Useful Tools

contact us contact tutor/group email to a friend accessibility options report a fault
Concepts of PCB Design

Concepts of Printed Circuit Design

Unit 2: Development of PCB Products

Unit Contents

 

2.1 Introduction

A product is something sold by an enterprise to its customers, in order to satisfy a want or need. Product development is the set of activities beginning with the perception of a market opportunity and ending in the production, sale and delivery of a product.

Although many categories of product exist, this unit focuses on products that are engineered, discrete and physical. It relates particularity to products that involve Printed Circuit Boards or Assemblies (PCBs or PCAs).

Product development can be both exciting and intense. For some products, the focused efforts of a small team result in a successful new product. For complex products, the development team can be very large. Most products, regardless of their complexity, generally last from two to five years. However, when thousands of components are involved, hundreds or thousands of people must develop these parts. The cost implications can be enormous. The challenges of co-ordination and integration can thus be substantial, making the activity a significant task for any development team.

Product development, although incorporating many creative ‘design-related’ activities, is often quite procedural, with sets of activities that can generally be documented, studied, and improved. It is the recognition of the structure of these processes that facilitates improvement of the practice of product development.

To develop a quality product, organisations have found that wide ranges of skills are required. The necessary talents may be found in various types of engineers, marketers, industrial designers, managers, manufacturing personnel, and others. Current best practice always involves a team of people representing the necessary disciplines and skills (a cross-functional team).

Generally, the designer works within the context of an existing production system that can only be minimally modified. However in some cases, the production system will be designed or re-designed in conjunction with the design of the product. When design engineers and manufacturing engineers work together to design and rationalise both the product and production and support processes, it is known as integrated product and process design. The designer's consideration of design for manufacturability, cost, reliability and maintainability is the starting point for integrated product development.

The concept of DFX has been described as a critical success factor. When properly implemented and executed, it will ensure a high degree of manufacturability and testability. The lack of a competent DFX culture and curriculum can result in ‘design for failure’.

Electronic assembly operations typically focus on three key metrics: quality, cost and delivery. For a company to remain successful, these metrics must be achieved successfully over and over again. A strong DFX program will help achieve these targets with the least amount of effort and risk. PCB industry has slowly but surely embraced DFX concepts.

For this reason, many teams practice Design for eXcellence (DfX) methodologies, where X may correspond to one of dozens of quality criteria such as reliability, robustness, serviceability, test, environmental impact or manufacturability. The unit will focus on three of these in particular.

The most common of these methodologies is Design for Manufacturing (DfM), which is of universal importance because it directly addresses manufacturing costs.

The Design for Fabrication (DfF) of PCBs relates to the production of the bare board itself and thus incorporates aspects of material selection, size, shape, panelisation, number of layers, trace widths, pad and hole relationships, copper and via distribution and some surface coatings.

The Design for Assembly (DfA) of PCBs deals with the population of the board with the desired component set. It therefore embraces placement strategy, component type (through-hole/surface-mount), component density (single or doubled-sided placement), component packages, component orientation/alignment and automatic and manual assembly considerations.1

1 Although strictly within the auspices of Design for Test (DfT), consideration should also be given to bare-board and assembled board test (in-circuit or functional).

[Back to Top]

2.2 Product Development

2.2.1 Development Domains

To comprehend fully the product development processes, it is essential to consider the three relevant domains: product, process, and organisation. In complex development situations, each of these three domains is decomposed in order to manage the complexity. Prior to any further discussion, it is useful to document the decomposition of each of the three domains:

2.2.2 Generic Development Process

Product development is the process of creating a new product to be sold by a business or enterprise to its customers.

Design refers to those activities involved in creating the styling, look and feel of the product, deciding on the product's mechanical architecture, selecting materials and processes, and engineering the various components necessary to make the product work.

Development refers collectively to the entire process of identifying a market opportunity, creating a product to appeal to the identified market, and finally, testing, modifying and refining the product until it is ready for production. A product can be any item from a book, musical composition, or information service, to an engineered product such as a computer, hair dryer, or washing machine.

The impetus for a new product normally comes from a perceived market opportunity or from the development of a new technology. Marketing would have the responsibility of determining how the technology should be packaged to have the greatest appeal to its customers. With either scenario, manufacturing is responsible for estimating the cost of building the prospective new product, and their estimations are used to project a selling price and estimate the potential profit for the company.

2.2.3 Structured Approach

Figure 1: Stage-Gate Model

Figure 1: Stage-Gate Model

As shown in figure 1, a process is a sequence of steps that transforms a set of inputs into a set of outputs. A product development process is thus the sequence of steps that an organisation employs to conceive, design and commercialise a product. This process is unique to the organisation, with some companies having a very precise and detailed process, whereas other enterprises may not even be able to describe their process. However, a well-designed process is useful for the following reasons:

The structured generic process, as shown in figure 1, consists of five phases of development and their associated gates. These may now be described.

Phase 0: Idea validation

Within this stage, the needs of the target market are identified, alternative product concepts are generated and evaluated, and one or more concepts are selected for further development and testing. A concept is a description of the form, function, and features of a product and is usually accompanied by a set of specifications, an analysis of competitive products, and an economic justification of the project. Passing Gate 0 launches the project.

Phase 1: Conceptual design

The conceptual design phase includes the definition of the product architecture and the decomposition of the product into subsystems and components. The final assembly scheme for the production system is usually defined during this phase as well. The output of this phase usually includes a geometric layout of the product, a functional specification of each of the product's subsystems, and a preliminary process flow diagram for the final assembly process. Passing Gate 1 approves the project for implementation.

Phase 2: Specification and design

This phase includes the complete specification of the geometry, materials, and tolerances of all of the unique parts in the product and the identification of all of the standard parts to be purchased from suppliers. A process plan is established and tooling is designed for each part to be fabricated within the production system. The output of this phase is the control documentation for the product i.e. the drawings or computer files describing the geometry of each part and its production tooling, the specifications of the purchased parts, and the process plans for the fabrication and assembly of the product. Reaching Gate 2 releases the design for prototype manufacture.

Phase 3: Prototype production and testing

The testing and refinement phase involves the construction and evaluation of multiple pre-production or prototype versions of the product. Early (alpha) prototypes are usually built with production-intent parts i.e. parts with the same geometry and material properties as intended for the production version of the product but not necessarily fabricated with the actual processes to be used in production. Alpha prototypes are tested to determine whether or not the product will work as designed and whether or not the product satisfies the key customer needs. Later (beta) prototypes are usually built with parts supplied by the intended production processes but may not be assembled using the intended final assembly process. Beta prototypes are extensively evaluated internally and are also typically tested by customers in their own use environment. The goal for the beta prototypes is usually to answer questions about performance and reliability in order to identify necessary engineering changes for the final product. Passing Gate 3 commits the product for volume manufacturing.

Phase 4: Manufacturing ramp-up

In this phase, the product is made using the intended production system. The purpose of this stage is to train the work force and to establish any remaining problems in the production processes. Products manufactured during ramp-up are sometimes supplied to preferred customers and are carefully evaluated to reveal any remaining flaws. The transition from ramp-up to volume manufacturing is usually gradual. At some point in this transition, the product is launched and becomes available for widespread distribution.

2.2.4 PCA Design Process Life Cycle

It is important for printed circuit designers to understand that board layout is just a part of the overall PCA design process, although that is particularly crucial to the successful development of a design.

Since PCAs form the basic building a block of most systems, the design process actually begins by defining the functions that each of these building blocks must perform. A design has little value until it is turned into operating, production-configuration hardware and integrated into a system or product. When integration has been implemented successfully, the design process is considered to be complete.

All the interrelated activities that occur within the boundaries defined by these two events define a PCA's design process life cycle It is generally recognised that the PCA design effort frequently paces a programme's overall hardware development schedule and constitutes a major part of its nonrecurring (engineering and design) cost.

Although it is not as well recognized, the actual board layout activity is not usually the largest cost or schedule element of a PCA design cycle It has been shown elsewhere that the costs associated with board layout activities are about 15 to 20 percent of the total PCA cost of design. Schedule time required for layout is about 12 to 15 percent of the total. The pie chart in Fig 2 graphically show the cost relationships that usually exist between the circuit design, layout and integration activities needed to produce a new PCA design.

Figure 2: Nonrecurring (design) cost

Figure 2: Nonrecurring (design) cost

2.2.5 Scope of the Design Process

As shown in the process flowchart in Fig 3, PCA design activities begin with the functional and physical partitioning of a system. Partitioning is a critical process step because the decisions made at this point will affect all downstream design and production activities.

Figure 3: PCA Design Progression

Figure 3: PCA Design Progression

Partitioning entails:

After circuitry has been allocated to an assembly, the specific operational requirement of that assembly should be identified, parametrically defined, and documented. This set of physical and circuit specifications becomes the basis for defining the PCA's design characteristics and will be used as the criteria for verifying its performance.

A detailed circuit design is then developed, resulting in the creation of a schematic diagram that identifies all electronic components required to implement the PCA's functions and describes how they are to be interconnected. The schematic also specifies the circuit interfaces between the PCA and the rest of the system/product.

Layout is usually performed with use of CAD tools, which may also be involved in the test stage.

Prototyping or modelling and simulation may be performed on all or portions of the design to predict how the circuitry will perform over its intended range of operating conditions. This activity also may serve to identify critical portions of the circuitry that will require special attention during physical layout of the printed board.

The last part of the design process involves integrating the PGA into its system/product. The basic purpose of this activity is to verify that the design will perform its intended functions in an operating environment.

[Back to Top]

2.3 Product Planning

2.3.1 Types of Product Development Projects

Product development can be categorised into four types:

New product platforms: This type of project involves a major development effort to create a new family of products based on a new, common platform. The new product family would address familiar markets and product categories. The development of a new, digital photocopier platform, is an example of this type of project.

Derivatives of existing product platforms: These projects extend an existing product platform to better address familiar markets with one or more new products. To develop a new photocopier based on an existing light-lens (not digital) product platform would be an example of this type of project.

Incremental improvements to existing products: These projects may only involve adding or modifying some features of existing products in order to keep the product line current and competitive. A slight change to remedy minor flaws in an existing photocopier product would be an example of this type of project.

Fundamentally new products: These projects involve radically different product or production technologies and may help to address new and unfamiliar markets. Such projects inherently involve more risk; however, the long-term success of the enterprise may depend on what is learned through these important projects. The first digital copier Xerox developed is an example of this type of project.

This is illustrated in Figure 4 by the Xerox photocopier product range.

Figure 4: Types of Product Development Project

Figure 4: Types of Product Development Project

2.4 Definition and Philosophy of Design for Manufacture (DfM)

A designer's primary objective is to design a functioning product within given economic and schedule constraints. However, research has shown that decisions made during the design period determine 70% of the product's costs while decisions made during production only account for 20% of the product's costs. Further, decisions made in the first 5% of product design could determine the vast majority of the product's cost, quality and manufacturability characteristics. This indicates the great leverage that DfM can have on a company's success and profitability.

Manufacturing cost is thus a key determinant of the economic success of a product. In simple terms, economic success depends on the profit margin earned on each sale of the product and on how many units of the product the firm can sell. Profit margin is the difference between the manufacturer's selling price and the cost of making the product. The number of units sold and the sales price are to a large degree determined by the overall quality of the product.

Economically successful design is therefore about ensuring high product quality while minimising manufacturing cost. DfM is one method for achieving this goal; effective DfM practice leads to low manufacturing costs without sacrificing product quality.

However, the application of DfM must consider the overall design economics. It must balance the effort and cost associated with development and refinement of the design to the cost and quality leverage that can be achieved. In other words, greater effort to optimise a products design can be justified with higher value or higher volume products.

Design effectiveness is improved and integration facilitated when:

2.4.1 Cross-Functional Teams

Design for manufacturing is one of the most integrative practices involved in product development. DfM utilises information of several types, including:

  1. sketches, drawings, product specifications, and design alternatives
  2. a detailed understanding of production and assembly processes
  3. estimates of manufacturing costs, production volumes, and ramp-up timing.

DfM therefore requires the contributions of most members of the development team as well as outside experts. DfM efforts commonly draw upon expertise from manufacturing engineers, cost accountants, and production personnel, in addition to product designers. Many companies use structured, team-based workshops to facilitate the integration and sharing of views required for DfM.

2.4.2 DfM Development Process

DfM begins during the concept development phase, when the product's functions and specifications are being determined.

When choosing a product concept, cost is almost always one of the criteria on which the decision is made - even though cost estimates at this phase are highly subjective and approximate. When product specifications are finalised, the team makes trade-offs between desired performance characteristics. For example, weight reduction may increase manufacturing costs. At this point, the team may have an approximate bill of materials (a list of parts) with estimates of costs.

During the system-level design phase of development, the team makes decisions about how to break up the product into individual components, based in large measure on the expected cost and manufacturing complexity implications. Accurate cost estimates finally become available during the detail design phase of development, when manufacturing concerns drives many more decisions.

2.4.3 Sample DfM Process

A DFM method is illustrated in Figure 10. It consists of five steps:

  1. Estimate the manufacturing costs
  2. Reduce the costs of components.
  3. Reduce the costs of assembly.
  4. Reduce the costs of supporting production.
  5. Consider the impact of DfM decisions on other factors.

Figure 10: DfM Method

Figure 1: DfM Method

 

As shown in Figure 1, the DfM method begins with the estimation of the manufacturing cost of the proposed design. This helps the team to determine, at a general level, which aspects of the design - components, assembly, or support are most costly. The team then directs its attention to the appropriate areas in the subsequent steps.

This process is iterative. It is not unusual to re-compute the manufacturing cost estimate and to improve the design of the product dozens of times before agreeing that it is good enough. As long as the product design is improving, these DfM iterations may continue even until pilot production begins.

DfM software tools are being increasingly utilised to facilitate this iterative process. Within the PCB industry, the Valor software toolset is used extensively for this purpose. (www.valor-us.com)

At some point, the design is frozen (or "released"), and any further modifications are considered formal "engineering changes" or become part of the next generation of the product.

Estimate the Manufacturing Costs: Step 1

Figure 11 shows a simple input-output model of a manufacturing system. The inputs include raw materials, purchased components, employees' efforts, energy, and equipment. The outputs include finished goods and waste. Manufacturing cost is the sum of all of the expenditures for the inputs of the system and for disposal of the wastes produced by the system2. Note that this latter requirement (i.e. cost) is becoming increasingly important in the PCB industry.

2 Note that this latter requirement (i.e. cost) is becoming increasingly important in the PCB industry.

Figure 11: Simple Model of a Manufacturing System

Figure 2: Simple Model of a Manufacturing System

As the metric of cost for a product, firms generally use unit-manufacturing cost, which is computed by dividing the total manufacturing costs for period (usually a quarter or a year) by the number of units of the product manufactured during that period. This simple concept, however, may be less than straightforward in practice.

Figure 12 shows one way of categorising the elements of manufacturing cost.

Figure 12: Elements of Manufacturing Cost

Figure 3: Elements of Manufacturing Cost

Manufacturing Costs

Under this scheme, the unit manufacturing cost of a product consists of costs in three categories:

Bill of Materials

Since manufacturing cost estimation is fundamental to DfM, it is important to keep this information well organised. Table 1shows a line entry for a typical Bill of Materials (BOM), without the associated cost information.

This approach to is widely used in electronic manufacturing industry to provide, in a list form, a record of each individual component in the product. The BOM is typically generated from the CAD system.

Table 1: Example BOM Entry
QTY DESCRIPTION REF DESG MANF DISTRIBUTOR
7 Diode, 1N5817, 1A, D1-D7 Diodes Inc: 1N5817MCT

Estimating the Costs of Standard Components

The costs of standard components are estimated by either:

  1. comparing each part to a substantially similar part the firm is already producing or purchasing in comparable volumes;
  2. soliciting price quotes from vendors or suppliers. The costs of minor components are usually obtained from the firm’s experience with similar components, while the costs of major components are usually obtained from vendor quotes.

In obtaining price quotes, the estimated production quantities are extremely important. For example, the unit price on a purchase of a dozen screws or inserts may be 10 times higher than the unit prices paid when purchasing 100,000 of these parts every month.

Some suppliers will design and fabricate a custom variation to a standard component if production quantities are high enough. For example, small electric motors, such as those found in powered hand tools, are often designed and built specifically for the product application. If the production quantities are high enough (say, 100,000 per year in this case), these custom motors are quite economical.

Estimating the Costs of Custom Components

Custom components, which are parts designed especially for the product, are made by the manufacturer or by a supplier. Custom parts are typically special-purpose parts, useful only in a particular manufacturer's products.

When the custom component is a single part, it is possible to estimate its cost by adding up the costs of raw materials, processing, and tooling.
While the design of a custom part or selection of a new part may be the most optimal approach to meet product requirements from the designer's point of view, it may not be the best overall approach for the company. Product cost and quality may be negatively affected by the proliferation of specialised items that require specialised capabilities or prevent efficient manufacture and procurement.

Estimating the Cost of Assembly

Products made of more than one part require assembly. In the assembly of electronic printed-circuit boards, it is normally principally done automatically, even at relatively low volumes. The cost may be estimated on a basis of machine time per component insertion and then summating for the whole product.

Manual assembly operations also exist, usually for non-standard electrical or mechanical fittings on the PCB (e.g. stiffeners). Summing the estimated time of each assembly operation and multiplying by a labour rate can estimate manual assembly costs.

Assembly labour, as an hourly rate, can cost very little in low-wage countries to a great deal more in industrialised nations.

Estimating the Overhead Costs

Accurately estimating overhead costs for a new product is difficult, and the industry practices are not necessarily very accurate.

The indirect costs of supporting production are very difficult to track and assign to particular product lines. The future costs of supporting production are even more difficult to predict for a new product. Most companies assign charges by adopting an overhead rate, based on the most accurate estimates available.

Reduce the Cost of Components: Step 2

For most highly engineered discrete goods (e.g. a PCB) the cost of purchased components will be the most significant element of the manufacturing cost. This section presents several strategies for minimising these costs. Many of these strategies can be followed even without the benefit of accurate cost estimates. In this case, these strategies become design rules, or rules of thumb, to guide DFM cost reduction decisions.

Understand the Process Constraints

Some component parts may be costly simply because the designers did not understand the capabilities and constraints of the production process.

For example, a designer may specify a small surface-mount component package without realising that physically inserting such a device into the PCB is outwith the capabilities of the pick-and-place machine. In a similar fashion, a designer may specify dimensions with excessively tight tolerances, without understanding the difficulty of achieving such accuracy in production.

It is often possible to redesign the part to achieve the same performance while avoiding costly manufacturing steps; however, to do this the design engineer needs to know what types of operations are difficult in production and what drives their costs. This lack of knowledge of the manufacturing function is by no means uncommon amongst designers.

In some cases, the constraints of a process can be concisely communicated to designers in the form of design rules. When this is possible, designers can avoid exceeding the normal capabilities of a process and thereby avoid incurring unusually high costs.

For processes whose capabilities are complex (e.g. highly-automated surface-mount assembly), the best strategy is to work closely with the people who deeply understand the production process. These manufacturing experts will generally have plenty of ideas about how to redesign products to reduce production costs.

Redesign Products to Eliminate Processing Steps

Careful scrutiny of the proposed design may lead to suggestions for redesign that can result in simplification of the production process. Reducing the number of steps in the fabrication or assembly process generally results in reduced costs as well.

Some process steps may simply not be necessary. For example, the component legend (known as Silkscreen) may not be required on a high-density surface-mount with no filed-service requirement.

In some cases, several steps may be eliminated through substitution of an alternative process step.

Choose the Appropriate Economic Scale for the Part Process

The manufacturing cost of a product usually drops as the production volume increases. This phenomenon is labelled economies of scale.

As production volumes increase, the firm may be able to justify a more costly part implementation, which, for example, required the purchase of a specialised machine.

Standardise Components and Processes

The principles of economies of scale also apply to components. As production volume increases, the unit cost of components decreases. For a given expected product volume, the benefits of substantially higher component volumes can be achieved through the use of standard components.

Standard components are those common to more than one product. This standardisation may occur within the product line of a single firm or may occur, via an outside supplier, across the product lines of several firms. For example, the use of the 1.8-litre turbo-diesel engine in several Ford cars is an example of internal standardisation. The use of a common 9 volt power supply across several manufacturers power toll products is an example of external standardisation. In either case, all other things being equal, the component unit cost is lower than if the component were used in only a single product.

Reduce the Costs of Assembly: Step 3

Design for Assembly (DfA) is a fairly well established subset of DfM that involves minimising the cost of assembly. Often as a result of emphasis on DfA, the overall parts count, manufacturing complexity, and support costs are all reduced along with the assembly cost.

Boothroyd and Dewhurst (1989) advocate maintaining an ongoing estimate of the cost of assembly. In addition to this absolute score, they propose the concept of assembly efficiency. This is measured as an index that is the ratio of the theoretical minimum assembly time to an estimate of the actual assembly time for the product. This concept is useful in developing an intuition for what drives the cost of assembly.

Maximise Ease of Assembly

Two products with an identical number of parts may nevertheless differ in required assembly time by a factor of two or three. This is because the actual time to grasp, orient, and insert a part depends on the part geometry and the required trajectory of the part insertion. The ideal characteristics of a part for an assembly are (adapted from Boothroyd and Dewhurst, 1989):

Reduce the Costs of Supporting Production: Step 4

In working to minimise the costs of components and the costs of assembly, the team may also achieve reductions in the demands placed on the production support functions.

For example, a reduction in the number of parts reduces the demands on inventory management. A reduction in assembly content reduces the number of workers and/or machines required for production and therefore reduces the cost of supervision and human resource management. Standardised components reduce the demands on engineering support and quality control.

Minimise Systemic Complexity

An extremely simple manufacturing system would utilise a single process to transform a single raw material into a single part-perhaps a system extruding a single diameter of plastic rod from plastic pellets. Unfortunately, few such systems exist. Complexity arises from variety in the inputs, outputs and transforming processes.

Many real manufacturing systems (e.g. mobile telephone manufacture) involve hundreds of suppliers, thousands of different parts, hundreds of people, dozens of types of products and many types of production processes. Each variant of suppliers, parts, people, products, and processes introduces complexity to the system.

These variants must usually be tracked, monitored, managed, inspected, handled and inventoried at considerable cost to the enterprise. However, much of this complexity is driven by the design of the product and can therefore be minimised by intelligent design decisions.

Impact of DfM on Other Factors: Step 5

Minimising manufacturing cost is not the only objective of the product development process.

The economic success of a product also depends on the quality of the product, the timeliness of product introduction and the cost of developing the product. There may also be situations in which the economic success of a project is compromised in order to maximise the economic success of the entire enterprise. In contemplating a DfM decision, these issues should be considered explicitly.

Impact of DfM on Development Time

Development time can be precious. For a large development project, time may be worth as much as several hundred thousand pounds per day. For this reason, DfM decisions must be evaluated for their impact on development time as well as for their impact on manufacturing cost. While saving £1 in cost on each one of a particular part would be worth perhaps £1 million in annual cost savings, it would almost certainly not be worth causing a six-month delay in a large development program.

The relationship between DfM and development time is complex. As such, the cost benefits of the DfM decision may not be worth the delay in project duration. This is particularly true for products competing in dynamic markets (e.g. computer peripherals).

Impact of DfM on Development Cost

Development cost closely mirrors development time. Therefore, the same caution about the relationship between DfM and development time applies to development cost. In general, however, teams that aggressively pursue low manufacturing costs as an integral part of the development process seem to be able to develop products in about the same time and with about the same budget as teams that do not. Part of this phenomenon certainly arises from the correlation between good project management practices and the application of sound DfM methods.

Impact of DfM on Product Quality

Before proceeding with a DfM decision, the team should evaluate the impact of the decision on product quality. Under ideal circumstances, actions to decrease manufacturing cost would also improve product quality. It is not uncommon for DfM efforts focused primarily on manufacturing cost reduction to also result in improved serviceability, ease of disassembly, and recycling.

However, in some cases actions to decrease manufacturing cost can have adverse effects on product quality (such as reliability or robustness), so it is advisable for the team to keep in mind the many dimensions of quality that are important for the product. Embracing the broader philosophy of Design For Excellence (DfX) helps to minimise this impact.

Impact of DfM on External Factors

Design decisions may have implications beyond the responsibilities of a single development team. In economic terms, these implications may be viewed as externalities. Two such externalities are component reuse and life cycle costs.

[Back to Top]

2.5 Design for Fabrication Guidelines (PCB Specific)

A key aspect of efficient design for fabrication practice is to develop a good relationship between the board designer and fabricator. This means that both parties are familiar with each other’s activities; maintain close communication and exchange appropriate information throughout the process

2.5.1 Board Size and Shape

PCB assemblers have a maximum size board they can handle. The PCB fabrication house's panel size is also important when mass-producing boards.

In terms of cost efficiency, it is desirable to fit as many boards as possible on the largest panel with as little wasted space (i.e. scrap) as possible. Normal board spacing for routing (the means by which boards are separated on a panel) is 0.3", plus there is typically a 1.0" to 2.0" border on the board for handling it during processing. This is illustrated in Figure 13.

Figure 13: 4 PCBs on one process panel

Figure 4: 4 PCBs on one process panel

As the above suggests, a modification to the size of an individual board can have a significant impact on the number of boards contained within the panel.

The panels should be designed for routing with little manual intervention. Solid panel designs are preferred over routed slots for structural stability, but solid panel designs require depaneling technologies such as laser, water jets or machine routing.

In some cases, grooves may be used for depaneling for manual separation. Instead of grooves, unplated holes also are used between pre-routed slots to separate these boards. Regardless of the depaneling technology used, depaneling costs must be taken into account.

It is thus essential that the concept and practical realisation of fabrication and assembly panelisation be well understood by the layout designer.

2.5.2 Board Thickness

Board thickness may also be specified. A standard thickness and type of board is .062" FR4 (standard epoxy resin with woven glass reinforcement). Other typical board thickness is .010", .020", .031", and .092"

2.5.3 Board Warpage

Board warpage must be minimised as it causes a number of problems, including insertion errors and laminate breakage. This is illustrated in Figure 14.

Figure 14: Board Warpage

Figure 5: Board Warpage

For Multilayer boards, a balanced construction is essential for realising a perfectly flat result. This is illustrated in Figure 15.

Figure 15: Balanced Multilayer Construction

Figure 6: Balanced Multilayer Construction

2.5.4 Board Location References

Of prime importance in board design and construction is the establishment of accurate datum points. Datum holes provide the locating references to which all holes are drilled or punched.

These references are also used to accurately position the board on the workboard holder and reduce printed circuit board tolerance accumulation. This is shown in Figure 16.

Figure 16: Datum Hole Position

Figure 7: Datum Hole Position

2.5.5 Base Copper

The amount of copper used on the board external and internal (if Multilayer construction) is referred to in ounces per square foot. There is a direct correlation between copper weight and thickness.

Normally, 0.5 oz copper is preferred on outer layers with 1 oz on internal layers. This is illustrated in Figure 17.

Figure 17: Copper Weight and Thickness

Figure 8: Copper Weight and Thickness

Note that increasing the weight 2 oz could add as much as 20% to the material cost.

2.5.6 Trace Width and Spacing

The chemical and photographic processes used to produce a PCB put restrictions on the minimum width of trace (i.e. the copper) and the minimum spacing between traces. If a trace is made smaller than this minimum width there is some chance it will open (no connection) when manufactured.

The other issue with trace width relates to the heat produced by current passing through the copper track. If the track is of insufficient width, there is a danger of overheating and possible destruction of the track. Values for track thickness in relation to current handling and heat dissipated are well documented.

If two traces are closer together than the minimum spacing there is some chance they will short when manufactured. This may also be the case if the voltage difference between the traces is of sufficient magnitude.

These parameters are usually specified as "x/y rules", where x is the minimum trace width and y is the minimum trace spacing. For example, "8/10 rules" would indicate 8 mil minimum trace width and 10 mil minimum trace spacing. These rules (especially spacing) apply to any metal on the PCB, including pad to track spacing.

Typical modern process rules are 8/8 rules with values as small as 2/2 rules being available.

In a general sense, it is important to balance the circuit pattern as evenly as possible across the board’s surface. This helps achieve uniform plating and avoid warpage.

2.5.7 Pad Sizes

The biggest issues with pad size are solderability and manufacturability.

Solderability is the ability to form a good quality solder joint between the pad and the component leg. It is related to factors such as the amount of heat applied, the cleanliness of the surface and the contact area.

Manufacturability is concerned with, for through-hole designs, whether or not the pad will be broken when the hole is drilled in it. This is mainly a function of the accuracy of the PCB manufacturer's drilling. If a drill hole is slightly off centre the pad may be broken at one edge possibly leading to an open in the circuit.

This is shown in Figure 18.

Figure 18: Pad and Annular Ring Relationship

Figure 9: Pad and Annular Ring Relationship

2.5.8 Hole Sizes

Most PCB manufacturers have a wide selection of drill (hole) sizes available. Some charge per drill size used, others offer a standard set of drill sizes for no charge and then charge for any non-standard drill sizes. It is important to note that the plate-through will cause the hole to effectively be narrower.

2.5.9 Solder Coating and Soldermask

Solder coating is typically applied to all component location points on a PCB. For bare copper areas where solder is not applied, such as the tracks, soldermask is applied to provide protection and insulation.

There are a number of process issues with soldermask that a designer should consider. When covering large copper areas, there is a danger of ‘blistering’ of the soldermask when heat is applied (e.g. during reflow).

Material thickness and feature resolution are critical soldermask characteristics and are interrelated (thinner material allows finer feature resolution). High density circuit boards require that very tight control be exercised over the position and dimensional accuracy of the soldermask features.

2.5.10 Gold Plating

Nickel-gold plating is commonly applied to edge-connectors where contact resistance must be kept to a minimum. This procedure adds costs to the board.

2.5.11 Silkscreen

This is textual or numeric information applied to the surface of the PCB. In today’s high-density designs, it can cause problems for the board fabricator due to the area it consumes. All such board markings must be legible and must be able to withstand all manufacturing processes and environmental conditions to which they are exposed. They must also be visible after all parts are installed.

Unless the board is to be manually assembled or there is a particular field-service requirement, it may be the case that silkscreen provides no real function and can be discarded.

[Back to Top]

2.6 Design for Assembly Guidelines (PCB Specific)

The technology used for mounting and attaching components o a PCB determines the basic processes that will be employed to produce a finished assembly.

Through-hole assemblies are the easiest and also may be least costly to produce. This technology is usually applied to designs where there are no size constraints or where a circuit requires a majority of parts that are only available in a through-hole configuration (such as transformers, large capacitors, inductors, connectors, etc.). Through-hole boards can be assembled manually or with automated equipment.

Assemblies that use surface-mount technology (SMT) mixed technology are much more difficult to build, requiring complex processes that are implemented with costly, automated equipment. While through-hole assemblies usually have parts mounted to one side of a board, SMT or mixed assemblies may have parts on both sides.

Also, the physical size and configuration of surface-mounted packages and their terminations require that the processes used to attach and connect them to a circuit board be very tightly controlled.

However, only automatic component insertion provides the consistency required to ensure the highest levels of circuit board quality, throughput, and process control.

When properly planned and implemented, automatic component insertion provides significant cost savings in the printed circuit board assembly process. The benefits realised from automating circuit board assembly processes span all areas of manufacturing. Ultimately, all of these benefits result in improved products and reduced production costs and, as such are very much the norm in today’s electronic industry.

Three key inputs affect the economics and logistics of PC board assembly:

  1. the circuit boards;
  2. the equipment used to assemble the boards;
  3. the components to be inserted.

By precisely understanding and standardising these three primary elements, manufacturers can improve the quality of the finished boards, increase the throughput of the assembly equipment and the system and more precisely define the process control standards to provide a basis for future applications. Standardising these elements reduces process variability and eliminates unnecessary process steps, which leads to increased insertion reliability, improved product quality, enhanced system price/performance and, ultimately, reduced production costs.

As surface mount technology now dominates the industry, it is this approach that will be the focus here.

2.6.1 Axis Considerations

An important consideration for manufacturing is proper component alignment on the printed circuit board. The guidelines for component orientation vary depending on type. Components of similar types should be aligned in the same orientation for ease of component placement, inspection and soldering.

Orienting similar components in the same direction is very desirable for ease of programming pick-and-place equipment. Orienting pin 1 of all devices in the same direction (e.g. North) would be key; however, this is not always feasible because of routing constraints. Therefore, pin 1 orientation should be limited to a maximum of two directions on the board. Uniform orientation also facilitates component inspection for misplacement and is very helpful for troubleshooting during testing.

Insertion machines are normally capable of inserting components at 0 or 90 degrees. This is shown in Figure 19 (note through-hole technology).

Figure 19: Component Orientation for Automated Assembly

Figure 10: Component Orientation for Automated Assembly

2.6.2 Fine Pitch Parts

Fine pitch SMT parts (such as Quad Flat-Packs [QFP]) can cause serious alignment problems for pick-and-place machines. It is now preferred to replace such packages with standard ball-grid arrays (BGA).

2.6.3 Insertion Head Clearance

It is essential that there is sufficient clearance around components to accommodate the insertion head of the placement machine. Layout designers frequently overlook this aspect of assembly.

2.6.4 High Defect Processes

The highest defect rates are found in processes such as hand soldering; wave soldering, excessive part handling and other secondary operations. This implies that the best practice involves only reflowed surface-mount parts with zero operator contact.

2.6.5 Single-Sided Board Mounting

For SMT designs, it is preferable to mount all components on one board side. Apart from making the assembly process simpler, it also facilities board test.

2.6.6 PTH to SMT Conversion

Converting as many plated-through-hole (PTH) as possible to SMT is a positive manufacturability step, as surface mount parts are smaller, require little or no manual handling, consume less board area and are readily placed by automated methods.

2.6.7 CTE Issues

Plastic packages provide a better CTE (Coefficient of Thermal Expansion) match with the circuit board than, for example, ceramic packages. This results in less strain being applied to the interface point between board and package and thus minimises solder joint failure.

2.6.8 Mechanical Hardware

Almost without exception, PCBs use mechanical parts and hardware. These include stiffeners, supports, component tie-downs, heat sinks, board extractors, terminals, eyelets and insulators.

When specifying the installation of these parts and hardware, the layout designer should be aware of some basic issues and requirements that are unique to these types of component. The forces that are employed to fit such parts can cause damage to board conductors or dielectric material.

Mechanical stress should also be considered when designing provision for board extraction. Plug-in connectors with large pin counts require large extraction forces for removal. This may require that the board be reinforced at the point of attachment.

Mechanical assembly of a heat sink is particularly critical, since it frequently requires the establishment of a good thermally conductive path between a part and the board, whilst being electrically insulated. This may require the provision of a thermally conductive adhesive or pastry, which must be compatible with other board materials and coatings.

Other issues such as part height and clearance, along with the assembly method (manual or automatic) and the stage at which they are fitted are of importance and should be considered by the layout designer.

2.6.9 Conformal Coatings

These coatings are part of the assembly process and are used to protect board circuitry from potentially damaging operating environments. They are especially applicable for complex, high-density designs.

Properly selected and applied conformal coatings will prevent degradation of performance due to the effects of humidity, dust and dirt, vibration and shock, corrosive chemicals, abrasion, fungi and stray conductive particles.

It is very important that the coating material be compatible with the surfaces and materials it makes contact with. Consideration should also be given to field repair or rework. Some coatings can be very difficult to remove, especially for the localised operations described.

[Back to Top]

2.7 Summary

During the 1980s, design-for-manufacturing practices were put into place in thousands of firms. Today DFM is an essential part of almost every product development effort. No longer can designers "throw the design over the wall" to production engineers.

In essence, comparing new product designs with those of earlier generations usually results in the identification of fewer parts, the use of new materials, the employment of more integrated parts (such as chips), the use of higher-volumes of standard parts and simpler assembly procedures.

Design for manufacturing (DfM) is aimed at reducing manufacturing costs while simultaneously improving (or at least not inappropriately compromising) product quality, development time, and development cost. The key element to the success of such an approach is that the organisation, as a whole, embraces a culture of simultaneous engineering. This is the cornerstone of effective DfM and essentially translates into all elements of the product development process working closely together.

Design for manufacturability (DfM) is the practice of designing board products that can be produced in a cost-effective manner using existing manufacturing processes and equipment.

The benefits of a manufacturable design are better quality, quicker time to market, lower labour and material costs, shorter throughput time, and fewer design iterations. The quickest way to waste large amounts of money is to produce an PCB design that cannot be assembled, repaired or tested using existing equipment. DfM essentially is a yield issue, hence a cost issue. It plays a critical role in reducing defects in printed circuit boards.

DfM is gaining more recognition as it becomes clear that manufacturing engineers alone cannot control cost reduction of PCBs. The printed circuit board designer also plays a critical role in cost reduction. The days of throwing the design over the fence to the manufacturing engineer are gone — if they ever really existed.

It is important to bear in mind that DfM alone cannot eliminate all defects in PCBs. Defects fall into three major categories:

  1. design-related problems;
  2. incoming material-related problems such as PCB, adhesive, solder paste, etc.;
  3. problems related to manufacturing processes and equipment.

Each defect should be analysed for its source to permit appropriate corrective action to be taken.

Design trade-offs determine whether a PCB meets cost, real estate and reliability constraints. They also should be considered an integral part of DfM.

It is not an exaggeration to say that the design team plays a critical role in the success and financial viability of any PCB product.

This is true in the through-hole as well surface-mount design, but not to the same degree. There is only one way to assemble a through-hole board: populate it (manually and/or automatically) and wave solder it.

In SMT, however, the designer has many options, depending on the assembly type. For example, in SMT, the same component can be placed on the top or bottom of the board, with different manufacturing consequences in each case. Clearly the design team must be thoroughly familiar with surface mount manufacturing processes.

This is not to belittle the importance of DfM in through-hole assemblies because DfM for auto-insertion is important. But unlike through-hole technology, SMT does not have the option of manual placement, and the rules for testability are very different. In short, DfM and the role of the design team in the success of SMT fabrication and assembly cannot be overemphasised.

[Back to Top]

2.8 Appendix 1: Simple DFM Evaluation Document

Appendix 1


2.9 Self Assessment Questions

Question 1

Sketch the organisation (in some appropriate graphical representation) of a consulting firm that develops new products for clients on a project-by-project basis. Assume that the individuals in the firm represent all of the different functions required to develop a new product. Would this organisation most likely be aligned with functions, be aligned by projects, or be a hybrid?

Question 2

Is there an analogy between a university and a product development organisation? Is a university a functional or project organisation?

Question 3

List 10 reasons why reducing the number of parts on a PCB might reduce production costs. Also list some reasons why costs might increase.

 

Question 4

Consider the following 10 "design rules" for electromechanical products. Do these seem like reasonable guidelines? Under what circumstances could one rule conflict with another one? How should such a trade-off be settled?

  1. Minimise parts count
  2. Use modular assembly
  3. Stack assemblies
  4. Eliminate adjustments
  5. Eliminate cables
  6. Use self-fastening parts
  7. Use self-locating parts
  8. Eliminate reorientation
  9. Facilitate parts handling
  10. Specify standard parts

 

Question 5

Is it practical to design a PCB with 100 percent assembly efficiency (DFA index = 1.0)? What conditions would have to be met?

 

Question 6

Is it possible to determine what a product really costs once it is put into production? If so, how might you do this?

 

Question 7

Can you propose a set of metrics that would be useful to predict changes in the actual costs of supporting production? To be effective, these metrics must be sensitive to changes in the PCB design that affect indirect costs experienced by the firm. What are some of the barriers to the introduction of such techniques in practice?

 

Question 8

Make a close inspection of either a PCB on its own, or a product containing a PCB. Scrutinise the board design and produce a list of suggestions as to how it might be better fabricated and assembled.

 

show answers

 

Site Search

Powered by Google
Site Map