Concepts of Printed Circuit Design
Unit 13: Design for excellence - SAQ Answers
Contents
Question 1
Question
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?
Answer
One possible diagram of a consulting firm organisation:
Consulting Firm Organisation
Consulting firms almost always have a strong project focus and are either project organisations or heavyweight project matrix organisations. Teams form and dissolve almost daily in order to meet the needs of new client engagements. In some firms, there will also exist some measure of functional organisation. For example, there may be a member of the firm responsible for ensuring that the mechanical engineers have the software tools they need or that the marketing professionals know the latest research techniques. Nevertheless, the dominant organisational structure is the project.
Question 2
Question
Is there an analogy between a university and a product development organisation? Is a university a functional or project organisation?
Answer
We can make the analogy between a product development organisation and a university by considering whether there are university equivalents to both the product development process and the product development organisation.
Although a university has many purposes, its primary purpose is the development of undergraduate students. The student development process is analogous to the product development process. The student development process takes as its (typical) input a secondary school pupil whose mission statement is to get an undergraduate degree.
Its output, instead of a product launch is the graduation of the student. Along the way there is a well-defined series of steps complete with milestones as in the product development process. Although the detailed series of steps in the student development process is different than in the product development process, the advantages of the well-defined process are similar. In both cases the process ensures the quality of the output of the process (student), the co-ordination of the resources on the development team (faculty/school, registry, physical plant), the timely completion of the project (graduation in 4 years), the management of the project, and the improvement of the process.
Having established the analogy between the product development process and the student development process, the next step is to look at the analogy between the product development organisation and the university. The two primary organisational types are the functional organisation where the primary links are among those who perform similar functions and the project organisation where the primary links are among those who work on the same project. A university is a functional organisation. Some of these functions include:
- faculty, whose primary role is to teach students.
- registry, who admit students to the process.
- physical plant, who maintains the facilities.
There are certainly other functions that make up a university but using just these three as an example we can see that they are indeed organised by the function they perform. Although they are all part of the development of the "product", the student, their strongest ties are to those who share the same function. As in a product development organisation, those sharing the same function are usually located near each other and report to the same manager.
If a more detailed view is taken of the organisation of the faculty, it may be seen that within the faculty the organisation continues to be a functional organisation. Breaking it down further we see:
- faculty, who all teach students.
- schools, who teach students in a particular field, e.g., engineering.
- subject groups, who teach students in a still more specified field, e.g., mechanical engineering.
The great strength of a functional organisation is that it maintains deep expertise in the functional areas. The great weakness is that the efficiency of co-ordination among the functions is not as good as it could be. This characteristic is seen among universities as well. A university’s great strength is the vast amount of knowledge contained in its areas of expertise, not its ability to co-ordinate among its different functions.
Question 3
Question
List 10 reasons why reducing the number of parts on a PCB might reduce production costs. Also list some reasons why costs might increase.
Answer
Ten important advantages of part count reduction are:
(1) Reduction of assembly time and cost:
This is because of
- lower number of parts to be assembled
- lower number of re-orientations, opportunity to eliminate difficult alignments and obstructed access for insertion of parts/fasteners
- less time spent inserting and securing fasteners
(2) Lower manufacturing costs:
Most often, the integral part will cost less to manufacture than the sum of the component costs so the manufacturing costs are reduced.
(3) Lower set-up costs:
Reducing the number of parts would mean that non-value-added costs like setup times would be reduced.
(4) Lower co-ordination costs:
A reduced number of parts would mean that the amount of co-ordination and bookkeeping required to get the parts to the right place at the right time (usually assembly) is reduced. This would mean less strain on the component tracking system, like bar coding systems.
(5) Inventory management:
The complexity of inventory management is greatly reduced with fewer parts. This would particularly affect in-process inventory. Also, the amount of shelf space required for inventory is reduced, leading to additional cost savings.
(6) Higher degree of automation possible:
The focus on reducing the number of parts would likely also take into account design for automated assembly. Use of these principles could enable the use of automated assembly, further reducing costs.
(7) Labour savings:
The combined effect of higher automation and reduced assembly time would lead to lower labour costs.
(8) Precise manufacture of complex parts:
Integrating component functions into fewer parts allows for enhanced control over the interaction between critical geometric features, since the geometry can now be governed by manufacturing and not by assembly.
(9) Purchasing and handling of materials:
Reducing the number of parts would reduce the burden on material purchasing and handling systems. This also reduces the complexity of parts routing and work order scheduling on the shop floor, leading to greater utilisation of the manufacturing system.
(10) Continuous improvement:
One of the greatest benefits of a drive to reduce the number of parts in a product is that it forces the team to reflect on the various aspects of the product. This might have far-reaching benefits to the quality of the product.
Drawbacks
One potential drawback of an effort to reduce the number of parts is that such a program can lead to the development of much more complex and intricate parts, whose design, development, and manufacture may be more difficult, slow, and expensive. For instance, the design team for the integrated part would now have to interact and iterate their design with a lot more product development teams. This might drive up the development costs, and delay the introduction of the product to the market. Also, in many cases the manufacturing cost of a larger number of simpler parts (which may be standard parts) can be less than the manufacturing costs of one integrated part.
To reconcile the benefits and drawbacks of parts count reduction, a development team must consider the total cost of the product, not just the assembly and overhead costs. Furthermore, these detail design decisions must be closely coupled with the product strategy, manufacturing capabilities, and specific project goals.
Question 4
Question
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?
- Minimise parts count
- Use modular assembly
- Stack assemblies
- Eliminate adjustments
- Eliminate cables
- Use self-fastening parts
- Use self-locating parts
- Eliminate reorientation
- Facilitate parts handling
- Specify standard parts
Answer
The given design rules are reasonable "guidelines" but they cannot be interpreted as hard-and-fast rules which are never to be broken. In fact, situations where one rule conflicts with another are quite common. For example, "minimise parts count" could potentially conflict with every one of the other rules. When this happens, the trade-off can be settled by considering the issues that really matter, such as: total cost of the product, repair and servicing, recyclability, product performance and functionality, minimisation of variations, etc.
Below are two such conflicting scenarios:
- Minimising Parts vs. Specifying Standard Parts:
While it is generally less expensive to use standard parts rather than custom components, sometimes a single custom component can replace many standard parts. When one or more custom components would be required to achieve the desired product functionality, in these cases it makes sense to attempt to integrate other component functions into the custom components.
- Minimising Parts Count vs. Using Modular Assembly:
In order to create a modular assembly, generally more parts are required to achieve the modular interfaces. Of course, modularity offers many advantages as discussed in the chapter "Product Architecture", and these benefits may outweigh any cost penalty of the additional parts required to achieve modularity.
Question 5
Question
Is it practical to design a PCB with 100 percent assembly efficiency (DFA index = 1.0)? What conditions would have to be met?
Answer
It might seem theoretically practical to produce a PCB with 100% assembly efficiency. However, because DFA has implications on other parts of a design and manufacturing process (that is, there is coupling between these issues), a 100% assembly efficient product can be counterproductive in several respects.
One drawback of 100% efficiency is part replacement. The integration of components that is inherent to DFA could lead to having to replace a large expensive component if a small section of a part fails. The customer might be extremely upset if it is necessary to replace a $200 part simply because a dry joint on the PCB could not be accessed.
Another problem with 100% assembly efficiency is customer maintenance. For example, imagine a product that requires occasional cleaning as a maintenance procedure. If the product is highly integrated there is a possibility that dust or dirt will accumulate in sections that can't be pulled apart or that require special tooling to separate.
From a cost standpoint it might not be practical to produce a product with 100% assembly efficiency. This requires incurring high fixed costs for expensive equipment and/or tooling. If the product volume is not high enough, the savings in assembly costs does not justify the expense incurred in fixed costs. Similarly, it might be the case that labour cost (the driving factor in manual assembly) is so low, that the time saved in simpler assembly procedures is not significant.
To achieve a product with 100% assembly efficiency it is necessary to integrate components to the level where only parts that need to move relative to each other or that need to be of different materials remain separate. It also requires parts that need not be oriented when assembling, which might be difficult if complex shapes are involved
Question 6
Question
Is it possible to determine what a product really costs once it is put into production? If so, how might you do this?
Answer
There are two broad limitations that make it impossible to determine what a product really costs in production. These areas follows:
- Variation:
Any system as complex as a manufacturing process is subject to many types of variation. Parts being produced by the process will have variations in true cost, For example, there are variable true labour rates due to workers’ differing performance and costs to the firm, variable costs of energy to power machines during various periods of the day, product liability costs incurred by faulty products, etc. These variations are so many and so complex that it is impossible to track them all. Therefore, it is not really possible to determine the exact cost of even the simplest products. The best we can hope for is some aggregate average cost.
- Shared Resources:
Many products flow through a production system, and so there are many resources they must share, including production personnel, machinery, floor space, engineering attention, etc. It is very difficult to allocate shared resource expenses properly to each product, which is why we simply "charge overhead" instead. This can only be accurate in the aggregate, not for each individual product. However, by using activity-based cost accounting techniques, we can aggregate less than with traditional accounting systems, and we can obtain somewhat higher accuracy of tracking costs.
Fortunately, the exact cost of a product is not necessary for management to be able to make good decisions. These decisions rely heavily on the specific and general knowledge associated with the systems and members of the organisation that are involved in formal or informal collection of production cost data. For a facility with a single product with dedicated equipment and a single customer, measuring the monthly or quarterly expenses associated with operating the plant and dividing by the number of units produced will give a very accurate estimate of the unit cost of the product. While the process becomes considerably more difficult for a complex plant with multiple product lines and considerable shared indirect resources, good management is still able to make effective decisions if they are attentive to appropriate information and they have a sound business strategy.
Question 7
Question
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?
Answer
Support costs are related to the demands on production support functions. To predict changes in these costs due to a particular product design, we can consider the product’s demands on the drivers of support costs. Certainly the categories of support functions and their associated drivers would be different for each firm, however some typical ones might be:
| Support Function | Cost Driver |
|---|---|
| Inspection and test | number of parts to be tested |
| materials handling | size or weight of component parts |
| purchasing | number of new parts or new vendors |
| manufacturing engineering | number of process steps |
| equipment maintenance | processing time |
| training | number of new processes |
In order to utilise such a set of metrics to predict costs, the charges (rates) associated with each driver must be known. The firm can divide its actual support costs by the total volume of each driver. For instance, if the firm experiences £300,000 in training costs in the year, and its products involved 30 new processes over that period, then the rate for this driver would become £10,000 per new process. Future designs could use this information.
In fact, the scheme above is basically the way that an activity-based costing (ABC) system functions. Many firms have now adopted such techniques and feel comfortable about their ability to track costs of production better than with the traditional systems. However, the implementation of such a scheme requires difficult decisions to be made about which drivers are important and about which of the firm's costs should be associated with each driver. If these decisions are not clear, the ABC system may not prove more accurate than earlier systems.
Question 8
Question
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.
Answer
By following the fabrication and assembly design guidelines in the unit, you should produce a list of manufacturabilty improvements for the selected PCB.
It would also be beneficial to consider the costings applied to these suggested changes as well as any implications on product quality or process variation.