Protection Device Selection

The Critical Art of Circuit Protection

Protection parts on a schematic do not automatically translate to a robust product in the field. Real transients (ESD, EFT, surge, and cable discharge events) depend on the installation, grounding, and where the energy actually flows.

Protection devices alone do not make a protected circuit. Coordination between stages, low-impedance return paths, thermal considerations, and realistic expectations of what each device can handle determine whether the system survives or fails.

This guide compares common protection devices (TVS, MOV, PTC, GDT, and more) and focuses on selection and coordination details that matter in real products.

The Workhorses: TVS Diodes

TVS diodes are the default choice for most circuit protection: fast (nanosecond response), precise clamping, and reliable. They use avalanche breakdown: when voltage exceeds the breakdown threshold, avalanche multiplication floods the junction with carriers and effectively short-circuits the transient energy to ground. The clamping voltage (the voltage across the device while it's conducting the specified peak pulse current) is what determines the maximum stress your protected circuit actually sees.

Peak pulse power versus pulse duration follows the Wunsch-Bell model in the intermediate (thermal-diffusion) regime: $P_{peak} = K \times t^{-0.5}$, where K depends on device construction and thermal mass. The square-root dependence makes sense physically because shorter pulses give heat less time to spread, so you need more peak power to cause damage. For very short pulses (under roughly 100 ns) the response is closer to adiabatic, with $P \propto t^{-1}$, and for very long pulses it tends to a steady-state thermal limit that is independent of duration.

MOVs: High Energy, But They Wear Out

MOVs work on completely different physics. Sintered zinc oxide grains form thousands of back-to-back diode junctions in series, creating a highly nonlinear resistance. The V-I relationship follows a power law: $I = K \times V^{\alpha}$, where alpha typically ranges from 25 to 50, a very sharp transition from insulating to conducting. Large MOVs can absorb thousands of joules, which is their main advantage over TVS diodes.

The trade-off: each surge event damages grain boundaries, gradually degrading the device. You'll see increasing leakage current and worse clamping over time, which can eventually lead to thermal runaway if you don't account for it. How fast they degrade depends on surge amplitude, duration, and how often events occur. Predicting lifetime is difficult, which is why MOVs in critical applications need monitoring or periodic replacement.

PTCs: Resettable Overcurrent Protection

PTCs (resettable fuses) are conductive particles in a crystalline polymer matrix. At normal temperatures the particles form continuous conductive paths, which gives low resistance. When current heats the device past the polymer's phase-transition temperature, the matrix expands and breaks those paths. Resistance jumps by up to six orders of magnitude, effectively cutting off current flow.

Trip time follows from the thermal balance: $t_{trip} = (m \times c \times \Delta T) / (I^2 \times R - P_{dissipated})$, where the dissipated power depends on ambient temperature and thermal resistance to the surroundings. Unlike traditional fuses, PTCs reset when power is removed and temperature drops. But they're not infinite-life devices. Repeated trips cause resistance drift and eventual failure.

Gas Discharge Tubes: The Heavy Artillery

GDTs handle very large surges, with the rating depending strongly on package and waveform. Small SMD signal-line GDTs are typically rated 0.5 to 5 kA on the standard 8/20 µs surge waveform; mid-size telecom SMD parts reach about 25 kA, and the largest AC-power and lightning-arrester GDTs are rated up to around 100 kA. When voltage exceeds the breakdown threshold, the gas ionises and creates a plasma arc that can conduct very large current with low voltage drop. They're common as primary protection on telecom lines and lightning-exposed equipment.

The catch: GDTs are relatively slow (often microseconds to fire) and can have "follow current" issues in AC systems where the arc persists after the surge passes. Their breakdown voltage also varies with temperature and aging, so you need margin and should use the datasheet curves. They need coordination with faster secondary protection to be effective.

Coordination: Making Multi-Stage Protection Work

Primary protection (GDTs, MOVs) absorbs the bulk energy but responds slowly. Secondary protection (TVS diodes) clamps faster but can't handle as much energy. The challenge: the primary must fire before the secondary burns out.

Impedance between stages is the key. Series resistance or inductance limits current and creates voltage division so the primary sees enough voltage to trigger. Too little impedance and your TVS dies before the GDT fires. Too much and you impact signal integrity.

Let-Through Energy: What Gets Past Your Protection

Clamping voltage gets all the attention, but let-through energy is what actually damages circuits. During the very fast leading edge of an ESD or surge event, some voltage and energy can get past the clamp due to parasitic inductance in the path (and the finite response of real devices). That energy can concentrate in microscopic areas, where it can fuse metallization or damage junctions even if the total energy seems small.

With slower protection devices, let-through energy can be substantial. The protected circuit has to survive whatever gets through. Protection design is about matching device response to threat characteristics and circuit vulnerability. Device specs feed that match, but on their own they don't tell you whether the part is the right one for your front end.

High-Speed Signal Protection

At multi-gigabit data rates, every protection device you add degrades signal quality through parasitic capacitance. For a digital signal with rise time $t_r$, bandwidth is roughly $BW \approx 0.35 / t_r$. That capacitance forms a low-pass filter with source and load impedances, softening edges and attenuating high-frequency content. Modern low-capacitance TVS devices get down to sub-picofarad levels using thyristor-based designs that separate the protection junction from the signal path, but they typically give up some protection robustness to get there. That's the fundamental trade-off: protection level versus signal integrity. Differential signaling helps because common-mode rejection handles part of the problem, but your protection devices need matched characteristics or you'll convert common-mode noise into differential-mode noise.

Each protection device type fails differently, and knowing how matters for your design. TVS diodes usually fail short-circuit when they absorb too much energy, which is a safe failure mode that kills the circuit but keeps protection in place. MOVs are trickier: they degrade gradually, leakage climbs, and eventually they can go into thermal runaway. This is a fire risk if you don't have a series fuse. PTCs fail from polymer degradation after too many thermal cycles, rising contact resistance from oxidation, or mechanical damage from thermal stress. GDTs can lose their gas seal, changing breakdown characteristics, or fail to clear follow current in AC applications. These failure modes drive practical choices: fuse your MOVs, monitor PTC health, and add redundancy where failure would be dangerous.

Standards drive a lot of protection decisions. The IEC 61000-4 series covers the main threats: ESD (61000-4-2), electrical fast transients (61000-4-4), and surge (61000-4-5). You don't just have to survive the test; you have to maintain functionality according to defined performance criteria. Medical devices add IEC 60601 requirements including patient leakage current limits that constrain what protection topologies you can use. Automotive (ISO 7637-2, ISO 16750-2) throws unique transients at you, from fast spikes to load dump events. On a 12 V system, the ISO 16750-2 load dump pulse lasts 40 to 400 ms; un-suppressed (Test A) it reaches 65 to 87 V above the battery voltage, and a centrally suppressed system (Test B) clamps it to roughly 35 V. The really brutal numbers belong to 24 V trucks, where the un-suppressed pulse runs 123 to 174 V above battery. Telecom equipment has Telcordia GR-1089 for central office gear. Each domain has its own threat profile, and the best protection often requires creative combinations of devices to meet everything at once.

Don't ignore the thermal side. During normal operation, leakage current through protection devices generates steady heat that your thermal design needs to handle. MOVs are particularly problematic here because their leakage increases with temperature, creating a positive feedback loop that can run away. During surges, protection devices dissipate enormous peak power and rely on thermal mass to absorb it before the junction overheats. Temperature rise follows $\Delta T = P_{average} \times Z_{th}(t)$, where thermal impedance is time-dependent because of thermal capacitance. Watch out: the manufacturer's pulse rating curves assume specific mounting and ambient temperature that may not match your board. Give protection devices adequate copper area for heat spreading, use thermal vias to conduct heat into inner layers, and check what else nearby is adding heat.

Protection components are cheap compared to the rest of your BOM, but their impact on field reliability can dominate your total product cost through warranty claims and reputation damage. Over-protect and you waste board space and money while potentially hurting signal quality. Under-protect and you risk field failures. The right level depends on how likely the threat is, how bad a failure would be, and how long the product needs to last. High-volume consumer products sometimes accept minimal protection betting on low exposure, while industrial or medical equipment needs comprehensive protection regardless of cost. You can also design protection as swappable modules that vary by market or installation environment. Multi-function protection ICs are getting better too. Combining several protection functions in one package saves board space and reduces parasitics.

Active protection circuits using MOSFETs or thyristors can get close to ideal (zero leakage, minimal capacitance, precise trigger thresholds) but at much higher complexity and cost. Hybrid devices like GDT-TVS combinations give you high energy handling with fast response. Silicon avalanche suppressors are blurring the line between discrete protection and ICs, adding temperature compensation, current limiting, and diagnostic features on the same die.

I've developed protection strategies for everything from delicate sensor interfaces to high-power industrial controls. I've debugged failures ranging from subtle let-through energy damage to spectacular MOV fires. This experience builds understanding of what works in practice versus what looks good on datasheets.

If you're designing protection for a challenging environment or troubleshooting field failures, let's talk. Sometimes understanding why protection failed is the key to designing protection that works.

Robust circuit protection is designed into every project. Explore my electronic design services to learn how we protect hardware against real-world electrical threats.