Electronics Guide

The Fretting Mechanism

Fretting occurs when two surfaces in contact experience small-amplitude oscillatory motion, typically in the range of micrometers to millimeters. In electrical connectors, vibration causes the male and female contacts to slide against each other. This micromotion has several consequences:

Oxide formation: Fresh metal surfaces exposed by the sliding motion rapidly oxidize in the presence of atmospheric oxygen. These oxide films are typically non-conductive or poorly conductive, increasing contact resistance.

Wear debris accumulation: The sliding motion generates metallic and oxide particles that accumulate at the contact interface. This debris can physically separate the contacts and further impede current flow.

Contact surface damage: Repeated motion causes mechanical wear of the contact plating, eventually exposing base metals that corrode more readily than the original plating material.

Contact force reduction: Spring contacts can lose normal force due to material fatigue or settling, reducing the pressure that maintains intimate metal-to-metal contact.

The result is progressive increase in contact resistance, often accompanied by intermittent open circuits when vibration momentarily separates the contacts. These effects generate both conducted and radiated interference as the fluctuating contact resistance modulates circuit currents.

Electrical Effects of Fretting

Degraded connectors affect circuit behavior in multiple ways:

Resistance modulation: As contact resistance varies with vibration, the voltage drop across the connection fluctuates. In power circuits, this appears as supply voltage variation. In signal circuits, it represents noise injection.

Intermittent opens: Complete momentary loss of contact creates open-circuit transients. For digital signals, this may cause data errors. For power connections, it creates voltage spikes due to inductive energy release.

Arcing and sparking: When contacts separate while carrying current, arcing can occur. Arc events generate broadband electromagnetic interference and can cause permanent contact damage.

Noise injection: The fluctuating contact resistance introduces noise at vibration frequencies and their harmonics. This noise often has components in the audio frequency range (20 Hz to 20 kHz) corresponding to typical vibration spectra.

These effects are particularly troublesome for low-level analog signals where small noise voltages represent significant errors, and for high-speed digital circuits where impedance variations affect signal integrity.

Connector Design for Vibration Resistance

Connectors intended for vibrating environments incorporate features to resist fretting:

High normal force: Contacts designed with higher normal force maintain better contact even when disturbed by vibration. However, higher forces require more insertion force and can increase wear.

Precious metal plating: Gold and other noble metal platings resist oxide formation, so even if fretting motion occurs, the exposed surfaces remain conductive. Gold thickness specifications often distinguish between static (thin plating acceptable) and dynamic (thick plating required) applications.

Multiple contact points: Contacts with multiple independent contact points provide redundancy; if some points develop high resistance, others may maintain good contact.

Lubricants: Contact lubricants reduce friction and wear, and can help seal the contact area from oxidizing atmospheres. The lubricant must be carefully selected to remain stable over the expected temperature range and service life.

Strain relief: Proper strain relief prevents cable-transmitted vibration from reaching the contact interface. The strain relief should accommodate some cable motion while preventing that motion from affecting the contacts.

Connector locking: Positive locking mechanisms prevent connector separation under vibration and can also reduce relative motion between contacts.

Triboelectric Noise in Cables

Triboelectric effect is the generation of electrical charge when two dissimilar materials are separated. In coaxial cables, vibration causes relative motion between the center conductor, dielectric insulation, and shield. This motion generates charge that appears as noise voltage on the cable conductors.

The triboelectric noise mechanism operates as follows:

When the insulator and conductor move relative to each other, charge transfer occurs at the interface. As they separate, the separated charge creates a voltage difference. This voltage drives current through the cable's capacitance, appearing as noise at the cable output.

Triboelectric noise is particularly problematic in:

• High-impedance sensor circuits (piezoelectric accelerometers, pH probes, etc.)
• Low-level signal cables in vibrating environments
• Cables with loose construction that allows internal motion
• Applications requiring low-frequency response, as triboelectric noise often has significant low-frequency content

Low-noise cables designed for vibrating environments use special constructions to minimize triboelectric effect:

• Tight construction: Minimizing relative motion between cable elements reduces charge generation.
• Conductive layers: Semiconductive or graphite-impregnated layers between insulation and shield allow charge to bleed off before it can generate significant voltage.
• Special dielectric materials: Some insulating materials generate less triboelectric charge than others.
• Shield material selection: Different shield constructions (solid, braid, spiral) have different susceptibility to triboelectric noise.

Cable Fatigue and Failure

Repeated flexing eventually causes conductor fatigue and failure. As conductors break, the cable's electrical properties change, and intermittent contact can generate significant interference:

Gradual resistance increase: As conductor strands break, the remaining strands carry more current, heating increases, and resistance rises further. This progressive degradation may be detected by monitoring circuit parameters.

Intermittent opens: Before complete failure, broken conductors may make intermittent contact during flexing, generating noise similar to connector fretting.

Shield degradation: Shield fatigue reduces shielding effectiveness and increases both emissions and susceptibility. Braided shields typically have better flex life than spiral shields or foil shields.

Insulation damage: Repeated flexing can crack insulation, creating paths for leakage current and potentially leading to shorts.

Cable flex life is specified for cables intended for continuous flexing applications. Proper cable selection, routing, and strain relief are essential for reliable operation in vibrating environments.

Cable Routing and Support

How cables are routed and supported significantly affects their response to vibration:

Service loops: Providing extra cable length in the form of service loops prevents strain when relative motion occurs between connection points. The loops should be sized to accommodate the expected motion without excessive slack.

Support spacing: Cable supports should be spaced to prevent cable resonance at expected vibration frequencies. Unsupported cable spans can develop standing wave oscillations that accelerate fatigue.

Clamp design: Cable clamps should grip firmly enough to prevent abrasion but not so tightly as to damage conductors. Rubber-lined clamps provide cushioning and grip without excessive pressure.

Routing path: Cables should be routed to minimize bend angles and avoid sharp edges. Corners should be rounded, and cables should not be routed under tension.

Harness construction: Bundling cables into harnesses can provide mutual support and damping. However, harness ties should not be overtightened, and cables requiring flexibility should not be bundled with fixed cables.

Component Resonance

Every component and assembly has natural frequencies at which vibration response is amplified. When excitation occurs at or near a resonant frequency, displacement can be many times greater than the input displacement:

The amplification factor (Q) at resonance depends on damping. Lightly damped components can have Q values of 10-100 or more, meaning that 0.1 mm of input vibration produces 1-10 mm of component motion at resonance.

Component resonance creates several problems:

• Mechanical stress: Amplified motion creates stress in component leads, solder joints, and mounting points that can lead to fatigue failure.
• Microphonics: Resonating components generate larger electrical noise signals due to greater mechanical displacement.
• Parameter modulation: Component values may be significantly affected by large-amplitude vibration at resonance.
• Contact to adjacent structures: Large displacements can cause components to contact adjacent components or enclosure walls, potentially causing shorts or mechanical damage.

Avoiding component resonance at operating vibration frequencies is a key design goal. This may involve selecting components with different resonant frequencies, adding damping, or stiffening the mounting to raise the resonant frequency above the excitation range.

Surface-Mount Component Behavior

Surface-mount components present particular challenges in vibrating environments:

Large components: Components such as large capacitors, transformers, and connectors have significant mass that creates high inertial forces during vibration. The solder joints must withstand these forces without cracking.

Tall components: Components with high center of gravity experience torque during vibration that stresses solder joints, particularly at the corners of the footprint.

Flex during PCB resonance: When a PCB resonates, it bends, creating strain in surface-mount solder joints. Large components resist bending and concentrate stress at their edges.

Ceramic component cracking: Ceramic components (particularly MLCCs) can crack if subjected to excessive PCB bending. Once cracked, they may exhibit intermittent behavior or parametric changes.

Mitigation approaches for surface-mount components include:

• Using underfill or staking compound to reinforce large components
• Locating heavy components near PCB supports where motion is minimized
• Using larger solder joint footprints that provide greater mechanical strength
• Avoiding placement of stress-sensitive components in high-strain areas of the PCB
• Using flexible termination capacitors in high-flex areas

Through-Hole Component Considerations

Through-hole components have leads that pass through the PCB, providing different mechanical characteristics:

Lead stress: Component leads between the body and PCB surface act as springs, absorbing some vibration energy. However, repeated flexing can cause lead fatigue, particularly at the board surface and at the component body.

Solder joint integrity: Through-hole solder joints have more solder volume and generally higher mechanical strength than SMT joints, but can still fail under repeated stress.

Component body support: Large through-hole components may require supplemental mounting (adhesives, clips, or tie-downs) to prevent the component body from moving relative to the leads.

Lead-free solder effects: Lead-free solders are generally more brittle than tin-lead solders and may be more susceptible to fatigue cracking under vibration.

PCB Resonance Characteristics

Printed circuit boards are thin, plate-like structures that exhibit multiple resonant modes. The fundamental frequency and mode shapes depend on board dimensions, material properties, mounting configuration, and mass distribution:

Fundamental mode: The first resonant mode typically has the lowest frequency and the largest displacement. For a rectangular board simply supported at its edges, the fundamental mode is a single half-wave with maximum displacement at the center.

Higher-order modes: Successive modes have higher frequencies and more complex deflection patterns with multiple peaks and nodes. Components located at nodal lines experience less displacement than those at antinodes.

Mass effects: Heavy components shift resonant frequencies and can create local displacement maxima. The added mass can lower the frequency of modes where the component is at a displacement peak.

Mounting effects: How the board is supported significantly affects resonance. Edge-clamped boards have different modes than corner-supported boards or boards mounted with standoffs.

PCB resonances typically range from tens of hertz for large, flexible boards to hundreds of hertz for small, stiff boards. Boards should be designed so that these resonances do not coincide with expected vibration excitation frequencies.

Enclosure and Chassis Resonance

The enclosure or chassis housing the electronics also has resonant frequencies:

Panel resonances: Large flat panels resonate at frequencies determined by their size, thickness, and edge conditions. Resonating panels can transmit vibration to internal components and radiate acoustic noise.

Structural resonances: The overall structure has modes where different parts move in characteristic patterns. Some modes may amplify motion at the PCB mounting points.

Local resonances: Mounting brackets, cable supports, and other internal structures have their own resonant frequencies that can locally amplify vibration.

Enclosure resonances can be managed through stiffening (adding ribs or gussets), damping (using constrained-layer damping materials), and design changes to shift resonances away from excitation frequencies.

Resonance Identification and Avoidance

Identifying resonances before they cause problems requires analysis and testing:

Finite element analysis: Structural FEA can predict resonant frequencies and mode shapes during design. This allows design changes before hardware is built.

Modal testing: Experimental modal analysis using accelerometers and impact hammers or shakers identifies actual resonances in hardware. Results can be compared with analysis to validate models.

Transmissibility testing: Measuring the ratio of response to input as a function of frequency reveals amplification at resonances. High transmissibility indicates problematic resonances.

Once identified, resonances can be addressed by:

• Shifting resonant frequency by changing stiffness or mass
• Adding damping to reduce amplification
• Isolating the assembly from vibration input at problematic frequencies
• Relocating sensitive components away from displacement maxima

Solder Joint Fatigue

Solder joints are often the weakest link in vibrating electronic assemblies. The combination of relatively low fatigue strength and concentration of stress at the joint makes solder joint fatigue a common failure mode:

Thermal and mechanical fatigue interaction: Solder joints experience both thermal cycling (due to power dissipation and ambient temperature changes) and mechanical cycling (due to vibration). These effects combine, often synergistically, to accelerate fatigue.

Crack initiation and propagation: Fatigue cracks typically initiate at stress concentrations such as corners, voids, or surface defects. Once initiated, cracks propagate through the joint with each stress cycle until complete fracture occurs.

Creep effects: Solder is subject to creep (time-dependent deformation under stress) at typical operating temperatures. Creep interacts with fatigue, particularly in thermal cycling, to accelerate failure.

Joint geometry effects: Solder joint shape affects fatigue life. Fillets that distribute stress and avoid sharp corners improve fatigue resistance. Insufficient solder or excessive solder creating large stand-off heights both degrade fatigue performance.

Improving solder joint fatigue life involves proper solder paste and reflow profiles, good joint geometry, appropriate pad and lead design, and avoidance of contamination and defects that create stress concentrations.

Component Lead Fatigue

Component leads undergo repeated bending during vibration, leading to fatigue:

Lead material properties: Different lead materials (copper, alloy 42, copper-clad steel, etc.) have different fatigue characteristics. More ductile materials generally have better fatigue life but may have other drawbacks.

Lead forming: The forming process creates residual stresses and work hardening that affect fatigue life. Sharply formed leads are more susceptible to fatigue than gently curved leads.

Stress concentration: Fatigue cracks typically initiate at stress concentrations such as the point where leads exit the component body or enter the solder joint.

Lead length: Longer leads are more flexible and experience lower stress for a given displacement, but may also have lower resonant frequencies that can lead to amplified motion.

Structural Fatigue

Beyond components and joints, structural elements of the electronic assembly can experience fatigue:

PCB fatigue: Printed circuit boards can develop fatigue damage when flexed repeatedly. FR-4 and other PCB materials have finite fatigue life, and cracking can occur in severely stressed boards.

Mounting fastener fatigue: Screws, rivets, and other fasteners experience cyclic loads during vibration and can fail through fatigue. Properly designed joints distribute load and avoid stress concentration.

Bracket and support fatigue: Mounting brackets and component supports that experience cyclic stress can crack, leading to increased component motion or complete detachment.

Characteristics of Intermittent Connections

Intermittent connections exhibit time-varying resistance or complete open/short conditions:

Position-dependent behavior: The connection may work in some orientations or positions but fail in others, making desk testing unable to replicate field failures.

Vibration-dependent behavior: The connection may work when static but fail under vibration, or may require specific vibration frequencies or amplitudes to induce failure.

Temperature-dependent behavior: Thermal expansion may change the relationship between mating parts, causing failures that occur only at certain temperatures.

Time-varying behavior: The connection may work initially but fail as fretting or fatigue progresses, or may improve temporarily if mechanical action cleans the contact surfaces.

EMC Effects of Intermittent Connections

Intermittent connections create several forms of electromagnetic interference:

Broadband noise: Fluctuating contact resistance generates noise across a wide frequency spectrum, with content depending on the mechanical vibration spectrum and the contact dynamics.

Transient generation: Sudden opens and shorts create fast transients that can have spectral content extending to high frequencies. These transients can couple into adjacent circuits and radiate from cables.

Signal integrity effects: In digital circuits, intermittent connections can cause data errors, timing violations, and protocol failures. In analog circuits, they can produce spikes, dropouts, and noise.

Power integrity effects: Intermittent power connections create supply voltage variations that affect all circuits on the affected rail, potentially causing widespread malfunction.

Detecting Intermittent Connections

Finding intermittent connections requires techniques that exercise the connection while monitoring for failures:

Vibration testing: Applying controlled vibration while monitoring circuit operation can reveal connections that fail under mechanical stress.

Thermal cycling: Temperature cycling while monitoring can reveal connections affected by differential thermal expansion.

Mechanical stress application: Flexing cables, pressing on assemblies, or tapping enclosures while monitoring can locate sensitive areas.

Event capture: Recording circuit operation over extended periods can capture intermittent events that occur randomly.

Time domain reflectometry: TDR can detect impedance variations associated with intermittent connections, and some instruments can capture intermittent events.

The Triboelectric Series

When two materials contact and separate, charge transfer occurs, with one material becoming positively charged and the other negatively charged. The triboelectric series ranks materials by their tendency to become positive or negative:

Materials near the positive end of the series (such as glass, nylon, and human skin) tend to donate electrons and become positive. Materials near the negative end (such as PTFE, silicone rubber, and PVC) tend to accept electrons and become negative. The greater the separation in the series between two materials, the greater the charge transfer.

This has several implications for electronic design:

• Avoid combinations of materials far apart in the series where relative motion can occur
• Use materials in the middle of the series for minimum charge generation
• Add conductive materials or coatings to drain charge before it can accumulate

Triboelectric Noise in Sensors

High-impedance sensors are particularly susceptible to triboelectric effects:

Piezoelectric sensors: The cables connecting piezoelectric accelerometers, microphones, and other sensors are notorious for triboelectric noise. Special low-noise cables are essential for accurate measurements in vibrating environments.

Electrostatic sensors: Capacitive and electrostatic sensors have high-impedance inputs that can be affected by triboelectric charge generated on nearby surfaces.

High-impedance voltage measurements: Any measurement with high source impedance (pH probes, ion-selective electrodes, photomultipliers, etc.) can be affected by triboelectric effects in cables and connectors.

Reducing triboelectric noise in sensor systems involves using specialized low-noise cables, minimizing cable motion, providing proper cable support, and using charge amplifiers or impedance buffers close to the sensor to reduce the impedance of long cable runs.

Triboelectric Effects in Components

Triboelectric charging can also affect components within electronic assemblies:

Relay and switch contacts: Rubbing action as contacts open and close can generate charge that affects circuit operation.

Component motion: Components that shift relative to their mounting or to adjacent structures can generate triboelectric charge.

Cable bundle rubbing: Cables rubbing against each other within a harness can generate charge that couples into the cables.

These effects are generally smaller than cable triboelectric noise but can be significant in very sensitive circuits.

Shock Versus Vibration

Shock differs from vibration in several important ways:

Duration: Shock events are brief, typically measured in milliseconds, while vibration may be continuous or last for extended periods.

Amplitude: Shock accelerations often exceed steady-state vibration levels by large factors. Peak accelerations of hundreds or thousands of g's are common in shock events.

Frequency content: Shock transients contain energy across a broad frequency spectrum determined by the pulse shape and duration. The shock response spectrum (SRS) characterizes this content.

Response: The system response to shock depends on the relationship between shock duration and system natural frequencies. Short shocks excite high-frequency modes, while long shocks excite low-frequency modes.

EMC Effects of Shock

Shock events can cause immediate EMC effects and longer-term damage:

Contact opening: High acceleration forces can cause connectors to momentarily separate and relay contacts to chatter, generating transient interference.

Component displacement: Shock can cause large component displacements that affect circuit parameters, cause contact with adjacent structures, or result in permanent position shifts.

Structural damage: Severe shocks can crack solder joints, break component leads, and damage other structural elements. This damage may manifest immediately or may weaken the structure so that failure occurs later under lesser stress.

Relay chatter: Electromechanical relays are particularly susceptible to shock-induced contact chatter. This is an important consideration for circuits that control critical functions.

Shock Hardening Approaches

Making electronic systems resistant to shock involves:

Shock isolation: Mounting electronic assemblies on shock mounts that attenuate high-frequency shock content can dramatically reduce transmitted acceleration. The mount natural frequency must be lower than the shock spectrum corner frequency for effective isolation.

Structural stiffening: For assemblies that cannot be isolated, increasing stiffness reduces relative motion and associated stress. This approach shifts resonances to higher frequencies where the shock spectrum may have less energy.

Component selection: Choosing components rated for shock environments ensures they can survive expected conditions. Solid-state devices instead of electromechanical devices eliminate mechanical failure modes.

Encapsulation: Potting or conformal coating can support components and reduce relative motion during shock. The encapsulant must remain compliant enough to avoid creating additional stress from differential thermal expansion.

Types of Damping

Several damping mechanisms are available for electronic system design:

Material damping: All materials have some internal damping due to friction between molecular structures or crystal grains. Viscoelastic materials (rubber, certain polymers) have particularly high damping.

Structural damping: Friction at joints and interfaces dissipates energy. This damping is often variable depending on joint preload and condition.

Viscous damping: Motion through a fluid generates velocity-dependent forces that dissipate energy. Fluid dampers use this principle.

Constrained-layer damping: A viscoelastic material sandwiched between two stiff layers is forced to shear when the structure bends, effectively dissipating energy. This is particularly effective for panel resonances.

Tuned mass dampers: A mass-spring-damper system tuned to a problematic resonant frequency can absorb energy at that frequency, reducing the main structure's response.

Damping Materials for Electronics

Several damping materials are compatible with electronic assemblies:

Silicone rubber: Excellent damping over a wide temperature range, compatible with electronics, and available in sheet, pad, and potting forms.

Urethane materials: Good damping properties, can be formulated for specific applications, available as foam, elastomer, and coatings.

Acrylic tape products: Damping tapes with viscoelastic layers can be applied to vibrating panels to reduce resonant amplification.

Potting compounds: Some potting materials provide significant damping in addition to environmental protection and mechanical support.

Conformal coatings: While primarily for environmental protection, conformal coatings add some damping to component motion.

Vibration Isolation

When damping at the source is insufficient, isolation mounts can reduce vibration transmitted to sensitive assemblies:

Isolator selection: Isolators must be chosen based on the weight to be supported and the frequency range to be attenuated. The isolator natural frequency should be well below the lowest problematic excitation frequency.

Multi-axis isolation: Vibration can occur in any direction, so isolators must provide adequate compliance in all axes. Some designs use separate isolators for different axes; others provide multi-axis isolation with a single mount.

Damping in isolators: Isolators need some damping to limit amplification at resonance. However, excessive damping reduces isolation effectiveness at higher frequencies.

Static deflection: Soft isolators that provide good isolation may have large static deflection that must be accommodated in the mechanical design. Nonlinear stiffness can help manage this tradeoff.

Shock and vibration: Isolators designed for vibration may not be effective for shock, and vice versa. Combined environments may require specialized or staged isolation systems.