Detailed Design, Prototyping and Implementation

Introduction

The transition from a project concept to a working electrical-engineering prototype is a design activity, not merely an assembly exercise. At this stage, requirements are converted into component values, circuit schematics, printed-circuit-board layouts, firmware, mechanical arrangements, test procedures, and manufacturing records. Decisions made here strongly influence safety, performance, cost, reliability, maintainability, and the credibility of the final dissertation.

A successful prototype must do more than operate once. It should perform its intended function under stated conditions, provide evidence that requirements have been met, be safe to operate, and be sufficiently documented for another engineer to inspect, reproduce, test, or modify it. Design-for-testability (DFT), design-for-manufacturability (DFM), design-for-reliability (DFR), and design-for-maintainability (DFMnt) should therefore be considered from the beginning rather than added after construction.

This chapter presents a structured approach to detailed design, component selection, schematic and PCB development, firmware or control implementation, integration, and prototype evaluation.

Learning Objectives

By the end of this chapter, a student should be able to:

• translate system requirements into verifiable electrical specifications;
• select components using electrical, thermal, mechanical, safety, availability, and lifecycle criteria;
• produce readable schematics and manufacturable PCB layouts;
• estimate power dissipation, voltage margins, current demand, and basic thermal requirements;
• implement and test firmware or control logic using suitable verification practices;
• integrate hardware, software, mechanical parts, and power sources systematically;
• apply DFT, DFM, reliability, maintainability, and safety principles;
• document a prototype so that its performance and limitations can be defended academically.

1. From Requirements to a Detailed Design

A requirement is useful only when it can be interpreted consistently and verified. A statement such as “the device should be low power” is incomplete until a limit, operating condition, and measurement method are defined. A stronger requirement might be: “The prototype shall consume no more than 2 W from a 12 V supply during continuous operation at an ambient temperature of 25 °C.”

A requirements traceability matrix links each requirement to a design element and a verification method.

Requirements should be classified as functional, performance, interface, environmental, safety, regulatory, and maintenance requirements. Each should have an owner, a verification method, and a status. This prevents a common project failure: a prototype that demonstrates a function but cannot demonstrate compliance with the original specification.

A useful design decomposition is:

1. system boundary and interfaces;
2. power-input and protection stage;
3. sensing or input stage;
4. processing or control stage;
5. actuation or output stage;
6. communication and user-interface stage;
7. mechanical packaging and thermal path;
8. test points and diagnostic facilities.

The decomposition should preserve clear interfaces. For example, the sensor subsystem may specify an analogue range of 0–3.0 V, the controller may accept 0–3.3 V, and the firmware may represent the measurement in degrees Celsius. These interfaces must be documented rather than assumed.

2. Component Selection as an Engineering Decision

Component selection is a constrained engineering decision. The lowest purchase price is rarely the correct criterion by itself. Selection should consider:

• voltage, current, frequency, accuracy, and dynamic range;
• tolerance, temperature coefficient, drift, and ageing;
• transient, reverse-polarity, and fault withstand capability;
• package, assembly method, heat dissipation, and board area;
• availability, approved suppliers, obsolescence risk, and lead time;
• software support, documentation, and development tools;
• compliance with relevant safety and electromagnetic-compatibility requirements;
• repairability and replacement cost.

2.1 Electrical derating

A component should not normally be operated at its absolute maximum rating. If a resistor is rated at 0.25 W and the design calculation predicts 0.20 W, the remaining margin may be inadequate when temperature, tolerance, and abnormal conditions are considered.

For a resistive component, power is calculated using:

where P is power in watts, V is voltage in volts, I is current in amperes, and R is resistance in ohms. These relationships assume an ideal resistor and steady-state electrical quantities. For non-sinusoidal or time-varying signals, average power may need to be calculated over a period.

A simple derating factor may be defined as:

where D is the fraction of the rated power being used. This ratio is dimensionless. A design policy might require D ≤ 0.5, although the appropriate value depends on the component, ambient temperature, reliability target, and manufacturer data. Derating is not a substitute for checking the manufacturer’s temperature-dependent rating curve.

2.2 Power-budget example

Suppose a controller contains a microcontroller consuming 120 mA at 3.3 V, a sensor consuming 20 mA at 3.3 V, and a communication module consuming 80 mA at 5 V. The nominal load power is:

If the 5 V rail is produced from a 12 V source by a regulator with efficiency 85%, its input power for the communication module is approximately:

where η is efficiency, expressed as a decimal. The 3.3 V rail also requires an input-power estimate based on its regulator type. Startup current, radio transmission peaks, actuator current, and capacitor charging may exceed the nominal budget and must be included in supply and protection decisions.

2.3 Tolerance and worst-case design

For a voltage divider, the ideal output is:

where R₁ and R₂ are resistances in ohms. The formula assumes negligible loading by the connected circuit. If the load resistance is significant, it appears in parallel with R₂ and must be included.

For example, with V_in = 12 V, R₁ = 30 kΩ, and R₂ = 10 kΩ,

If the analogue input must never exceed 3.3 V, the designer should also examine the maximum input voltage and resistor tolerances. A nominal calculation alone does not prove safety.

3. Schematic Design

A schematic is a functional and electrical description of a circuit. It should communicate intent clearly to a reviewer, assembler, troubleshooter, and future project member. It is not simply a graphical record of how components were placed on a breadboard.

A professional schematic should include:

• functional blocks arranged from input to output or from power entry to loads;
• reference designators and values;
• clearly named power rails and signal nets;
• connector pin numbers and interface descriptions;
• protection components near the relevant interface;
• decoupling capacitors near integrated-circuit supply pins;
• test points for important voltages, clocks, communication lines, and resets;
• notes concerning assembly, polarity, shielding, or special components;
• design rules and revision information.

Power distribution deserves special attention. A typical arrangement is input connector, fuse or resettable protection, reverse-polarity protection, transient suppression, filtering, regulation, and local decoupling. The exact order depends on the supply and fault model. Protection components must be selected for actual surge energy, clamping voltage, current, and response time rather than symbolically added to the schematic.

Analogue and digital sections should be separated where noise sensitivity requires it. Ground strategy must be explicit. A ground plane is often beneficial, but a plane does not automatically solve poor return-current paths. High-current switching returns should not share narrow traces with precision analogue signals.

4. PCB Layout and Design for Manufacturability

PCB design converts the schematic into physical geometry. The layout must satisfy electrical, thermal, mechanical, assembly, and inspection constraints simultaneously.

A practical sequence is:

1. confirm the component library footprints and pin numbering;
2. define board outline, mounting holes, keep-outs, and connector locations;
3. place safety-critical, mechanically constrained, and high-current components;
4. place sensitive analogue, clock, and communication components;
5. route power and high-current paths with suitable width and copper area;
6. route controlled or sensitive signals;
7. add planes, stitching vias, labels, and test points;
8. run electrical-rule and design-rule checks;
9. inspect the board using a schematic-to-layout comparison;
10. generate manufacturing and assembly outputs.

Trace width is not chosen solely by aesthetic preference. It depends on current, allowable temperature rise, copper thickness, external or internal layer position, and manufacturer guidance. A simplified resistive estimate for a copper trace is:

where R is resistance in ohms, ρ is copper resistivity in ohm-metres, L is trace length in metres, and A is cross-sectional area in square metres. This equation estimates DC resistance and does not by itself determine permissible PCB current. Temperature rise, thermal spreading, vias, pulse duration, and applicable PCB standards must also be considered.

DFM practices include consistent component orientation, adequate solder-mask and courtyard clearances, suitable pad geometry, accessible inspection areas, minimized unnecessary component varieties, and avoidance of parts that cannot be sourced or assembled reliably. Fiducials, tooling holes, polarity marks, and clear reference designators assist automated and manual assembly.

Design for testability means that faults can be localized efficiently. Test points should expose supply rails, ground, reset, programming interfaces, sensor outputs, actuator commands, and communication buses. A test point should be physically accessible and clearly labelled. A connector used only for programming should not be required for routine measurement if a safer alternative is practical.

5. Firmware and Control Implementation

Firmware should be designed as an engineered subsystem. Its architecture may contain hardware-abstraction drivers, device services, control logic, communication handling, fault management, and a user interface. Separating these concerns improves portability and testing.

A finite-state machine is often suitable for embedded projects. States such as INITIALISE, SELF_TEST, IDLE, RUN, FAULT, and SHUTDOWN should have defined entry conditions, permitted transitions, outputs, and fault responses. A state table is usually clearer than a long collection of nested conditional statements.

Timing must be explicit. A periodic task with sampling interval T_s has sampling frequency:

where f_s is in hertz and T_s is in seconds. The selected sampling rate must be appropriate for the signal bandwidth and control dynamics. Sampling too slowly can miss events or cause aliasing; sampling too quickly can waste processing time and increase noise sensitivity.

For a simple proportional controller:

where e(t) is the error between reference and measured output, K_p is proportional gain, and u(t) is the control command. This relationship is applicable to a proportional control law within actuator limits. Saturation, sensor delay, noise, and system dynamics may make proportional control insufficient. Integral and derivative terms should not be added without considering wind-up, noise amplification, and stability.

Firmware implementation should include input validation, watchdog strategy, timeout handling, range checks, safe startup outputs, and defined behaviour after communication loss. A watchdog is useful only when the software architecture can recover safely; repeatedly resetting a hazardous system is not an acceptable fault strategy.

Unit testing can verify drivers and calculations. Integration testing verifies interfaces between firmware and hardware. System testing verifies behaviour under realistic operating conditions. Boundary-value tests are particularly valuable for thresholds, overflow, sensor disconnection, and invalid communication frames.

6. Integration and Prototype Construction

Integration should proceed incrementally rather than by connecting every subsystem simultaneously. A recommended order is:

1. inspect boards and components for damage, polarity, and workmanship;
2. perform continuity and short-circuit checks with power removed;
3. energize the power-input stage using a current-limited supply;
4. verify each regulated rail under no-load and expected-load conditions;
5. program and test the controller in isolation;
6. connect sensors and communication interfaces one at a time;
7. connect actuators using dummy loads before real loads;
8. execute functional and fault tests;
9. record measured results, deviations, and corrective actions.

Current-limited laboratory supplies reduce the consequences of assembly errors. However, they do not replace fuses, isolation, protective earth, safe enclosures, or suitable personal protective equipment. Mains-connected prototypes require particular care, competent supervision, appropriate creepage and clearance, isolation barriers, and compliance with institutional safety procedures.

A prototype build record should contain board revision, firmware revision, component substitutions, calibration data, photographs, test equipment identification, and deviations from the approved design. Without this information, experimental results may not be reproducible.

7. Reliability, Maintainability, and Safety

Reliability is the ability of a system to perform its required function for a stated time under stated conditions. Reliability analysis begins with likely failure modes. A simplified failure-mode table may include component failure, loose connector, overheating, software lock-up, sensor drift, and incorrect assembly.

For components with a constant failure rate λ, the exponential reliability model is:

where R(t) is the probability of survival to time t, λ is a constant failure rate per unit time, and t uses the same time unit. This model assumes a constant hazard rate and is not universally valid, especially during early-life failures or wear-out. It should therefore be used cautiously and with appropriate component data.

Maintainability is improved by modular construction, accessible fasteners, keyed connectors, replaceable fuses, diagnostic indicators, documented calibration procedures, and software logs. A design that requires complete disassembly to replace a common sensor is technically functional but poorly maintainable.

Safety analysis should identify hazards, initiating events, consequences, protective measures, and residual risk. Relevant hazards may include electric shock, fire, stored energy, moving mechanisms, hot surfaces, radiation, battery failure, and unexpected software-controlled motion. Emergency shutdown behaviour must be defined, and safety functions should not depend on an unreliable communication link alone.

Ethical engineering practice requires honest reporting. A student must not present a prototype as compliant with a standard unless the relevant tests and evidence support that claim. Measured limitations, failed tests, substituted parts, and unresolved risks belong in the dissertation.

8. Verification, Validation, and Design Review

Verification asks whether the design was built correctly against its specifications. Validation asks whether the resulting system fulfils the intended user need. A voltage measurement can be verified against an accuracy requirement, while validation may require demonstrating that the measurement is useful for the intended control or monitoring task.

A test specification should identify the objective, equipment, setup, procedure, measured variables, uncertainty or tolerance, acceptance criterion, and result. Measurements should include units and, where appropriate, uncertainty. Repeating a measurement does not remove systematic error caused by calibration, loading, or an unsuitable instrument.

Design reviews should occur at schematic freeze, PCB release, first power-up, firmware integration, and final acceptance. Reviewers should examine assumptions, interfaces, worst-case conditions, safety, manufacturability, and traceability—not merely visual appearance.

9. Illustrative Case: Environmental Monitoring and Ventilation Controller

Background

Consider an illustrative undergraduate project that monitors room temperature and humidity and controls a 12 V ventilation fan. The controller uses a 12 V input, a 5 V regulator for a sensor, and a 3.3 V microcontroller. A threshold algorithm activates the fan when temperature exceeds a set value.

Problem

The initial concept operated on a breadboard but experienced microcontroller resets whenever the fan started. The sensor readings also fluctuated during switching.

Analysis

The fan generated a transient current and electrical noise. The breadboard contained long shared supply and ground paths. The design also lacked reverse-polarity protection, a local bulk capacitor, and a defined flyback path for the switching device.

Approach and solution

The revised design used a protected input, separate filtering for the logic rail, a low-side MOSFET driver with an appropriately rated flyback diode for the inductive fan, and local ceramic and bulk capacitors. The sensor was physically separated from the switching node. Firmware included startup self-test, a hysteresis band, a fan timeout, and a fault state for invalid sensor readings.

If the fan turns on at 0.8 A from 12 V, its nominal electrical power is:

The MOSFET conduction loss can be estimated by:

where R_DS(on) is the on-state drain-source resistance in ohms. If R_DS(on) = 0.10 Ω, the estimated conduction loss is:

This is a conduction estimate only; switching loss, gate-drive conditions, temperature dependence, and PCB thermal spreading must also be considered.

Outcome

The revised prototype completed repeated start-stop tests without resets. Sensor variation was reduced, and the test record documented supply voltage, fan current, temperature, firmware revision, and fault responses. The project did not claim regulatory certification; instead, it identified the prototype’s operating limits and recommended enclosure and EMC testing for future development.

Learning points

The case demonstrates that integration problems often arise at interfaces rather than within individual blocks. Power integrity, return-current paths, actuator transients, firmware fault handling, and physical placement must be designed together.

10. Contemporary Tools and Responsible Use of AI

Modern projects may use electronic-design-automation tools, simulation packages, automated PCB rule checking, version-control systems, and AI-assisted coding or documentation. These tools improve productivity but do not transfer engineering responsibility. Generated firmware must be reviewed for timing, memory safety, security, licensing, and hardware assumptions. Generated circuit suggestions must be checked against datasheets and worst-case conditions.

Version control should preserve schematic, PCB, firmware, test scripts, calibration data, and release notes. Each prototype should be identifiable by hardware and software revision. Reproducible builds and automated tests are increasingly valuable, particularly when several students collaborate.

Best-Practice Checklist

Before releasing a prototype, confirm that:

• every requirement has a verification method;
• component ratings include suitable electrical and thermal margins;
• footprints and pin assignments have been independently checked;
• power-up, shutdown, fault, and communication-loss behaviour are defined;
• PCB clearances, return paths, test points, and mounting features are adequate;
• firmware has been tested at normal and boundary conditions;
• dangerous outputs default to a safe state;
• manufacturing, programming, calibration, and repair instructions exist;
• test equipment and results are recorded with units and revisions;
• limitations and unresolved risks are reported honestly.

Chapter Summary

Detailed design converts project requirements into implementable and testable engineering decisions. Component selection must consider ratings, tolerances, thermal behaviour, availability, safety, and lifecycle rather than price alone. Power, voltage-divider, resistance, efficiency, sampling, control, and reliability relationships are useful only when their assumptions and units are understood.

A good schematic communicates function and interfaces; a good PCB layout also addresses current paths, noise, heat, assembly, inspection, and test access. Firmware should have explicit architecture, timing, fault handling, and verification. Integration is best performed incrementally with current-limited power, dummy loads, controlled tests, and complete build records. DFT, DFM, reliability, maintainability, and safety are design principles, not final-stage corrections. Finally, a defensible dissertation reports evidence, limitations, revisions, and risks as carefully as it reports successful operation.

Key Terms

Requirement: A statement of an expected function, performance, constraint, or condition.

Traceability matrix: A record linking requirements to design elements and verification evidence.

Derating: Operating a component below its maximum rating to provide design margin.

Tolerance: The permitted deviation of a component or parameter from its nominal value.

Worst-case analysis: Evaluation using combinations of limiting parameter values.

Schematic: A symbolic representation of electrical connectivity and function.

PCB: Printed circuit board supporting and interconnecting electronic components.

DFT: Design for testability; designing access and features that simplify testing and diagnosis.

DFM: Design for manufacturability; designing for consistent, economical production.

Reliability: Probability that a system performs its required function under stated conditions and time.

Maintainability: Ease and speed with which a system can be inspected, repaired, or restored.

Finite-state machine: A control model comprising defined states and permitted transitions.

Watchdog: A hardware or software timer used to detect failure of expected program activity.

Decoupling capacitor: A capacitor that supplies transient local current and reduces supply disturbance.

Flyback diode: A diode providing a current path for an inductive load when its switch opens.

Verification: Demonstration that a design meets specified requirements.

Validation: Demonstration that the system fulfils its intended purpose.

Derating factor: Ratio of design operating stress to rated component stress.

Fault mode: A manner in which a component or system can fail to perform its function.

Design review: A structured examination of design assumptions, evidence, risks, and readiness.

Review Questions

Short Answer Questions

1. Why should requirements include an acceptance criterion and verification method?
2. Distinguish verification from validation in an engineering prototype.
3. State two purposes of PCB test points.
4. Why can a nominal power calculation be insufficient for component selection?
5. What is the purpose of a watchdog, and why is it not by itself a complete safety strategy?

Long Answer Questions

1. Explain a systematic process for selecting components for an embedded electrical system, including electrical, thermal, manufacturing, lifecycle, and safety considerations.
2. Describe how a schematic should be organized to support design review, construction, troubleshooting, and future maintenance.
3. Discuss the principal PCB-layout considerations for power integrity, analogue signal quality, thermal management, and manufacturability.
4. Explain how firmware architecture, finite-state machines, timing, boundary testing, and fault handling contribute to dependable prototype operation.
5. Describe an incremental integration and verification plan for a controller containing a regulated power supply, sensors, a microcontroller, and an inductive actuator.

Analytical / Application Questions

1. A 12 V prototype contains a 0.6 A actuator, a 3.3 V controller drawing 0.15 A, and a 5 V sensor subsystem drawing 0.10 A. Propose a power-budget method that includes regulator efficiency, startup current, protection, and design margin. State the additional data required before selecting the input supply.
2. A prototype resets whenever an inductive load is switched off. Analyse at least four possible electrical, layout, and firmware causes, then propose an ordered diagnostic experiment that distinguishes among them.
3. Design a verification plan for a sensor-controlled safety actuator. Include normal operation, sensor disconnection, power interruption, communication loss, over-temperature, mechanical obstruction, and recovery testing. Define suitable evidence for each test.

Multiple Choice Questions with Answers

Question 1

Which document most directly links a system requirement to its design implementation and verification evidence?

A. Bill of materials

B. Requirements traceability matrix

C. Assembly drawing

D. Firmware binary

Correct Answer: B. Requirements traceability matrix

Question 2

A resistor has 10 V across it and resistance 100 Ω. What is its power dissipation?

A. 0.01 W

B. 0.1 W

C. 1 W

D. 1000 W

Correct Answer: C. 1 W

Question 3

Which practice most directly supports design for testability?

A. Removing all unused connectors

B. Adding accessible, labelled test points

C. Using the smallest possible components

D. Hiding calibration values in firmware

Correct Answer: B. Adding accessible, labelled test points

Question 4

For a voltage divider, the ideal output formula assumes that the connected load has what characteristic?

A. It has negligible effect on the divider

B. It always equals R₁

C. It is connected in series with R₁

D. It supplies power to the divider

Correct Answer: A. It has negligible effect on the divider

Question 5

What is the principal purpose of a flyback diode across a switched inductive load?

A. To increase the load resistance

B. To provide a current path for inductive energy when switching off

C. To amplify the control signal

D. To measure the load temperature

Correct Answer: B. To provide a current path for inductive energy when switching off

Question 6

Which activity is most appropriately performed before first powering a newly assembled PCB?

A. Connect the maximum-rated actuator

B. Disable all protective devices

C. Perform continuity and short-circuit checks

D. Increase the supply voltage above nominal

Correct Answer: C. Perform continuity and short-circuit checks

Question 7

A finite-state machine is especially useful because it defines what?

A. Only the PCB trace width

B. States, events, transitions, and associated behaviour

C. The market price of the prototype

D. The chemical composition of solder

Correct Answer: B. States, events, transitions, and associated behaviour

Question 8

If a component has a constant failure rate λ, which expression represents the exponential reliability model?

A. R(t) = λt

B. R(t) = 1 − λt

C. R(t) = e^(−λt)

D. R(t) = e^(λ/t)

Correct Answer: C. R(t) = e^(−λt)

Question 9

Which statement best distinguishes validation from verification?

A. Validation checks intended usefulness; verification checks conformance to specifications

B. Validation checks only solder joints; verification checks only software

C. Validation is performed before requirements exist

D. There is no practical distinction

Correct Answer: A. Validation checks intended usefulness; verification checks conformance to specifications

Question 10

A MOSFET carries 0.8 A and has R_DS(on) = 0.10 Ω. Neglecting switching loss, what is its conduction loss?

A. 0.008 W

B. 0.064 W

C. 0.80 W

D. 6.4 W

Correct Answer: B. 0.064 W