Two major difficulties in the field of microelectronics is the processing of high-frequency signals and weak signals, PCB production level in this regard is particularly important, the same principle design, the same components, different people produced PCB have different results , Then how to make a good PCB board?
How to get a high quality PCB?
First, we must clearly design goals
Received a design task, we must first clear the design goal is ordinary PCB board, high-frequency PCB board, small signal processing PCB board or both high frequency and small signal processing PCB board, if it is a common PCB board, As long as the layout and wiring to achieve reasonable and tidy, mechanical dimensions can be accurate, if the load line and long line, it is necessary to use certain means for processing, reduce the load, to strengthen the drive long-term, focusing on preventing long-term reflection.
When there are more than 40MHz signal lines on the board, special consideration should be given to these signal lines, such as crosstalk between lines, etc. If the frequency is higher, the length of the wiring is more restricted, and according to the distribution parameter Network theory, the interaction between high-speed circuit and its connection is the decisive factor in the system design can not be ignored as the gate transmission speed increases, the opposition on the signal line will be a corresponding increase in crosstalk between adjacent signal lines will be Proportionally increase, usually high-speed circuit power dissipation and heat dissipation are also very large, high-speed PCB should be paid enough attention.
When there are weak signals on the board of millivolts or even microvolts, special care should be taken for these signal lines. Small signals are too weak to be easily disturbed by other strong signals. Shielding measures are often necessary. Otherwise, Greatly reducing the signal to noise ratio so that the useful signal is submerged by noise and can not be effectively extracted.
The commissioning of the board should also be considered in the design stage. The physical location of test points and the isolation of test points should not be neglected. Because some small and high-frequency signals can not be added directly to the probe for measurement.
In addition to consider some other relevant factors, such as the number of board layers, the use of component package shape, the mechanical strength of the board, etc. Before doing PCB board, to make the design goals of the design.
Second, understand the function of the components used in the layout of wiring requirements
We know that there are some special components in the layout of wiring has special requirements, such as LOTI and APH used analog signal amplifier, analog signal amplifier to the power requirements of a stable, small ripple. Analog small signal part should try to stay away from power devices. OTI board, small signal amplification part of the special shield also added to shield the stray electromagnetic interference.
NTOI GLINK chips used in the board is the ECL process, power powerful fever, the heat must be in the layout of the issue must be given special consideration, if the natural cooling, it is necessary GLINK chip on the air circulation is relatively smooth , And the heat dissipated can not yet have a big impact on other chips.If the board is equipped with speakers or other high-power devices, it may cause serious pollution of the power of this point should also be given sufficient attention.
Third, the component layout considerations
Component layout First factor to be considered is the electrical properties of the connection to close the components as much as possible, especially for some high-speed lines, the layout should make it as short as possible, the power signal and small signal devices To separate. In the premise of meeting the circuit performance, but also consider the components placed neatly, beautiful, easy to test, the board's mechanical size, socket location also need to seriously consider.
The transmission delay time on the ground and interconnect lines in high-speed systems is also the first factor to be considered in the system design. The transmission time on the signal lines has a great impact on the overall system speed, especially for high-speed ECL circuits, although the integration The circuit block itself is very fast, but the system speed is greatly reduced due to the increased delay time on the backplane with normal interconnects (about 2 ns per 30cm line length). Like shift registers, Synchronous counter synchronization of the work of this component is best placed on the same piece of the board, because the different plug-in board clock signal transmission delay time is not equal, may make the shift register owner error, if not on a board, Then where synchronization is critical, the clock lines connected from the common clock source to each board must be the same length.
Fourth, the wiring considerations
With the design of OTNI and star optical fiber network completed, there will be more boards with high-speed signal lines above 100MHz that need to be designed. Here we will introduce some basic concepts of high-speed line.
Transmission line:
Any "long" signal path on a printed circuit board can be considered as a transmission line. If the transmission delay of the line is much shorter than the rise time of the signal, the primary reflection produced during the signal rise will be submerged With overshoot, kickback and ringing no longer present, for most MOS circuits today, the trace can be long in meters without signal distortion due to the much larger rise time versus line propagation delay time. For faster logic circuits, especially ultra-fast ECLs.
In the case of integrated circuits, the length of the traces must be significantly shortened in order to maintain the integrity of the signal due to faster edge speeds.
There are two ways to make high-speed circuits operate over relatively long lines without severe waveform distortion. The TTL uses a Schottky diode clamp for the fast-descend edges, clamping the overshoot to a diode drop below ground , Which reduces the backlash. The slower rising edge allows for an overshoot, but it is attenuated by the relatively high output impedance (50 ~ 80Ω) of the circuit at level 'H' In addition, due to the level 'H' state of the larger immunity to the recoil problem is not very prominent, HCT Series devices, if the combination of Schottky diode clamp and series resistance termination method, which The effect of improvement will be more obvious.
The TTL shaping method described above appears to be somewhat less efficient at higher bit rates and faster edge rates when fanning out along the signal lines because of the reflected waves in the line that will tend to be synthesized at high rates , Resulting in serious signal distortion and reduced anti-interference ability.Therefore, in order to solve the reflection problem, ECL system usually uses another method: line impedance matching method.With this method can make the reflection is controlled, the signal integrity Guarantee.
Strictly speaking, transmission lines are not very needed for conventional TTL and CMOS devices with slower edge speeds, and transmission lines are not always needed for high speed ECL devices with faster edge speeds, but when using transmission lines , They have the advantage of predicting connection delays and controlling reflections and oscillations through impedance matching.
1, the basic factors that determine whether to adopt the transmission line are the following five:
(1) the system signal along the rate,
(2) Connection distance;
(3) capacitive load (fan out of how much),
(4) resistive load (line termination);
(5) Permissible recoil and overshoot percentage (reduction of AC immunity).
2, several types of transmission lines
(1) Coaxial cable and twisted pair: They are often used in the system and the system connection.Coaxial cable characteristic impedance is usually 50Ω and 75Ω, twisted pair is usually 110Ω.
(2) PCB on the microstrip line
The microstrip line is a strip conductor (signal line) that is separated from the ground plane by a dielectric. If the line thickness, width, and distance from the ground plane are controllable, its characteristic impedance Is also controllable. The characteristic impedance Z0 of the microstrip line is:
(3) strip line in the PCB
A stripline is a strip of copper tape in the middle of a dielectric between two conductive planes.If the thickness and width of the line, the permittivity of the medium, and the distance between two conductive planes are controllable, then the characteristics of the line The impedance is also controllable, the characteristic impedance of the stripline is:
3, termination transmission line
Termination of a line at the receiving end with a resistance equal to the characteristic impedance of the line means that the transmission line is a parallel termination which is used primarily to obtain the best electrical performance, including driving distributed loads.
Sometimes in order to save power consumption, a series of 104 capacitors are connected in series with the terminating resistor to form an AC termination circuit, which can effectively reduce the DC loss.
A resistor is connected in series between the driver and the transmission line, and the terminations of the line are no longer connected to the terminating resistor, this terminating method is called terminating in series. Overshoot and ringing on longer lines is available with series damping or series termination Technology to control the series damping is the use of a drive gate in series with a small resistance (typically 10 ~ 75Ω) to achieve this damping method is suitable for the impedance associated with the control of the line associated with (such as the floor wiring, No ground plane circuit boards and most of the winding and so on.
The sum of the series resistance value and the output impedance of the circuit (drive gate) is equal to the characteristic impedance of the transmission line when the series connection is terminated The series connection connection has the disadvantage that the terminal can only use the lumped load and the transmission delay time is long. However, This can be overcome by using extra tandem terminated transmission lines.
4, non-terminated transmission line
If the line delay is much shorter than the rise time of the signal, it is possible to use the transmission line without series or parallel termination. If a non-terminal connection has a two-way delay (the signal travels once on the transmission line) The rise time of the signal is short, then the kickback due to non-terminated is approximately 15% of the logic swing. The maximum open-circuit length is approximately:
Lmax Where: tr for the rise time
tpd is the transmission delay time per unit line length
5, several ways to compare the termination
Parallel termination and series termination have their own advantages, which one or two are used, depending on the designer's hobby and system requirements.
The main advantage of parallel termination is that the system is fast and that the signal is transmitted intact over the wire. The long-term load neither affects the propagation delay of the driver's gate driving the long wire nor does it affect its signal edge speed, The propagation delay of the signal along the long line increases.When driving a large fan-out, the load can be distributed along the branch short lines rather than the terminal where the load must be lumped together as in series termination.
The series termination method enables the circuit to drive several parallel load lines. The delay of series termination due to capacitive load is about twice as large as that of the corresponding parallel termination, while for short lines, the edge However, the crosstalk of the series termination is smaller than that of the parallel termination. The main reason is that the amplitude of the signal transmitted along the series termination is only half of the logic swing, and therefore Switching current is only half the switching current parallel termination, the signal energy is small crosstalk small.
PCB is the choice of double-sided board or multi-layer board, depends on the maximum operating frequency and the complexity of the circuit system and the assembly density requirements to decide.When the clock frequency over 200MHZ is best to use multi-layer board.If the operating frequency exceeds 350MHz , The best choice to use PTFE as the dielectric layer of the printed circuit board, because of its high-frequency attenuation to be smaller, smaller parasitic capacitance, the transmission speed to be faster, but also because of the larger Z0 and power consumption , The printed circuit board alignment has the following principles and requirements:
(1) All parallel signal lines should be kept as large as possible to reduce the crosstalk. If there are two signal lines close together, it is best to take a ground line between the two lines so as to shield effect.
(2) Design signal transmission line to avoid sharp corners to prevent the transmission line characteristics of the sudden changes in the impedance reflection, as far as possible designed to have a uniform arc of a certain size.
(3) The width of the printed line can be calculated according to the characteristic impedance calculation formula of the microstrip line and the strip line, and the characteristic impedance of the microstrip line on the printed circuit board is generally between 50 and 120Ω. Impedance, line width must be done very narrow, but the very thin lines are not easy to make.According to various factors, the general choice of impedance value of about 68Ω more appropriate, because the choice of 68Ω characteristic impedance, delay time and power consumption To achieve the best balance between. A 50Ω transmission line will consume more power; larger impedance can reduce the power consumption, but will make the transmission delay time hate .Because of the negative line capacitance will cause the transmission delay time increases and However, the intrinsic capacitance per unit length of the line segment with low characteristic impedance is relatively large, so the propagation delay time and the characteristic impedance are less affected by the load capacitance.One of the important features of a properly terminated transmission line is that the branch Short-term on-line delay time should have no effect.When Z0 is 50Ω, the length of short branches must be limited to 2.5cm, in order to avoid large ringing.
(4) For double-sided board (or four-layer board to go four lines) .The circuit board on both sides of the line perpendicular to each other, in order to prevent each other to produce the main crosstalk.
(5) Printed circuit board if equipped with high-current devices, such as relays, lights, speakers, etc., their best to separate ground separately to reduce the noise on the ground, these high-current devices should be ground Connect to a separate ground bus on the board and backplane, and these individual ground lines should be connected to the ground of the entire system.
(6) If there is a small signal amplifier on the board, the weak signal line before amplification should be far away from the strong signal line, and the wiring should be as short as possible, and if possible, it should be shielded with the ground wire.