What Exactly is a PCB? General Principles of PCB Design

When it comes to PCBs, many people will think of them as being everywhere around us. From all household appliances, various components in computers, to all kinds of digital products, PCBs are used in almost all electronic products. So what exactly is a PCB?

PCB stands for Printed Circuit Board, a substrate with circuits for mounting electronic components. A copper-clad substrate is printed with anti-etching circuits by printing, and the circuits are etched and rinsed out.

PCBs can be classified into single-layer boards, double-layer boards, and multi-layer boards.

Various electronic components are integrated on the PCB. On the most basic single-layer PCB, components are concentrated on one side, and wires are concentrated on the other side.

In this case, we need to drill holes on the board so that the pins can pass through the board to the other side. Therefore, the pins of the components are soldered on the other side.

For this reason, the two sides of such a PCB are called the Component Side and the Solder Side respectively. A double-layer board can be regarded as two single-layer boards bonded together, with electronic components and traces on both sides of the board.

Sometimes it is necessary to connect a single trace on one side to the other side of the board, which requires vias. A via is a small hole filled or coated with metal on the PCB, which can connect the wires on both sides.

Now many computer motherboards use 4-layer or even 6-layer PCBs, while graphics cards generally use 6-layer PCBs. Many high-end graphics cards, such as the NVIDIA GeForce 4Ti series, use 8-layer PCBs, which are so-called multi-layer PCBs.

The problem of connecting circuits between layers also occurs in multi-layer PCBs, which can also be achieved through vias. Since it is a multi-layer PCB, some vias do not need to penetrate the entire PCB. Such vias are called buried vias and blind vias, because they only penetrate several layers.

Blind vias connect several inner PCB layers to the surface PCB without penetrating the entire board. Buried vias only connect inner PCB layers, so they cannot be seen from the surface. In multi-layer PCBs, entire layers are directly connected to ground and power.

Therefore, we classify each layer as a Signal Layer, a Power Layer, or a Ground Layer. If components on the PCB require different power supplies, such PCBs usually have two or more power and ground layers.

The more layers a PCB has, the higher the cost. Of course, using more layers helps to improve signal stability.

The professional PCB manufacturing process is quite complex. Take a 4-layer PCB as an example.

Most motherboards use 4-layer PCBs.

During manufacturing, the middle two layers are first rolled, cut, etched, and oxidized and electroplated respectively. The four layers are the component side, power layer, ground layer, and solder press layer. Then these four layers are rolled together to form a motherboard PCB.

Next, holes are drilled and vias are fabricated. After cleaning, the circuits on the outer two layers are printed, copper-clad, etched, tested, and a solder mask and silkscreen are applied.

Finally, the full-size PCB (containing many motherboards) is punched into individual motherboard PCBs, which are then vacuum-packed after testing.

If the copper foil is not well bonded during PCB manufacturing, it may peel off easily, implying potential short circuits or capacitive effects (prone to interference).

Vias on the PCB must also be noted. If a hole is not drilled in the center but offset to one side, it will cause uneven matching, or easily contact the middle power or ground layer, resulting in potential short circuits or poor grounding.

The first step in manufacturing is to establish the wiring for connections between components.

We use the negative film transfer method to transfer the working film to the metal conductor. This technique involves laying a thin layer of copper foil over the entire surface and removing the excess parts.

Additive transfer is another less commonly used method, which involves laying copper wires only where needed, but we will not discuss it further here.

Positive photoresist is made of photosensitive material, which dissolves under light. There are many ways to apply photoresist to the copper surface, but the most common method is to heat it and roll it on the surface containing photoresist.

It can also be sprayed on in liquid form, but dry film provides higher resolution and can produce finer wires. A photomask is just a template for manufacturing PCB layers.

Before the photoresist on the PCB is exposed to UV light, the photomask covering it can prevent part of the photoresist from being exposed. The areas covered by the photoresist will become the wiring. After the photoresist is developed, the remaining bare copper parts are etched.

The etching process can immerse the board in an etching solvent or spray the solvent on the board. Ferric chloride is commonly used as an etching solvent. The remaining photoresist is removed after etching.

Generally, the width should not be less than 0.2mm (8mil). For high-density and high-precision PCBs, the spacing and trace width are generally 0.3mm (12mil). When the copper foil thickness is about 50um, a trace width of 1~1.5mm (60mil) can carry 2A.

The common ground is generally 80mil, especially for applications with microprocessors.

A signal is considered a high-speed signal when the rise/fall time of the signal is < 3~6 times the signal propagation time.

For digital circuits, the key is the steepness of the signal edge, i.e., the rise and fall time of the signal.

According to the theory in the classic book High Speed Digital Design, a signal is a high-speed signal if the time for the signal to rise from 10% to 90% is less than 6 times the wire delay! ——That is to say, an 8KHz square wave signal is still a high-speed signal as long as the edge is steep enough, and transmission line theory must be used for routing.

Multi-layer PCBs have better electromagnetic compatibility (EMC) design, enabling the PCB to meet EMC and sensitivity standards during normal operation. Proper stackup helps shield and suppress EMI. The following basic principles are generally followed in the layer division and stackup of multi-layer PCBs: ① The power plane should be as close as possible to the ground plane and under the ground plane; ② Wiring layers should be arranged adjacent to the reference plane layer; ③ The impedance between power and ground is minimized; ④ Striplines are formed in the middle layer, and microstrips are formed on the surface. The two have different characteristics; ⑤ Important signal lines should be close to the ground layer.

A very important advantage of multi-layer boards over ordinary double-layer and single-layer boards is that signal lines and power can be distributed on different board layers, improving signal isolation and anti-interference performance.

However, many engineers still struggle with PCB layer division and stackup. Take the commonly used 4-layer board as an example.

There are several stacking sequences for 4-layer boards (the advantages and disadvantages of each stacking are explained below):

First case: This should be the best among 4-layer boards. Because the outer layers are ground layers, they provide shielding against EMI. At the same time, the power layer is close to the ground layer, reducing the internal resistance of the power supply and achieving the best effect.

However, the first case cannot be used when the board density is relatively high. Because this will not ensure the integrity of the first ground layer, and the second signal layer will perform worse.

In addition, this structure cannot be used when the full-board power consumption is relatively high.

Second case: This is our most commonly used method. Structurally, it is also not suitable for high-speed digital circuit design. Because in this structure, it is difficult to maintain low power impedance.

Taking a 2mm-thick board as an example: Z0 is required to be 50ohm. With a trace width of 8mil and copper foil thickness of 35μm, the distance between the first signal layer and the ground layer is 0.14mm, while the distance between the ground layer and the power layer is 1.58mm, which greatly increases the internal resistance of the power supply.

In this structure, since radiation is directed into space, a shielding plate must be added to reduce EMI.

Third case: The signal line quality on the S1 layer is the best, followed by S2. It provides EMI shielding, but the power impedance is relatively high.

This board can be used when the full-board power consumption is high and the board is an interference source or close to an interference source.

Note: S1: First signal wiring layer; S2: Second signal wiring layer; GND: Ground layer; POWER: Power layer.

The amplitude of the reflected voltage signal is determined by the source reflection coefficient ρs and the load reflection coefficient ρL: ρL = (RL - Z0) / (RL + Z0) ρS = (RS - Z0) / (RS + Z0)

In the above formula, if RL=Z0, the load reflection coefficient ρL=0; if RS=Z0, the source reflection coefficient ρS=0.

Since the characteristic impedance Z0 of an ordinary transmission line usually needs to meet about 50Ω, and the load impedance is usually several thousand to tens of thousands of ohms.

Therefore, it is difficult to achieve impedance matching at the load end. However, since the output impedance at the signal source end is usually relatively small, about more than ten ohms.

Therefore, it is much easier to achieve impedance matching at the source end. If a resistor is connected in parallel at the load end, the resistor will absorb part of the signal, which is unfavorable for transmission (my understanding). When selecting the TTL/CMOS standard 24mA drive current, its output impedance is approximately 13Ω.

If the transmission line impedance Z0=50Ω, a 33Ω source matching resistor should be added. 13Ω+33Ω=46Ω (close to 50Ω, weak underdamping helps the signal setup time).

The matching impedance varies when other transmission standards and drive currents are selected. In high-speed logic and circuit design, it is recommended to add source matching resistors to some key signals such as clocks and control signals.

After this connection, the signal will still reflect back from the load end. Due to the source impedance matching, the reflected signal will not reflect back again.

Power lines should be as short as possible, routed in straight lines, and preferably in a tree shape instead of a ring shape.

Ground loop problem: For digital circuits, the ground loop current caused by the ground loop is only tens of millivolts, while the anti-interference threshold of TTL is 1.2V, and that of CMOS circuits can reach 1/2 of the power supply voltage. That is to say, the ground loop current will not adversely affect the operation of the circuit.

On the contrary, if the ground is not closed, the problem will be bigger, because the pulse power current generated by digital circuits during operation will cause unbalanced ground potentials at various points. For example, I measured that the ground current of 74LS161 is 1.2A when it flips (measured with a 2Gsps oscilloscope, ground current pulse width 7ns).

Under the impact of large pulse current, if a dendritic ground (trace width 25mil) is used, the potential difference between various points on the ground will reach the order of hundreds of millivolts. After adopting the ground loop, the pulse current will be distributed to various points of the ground, greatly reducing the possibility of interfering with the circuit.

With a closed ground, the measured maximum instantaneous potential difference of the ground of each device is 1/2 to 1/5 of that of an unclosed ground.

Of course, the measured data of circuit boards with different densities and speeds vary greatly. What I said above refers to the level approximately equivalent to the Z80 Demo board attached to Protel 99SE; for low-frequency analog circuits, I think the power frequency interference after the ground is closed is induced from space, which cannot be simulated and calculated anyway.

If the ground is not closed, there will be no ground eddy current. What is the theoretical basis for beckhamtao's statement that "the power frequency induced voltage will be larger if the ground is open-loop"?

Let me give two examples from senior engineers: ① Seven years ago, I took over a project from someone else, a precision pressure gauge using a 14-bit A/D converter, but the actual effective precision was only 11 bits.

After investigation, there was 15mVp-p power frequency interference on the ground. The solution was to cut off the analog ground loop of the PCB, and route the ground from the front-end sensor to the A/D in a dendritic shape with flying wires.

The PCB of the mass-produced model was later remanufactured according to the flying wire routing, and no problems have occurred so far.

② A friend loves audio DIY and built a power amplifier by himself, but there was always AC noise in the output. I suggested cutting off the ground loop, and the problem was solved.

Afterwards, this friend consulted the PCB diagrams of dozens of "Hi-Fi famous machines" and confirmed that no machine uses a ground loop in the analog part.

The Printed Circuit Board (PCB) is the support for circuit components and devices in electronic products.

It provides electrical connections between circuit components and devices. With the rapid development of electronic technology, the density of PCBs is getting higher and higher.

The quality of PCB design greatly affects the anti-interference ability. Therefore, when designing a PCB, the general principles of PCB design must be followed, and the requirements of anti-interference design must be met.

To achieve the best performance of electronic circuits, the placement of components and the routing of wires are very important. To design a PCB with good quality and low cost, the following general principles should be followed:

First, consider the size of the PCB. If the PCB size is too large, the printed lines will be long, the impedance will increase, the anti-noise ability will decrease, and the cost will increase; if it is too small, the heat dissipation will be poor, and adjacent lines will be easily interfered.

After determining the PCB size, determine the position of special components.

Finally, layout all components of the circuit according to the functional units of the circuit.

The following principles should be followed when determining the position of special components: (1) Shorten the connections between high-frequency components as much as possible, and try to reduce their distributed parameters and mutual electromagnetic interference. Interference-prone components should not be placed too close to each other, and input and output components should be kept as far away as possible.

(2) There may be a high potential difference between some components or wires, so the distance between them should be increased to avoid accidental short circuits caused by discharge. Components with high voltage should be arranged as far as possible out of reach of hands during debugging.

(3) Components weighing more than 15g should be fixed with brackets and then soldered. Large, heavy, and high-heat-generating components should not be mounted on the printed board, but on the chassis base of the whole machine, and heat dissipation should be considered. Thermosensitive components should be kept away from heating components.

(4) The layout of adjustable components such as potentiometers, adjustable inductance coils, variable capacitors, and micro switches should consider the structural requirements of the whole machine. If adjusted inside the machine, they should be placed on the printed board for easy adjustment; if adjusted outside the machine, their positions should correspond to the adjustment knobs on the chassis panel.

(5) Space should be reserved for the positioning holes of the printed board and the space occupied by the fixed brackets.

When laying out all components of the circuit according to the functional units of the circuit, the following principles should be followed: (1) Arrange the positions of each functional circuit unit according to the circuit flow, so that the layout facilitates signal circulation and keeps the signal direction as consistent as possible.

(2) Take the core component of each functional circuit as the center and layout around it. Components should be arranged evenly, neatly, and compactly on the PCB. Minimize and shorten the leads and connections between components.

(3) For circuits operating at high frequencies, consider the distributed parameters between components. For general circuits, arrange components in parallel as much as possible. This is not only beautiful, but also easy to install and solder, and easy to mass-produce.

(4) Components located at the edge of the circuit board should generally be no less than 2mm away from the edge of the circuit board. The best shape of the circuit board is rectangular, with an aspect ratio of 3:2 or 4:3. When the board size exceeds 200x150mm, consider the mechanical strength of the circuit board.

The routing principles are as follows: (1) Wires for input and output terminals should avoid adjacent parallel routing as much as possible. It is better to add ground wires between lines to avoid feedback coupling.

(2) The minimum width of printed wires is mainly determined by the adhesion strength between the wires and the insulating substrate and the current value flowing through them.

When the copper foil thickness is 0.05mm and the width is 1~15mm, a current of 2A can pass through without the temperature exceeding 3℃. Therefore, a trace width of 1.5mm can meet the requirements.

For integrated circuits, especially digital circuits, the trace width is usually 0.02~0.3mm. Of course, use wide wires as much as possible if allowed, especially power and ground wires.

The minimum spacing of printed wires is mainly determined by the insulation resistance and breakdown voltage between wires under the worst conditions.

For integrated circuits, especially digital circuits, the spacing can be as small as 5~8mm as long as the process allows.

(3) The corners of printed wires are generally arc-shaped, while right angles or included angles will affect electrical performance in high-frequency circuits.

In addition, avoid using large-area copper foil as much as possible. Otherwise, the copper foil will easily expand and fall off when heated for a long time.

If large-area copper foil must be used, it is better to use a grid shape, which helps to eliminate volatile gases generated by the heating of the adhesive between the copper foil and the substrate.

The center hole of the pad should be slightly larger than the lead diameter of the device. An overly large pad is prone to cold solder joints. The outer diameter D of the pad is generally not less than (d+1.2)mm, where d is the lead aperture. For high-density digital circuits, the minimum pad diameter can be (d+1.0)mm.

The anti-interference design of printed circuit boards is closely related to specific circuits. Only some common measures for PCB anti-interference design are described here.

Widen the power line as much as possible according to the current of the printed circuit board to reduce the loop resistance.

At the same time, make the direction of power and ground wires consistent with the direction of data transmission, which helps to enhance anti-noise ability.

The principles of ground line design are: (1) Separate digital ground from analog ground. If the circuit board has both logic circuits and linear circuits, separate them as much as possible.

The ground of low-frequency circuits should be grounded in parallel at a single point as much as possible. If it is difficult to route wires in practice, they can be partially connected in series and then grounded in parallel.

High-frequency circuits should be grounded in series at multiple points. The ground wire should be short and thick, and grid-shaped large-area ground foil should be used around high-frequency components as much as possible.

(2) Thicken the ground wire as much as possible. If the ground wire is a thin line, the ground potential changes with the current, reducing the anti-noise performance.

Therefore, thicken the ground wire so that it can pass three times the allowable current on the printed board. If possible, the ground wire should be more than 2~3mm.

(3) The ground wire forms a closed loop. For printed boards composed only of digital circuits, the ground circuit arranged in a closed loop can mostly improve the anti-noise ability.

One of the conventional practices in PCB design is to configure appropriate decoupling capacitors at each key position of the printed board.

The general configuration principles of decoupling capacitors are: (1) Connect an electrolytic capacitor of 10~100uF across the power input terminal. If possible, connect more than 100uF.

(2) In principle, each integrated circuit chip should be equipped with a 0.01pF ceramic chip capacitor. If there is not enough space on the printed board, one 1~10pF capacitor can be arranged for every 4~8 chips.

(3) For devices with weak anti-noise ability and large power changes when turned off, such as RAM and ROM memory devices, connect decoupling capacitors directly between the power and ground wires of the chip.

(4) The capacitor leads should not be too long, especially high-frequency bypass capacitors.

In addition, pay attention to the following two points: (1) When there are contactors, relays, buttons and other components on the printed board, large spark discharges will occur when operating them. The RC circuit shown in the attached figure must be used to absorb the discharge current.

Generally, R is 1~2KΩ and C is 2.2~47μF.

(2) CMOS has high input impedance and is susceptible to induction. Therefore, unused terminals should be grounded or connected to the positive power supply during use.

How to achieve high PCB routing completion rate and shorten design time? This article introduces the design tips and key points of PCB planning, layout and routing.

Now PCB design time is getting shorter, circuit board space is getting smaller, device density is getting higher, extremely strict layout rules and large-size components make the designer's work more difficult.

To solve design difficulties and speed up product launch, many manufacturers now tend to use dedicated EDA tools for PCB design.

However, dedicated EDA tools cannot produce ideal results or achieve 100% routing completion rate, and the routing is usually messy. It often takes a lot of time to complete the remaining work.

There are many popular EDA tool software on the market, but they are similar except for the terms and function key positions. How to use these tools to better implement PCB design?

Careful analysis of the design before routing and careful setting of the tool software will make the design more in line with the requirements. The following is the general design process and steps.

The board size and number of routing layers need to be determined at the initial stage of design.

If the design requires the use of high-density Ball Grid Array (BGA) components, the minimum number of routing layers required for routing these devices must be considered.

The number of routing layers and the stack-up method directly affect the routing and impedance of printed lines.

The size of the board helps to determine the stacking method and printed line width to achieve the desired design effect.

Over the years, people have always thought that the fewer layers a circuit board has, the lower the cost, but there are many other factors affecting the manufacturing cost of circuit boards.

In recent years, the cost difference between multi-layer boards has been greatly reduced.

It is better to use more circuit layers and evenly distribute copper cladding at the beginning of the design to avoid finding that a small number of signals do not meet the defined rules and space requirements near the end of the design, forcing the addition of new layers.

Careful planning before design will avoid many troubles in routing.

The auto-routing tool itself does not know what to do.

To complete the routing task, the routing tool needs to work under the correct rules and restrictions.

Different signal lines have different routing requirements. All signal lines with special requirements should be classified, and different design classifications are also different.

Each signal category should have a priority. The higher the priority, the stricter the rules.

Rules involve printed line width, maximum number of vias, parallelism, mutual influence between signal lines, and layer restrictions. These rules have a great impact on the performance of routing tools.

Careful consideration of design requirements is an important step for successful routing.

To optimize the assembly process, Design for Manufacturability (DFM) rules impose restrictions on component layout.

If the assembly department allows component movement, the circuit can be properly optimized to facilitate auto-routing. The defined rules and constraints will affect the layout design.

When laying out, consider routing channels and via areas.

These paths and areas are obvious to the designer, but the auto-routing tool only considers one signal at a time. By setting routing constraints and defining the layers for routable signal lines, the routing tool can complete the routing as the designer envisions.

In the fan-out design stage, to enable the auto-routing tool to connect component pins, each pin of a surface-mount device should have at least one via, so that the circuit board can perform inner-layer connections, In-Circuit Test (ICT) and circuit reprocessing when more connections are needed.

To maximize the efficiency of the auto-routing tool, use the largest possible via size and printed lines as much as possible. A spacing of 50mil is ideal.

Use the via type that maximizes the number of routing paths. When designing fan-out, consider circuit in-circuit testing.

Test fixtures may be expensive and are usually ordered when full production is about to start. It is too late to consider adding nodes to achieve 100% testability at this time.

After careful consideration and prediction, the circuit in-circuit test design can be carried out at the initial stage of design and implemented at the later stage of production. The via fan-out type is determined according to the routing path and circuit in-circuit test. Power and ground will also affect routing and fan-out design.

To reduce the inductance caused by the connection lines of filter capacitors, vias should be as close as possible to the pins of surface-mount devices. Manual routing can be used if necessary.

This may affect the originally envisaged routing path, or even make you reconsider which via to use. Therefore, the relationship between via and pin inductance must be considered and the priority of via specifications must be set.

Although this article mainly discusses auto-routing, manual routing is an important process in PCB design now and in the future.

Manual routing helps the auto-routing tool complete the routing work.

By manually routing and fixing selected nets, paths can be formed for auto-routing.

Regardless of the number of key signals, route these signals first, either manually or in combination with an auto-routing tool.

Key signals usually must pass careful circuit design to achieve the desired performance.

After routing is completed, the relevant engineering staff will check these signal routings, which is relatively easy.

After passing the inspection, fix these wires and then start auto-routing the remaining signals.

The routing of key signals needs to consider controlling some electrical parameters during routing, such as reducing distributed inductance and EMC. The same applies to the routing of other signals.

All EDA manufacturers provide a method to control these parameters.

Understanding the input parameters of the auto-routing tool and the impact of input parameters on routing can guarantee the quality of auto-routing to a certain extent.

General rules should be used to auto-route signals.

By setting restrictions and prohibited routing areas to limit the layers used for a given signal and the number of vias used, the routing tool can auto-route according to the engineer's design ideas.

If the layers and vias used by the auto-routing tool are not restricted, auto-routing will use every layer and generate many vias.

After setting constraints and applying the created rules, auto-routing will achieve results similar to expectations. Of course, some tidying up may be required, and space for other signals and net routing must be ensured.

After part of the design is completed, fix it to prevent being affected by the subsequent routing process.

Use the same steps to route the remaining signals.

The number of routing times depends on the complexity of the circuit and the general rules you define.

Current auto-routing tools are very powerful and can usually complete 100% of the routing.

However, when the auto-routing tool does not complete all signal routing, manual routing of the remaining signals is required.

(1) Slightly change the settings and try multiple routing paths; (2) Keep the basic rules unchanged, try different routing layers, different printed line and spacing widths, different line widths, and different types of vias such as blind vias and buried vias, and observe how these factors affect the design results; (3) Let the routing tool process the default nets as needed; (4) The less important the signal, the greater the routing freedom of the auto-routing tool.

If the EDA tool software you use can list the routing length of signals, check these data, and you may find that some signal routings with few constraints are very long.

This problem is relatively easy to handle. Manual editing can shorten the signal routing length and reduce the number of vias.

During the sorting process, you need to determine which routing is reasonable and which is not.

Like manual routing design, auto-routing design can be sorted and edited during the inspection process.

Previous designs often paid attention to the visual effect of the circuit board, but now it is different.

Auto-designed circuit boards are not as beautiful as manually designed ones, but they can meet the specified electronic characteristics and ensure the complete performance of the design.

About Maxipcb
Maxipcb enables advanced electronic innovation. We deliver one-stop solutions including circuit design, simulation, testing, PCB fabrication, component sourcing and SMT&PCBA assembly, to boost R&D efficiency, speed up mass production and control full-cycle risks. We serve global sectors like communication, industrial automation, aerospace, automotive and semiconductor, jointly forging a safer, connected intelligent future.