needed to reduce transient coupling to other circuits.

The ground wire of the low-level circuit must cross where the wires are perpendicular to each other.

Use balanced differential circuits to minimize the effects of ground circuit disturbances.

For circuits with a maximum size much smaller than λ/4, use a single-point grounded tight-twisted wire (depending on the actual condition) to make the device the best.

AC and DC lines cannot be tied together. The AC line itself has to be twisted together.

When terminating the cable shield, avoid using shielded lead wires.

When a signal needs to be transmitted over a coaxial cable, a signal loop is provided through the shield. The low frequency circuit can be grounded at a single point on the signal source; the high frequency circuit is grounded at multiple points.

The high-frequency, low-level transmission axis should be shielded by multiple layers, and each shielding layer should be grounded at a single point.

From the safety point of view, the ground wire of the test equipment is directly connected to the ground wire of the device under test; or from the safety, it is necessary to ensure that the grounding connection device can cope with the unexpected fault current, and can cope with the impact of the lightning current when the outdoor terminal is grounded.