An outdoor node is not protected when a TVS diode appears somewhere near the connector. It is protected when the unwanted energy has a shorter, lower-impedance and more survivable route than the one through the controller, radio, sensor front end or power converter. The protection design is a routing problem as much as a component choice.

The board edge is where the product meets the world. Power wires bring load dumps, long cable inductance, wrong adapters and nearby lightning transients. Data wires bring ESD from hands, tools and dry air, plus common-mode noise from long runs. A metal enclosure or shield can help, but only if the PCB gives the discharge a path to use. If the path is vague, the surge will invent one.

Outdoor protection starts by deciding where the strike goes

Surge and ESD protection on hardware that lives outdoors begins before selecting a package. The first decision is the boundary. Which conductors enter from outside? Which of them can be touched? Which ones may run tens of meters? Which cable shield is tied to chassis, logic ground or left floating? Which metalwork is allowed to carry a discharge? Without those answers, a protection part can be present and still route energy through the wrong plane.

ESD and surge are not the same event. ESD is fast, local and often arrives through a connector shell, pin or exposed surface. Surge is slower by comparison, carries more energy and often comes through cable inductance, power wiring or a nearby strike path. The front end may need both a fast clamp near a data line and a higher-energy device on a supply or field line. Putting one device in the schematic does not mean it handles every threat. The path, rating and coordination decide which device sees the first hit and which one finishes the job.

The layout can defeat a good part. A TVS connected by a long thin trace has inductance, and a fast event will raise the protected node before the clamp can be useful. A discharge current that returns through the digital ground under the MCU may keep the pin voltage in range while the whole logic domain jumps. A protector placed after a connector, fuse, choke and narrow neck may let the first pulse stress parts that were meant to sit behind it. Good protection sits where the outside line enters, has a broad return to the chosen dump node, and keeps the protected signal from sharing the dirty current path.

That is the discipline behind outdoor hardware. You do not only add clamps. You draw the fault current.

A powered interface needs a clamp that matches its normal voltage

SMBJ100CA protecting a powered IoT interface is the kind of choice made when a powered line may see transients above its normal operating range. The suffix and rating are not decorations. The designer has to compare the working voltage, the possible adapter or bus tolerance, the clamping behavior under pulse current and the parts downstream. If the clamp voltage is too low, it conducts during valid operation or margin events. If it is too high, the protected converter or IC may see more than it can survive.

Bidirectional and unidirectional behavior also matters. A DC power input often has a defined polarity, but field wiring may see reverse connection, inductive kick or cable discharge that does not follow a neat polarity story. A bidirectional TVS can be a practical choice in some field lines because it handles excursions in both directions. A unidirectional part may clamp one direction more tightly and suit a rail referenced to ground. The choice should follow the fault model, not a habit copied from another schematic.

The TVS has to coordinate with the upstream and downstream parts. A fuse or PTC may limit sustained fault current, but it will not react to the first fast edge. A series impedance may help share stress, yet it can also create voltage drop or heat. The power converter input capacitance may absorb small events and worsen inrush. The enclosure ground may be a better dump point for surge current than logic ground, provided the product has a real chassis reference. Protection is not a single part. It is a chain that must fail in the intended order.

High-energy TVS and fuses are not interchangeable

1.5KE100A on the power input of a field device points to the higher-energy side of transient suppression. A larger TVS package can absorb a larger pulse than a small data-line part, but it still has limits tied to waveform, repetition, ambient temperature and mounting. A part that survives one surge in a lab may not be meant to absorb repeated abuse caused by a bad installation. The design should ask whether the event is rare, repeated, or a sign of a wiring fault that should disconnect the product.

0697H9200-02 as a fast acting fuse for overcurrent handles a different job. A fuse protects against sustained overcurrent and fire risk. It does not clamp voltage. It does not save a microcontroller pin from a nanosecond ESD hit. It opens when current and time cross its characteristic curve, which means the protected system must tolerate what happens before the fuse clears. In a field device, the fuse and TVS often work together: the TVS diverts a transient, while the fuse opens if the condition becomes a longer fault.

The order around the connector matters. If the fuse sits before the TVS, a severe event may force current through the fuse path first and create a high voltage on the downstream side. If the TVS sits before the fuse, it may see more surge energy and needs a return path that can carry it. There is no universal placement rule because cable, supply, enclosure and safety requirements change the answer. The useful rule is to trace both fast transient current and sustained fault current, then choose which component should carry each one.

Vehicle and CAN interfaces need low-capacitance protection in the right place

AVR-M1608C080MTAAB for ESD on a vehicle interface belongs to a class of small protection parts used where the signal must remain clean while the outside contact remains touchable or cable-facing. A vehicle interface may see dry-air ESD, harness coupling, load switching and ground offsets. The protector must clamp fast enough, add little enough capacitance for the signal, and be placed close enough to the entry point that the discharge does not cross sensitive routing before it reaches the device.

NUP2105L for ESD protection on a CAN bus shows the same logic on a differential bus. CANH and CANL need protection that does not distort normal differential signaling or add an imbalance that hurts EMC. The protector should sit near the connector, usually before the transceiver pins see the cable. Its return path must be chosen with care: dumping a harsh event into a quiet logic ground right beside the transceiver can cause a reset even if the bus pins survive. The protected bus is only protected if the discharge route is shorter than the route through the transceiver.

The term low capacitance can be misleading if treated as a magic label. The needed capacitance depends on bus speed, line impedance, stub length and the existing loading from choke, connector and routing. A part with low capacitance but poor placement can perform worse than a slightly larger part mounted correctly. The board should keep the protector symmetrical on a differential pair, avoid stubs that become antennas, and route the dirty return away from the receiver input reference.

Many data lines need arrays, but arrays create shared behavior

SMF05C as an ESD array across several data lines is attractive when several GPIO, USB, keypad, sensor or service lines leave the quiet part of the board. A multi-line array saves area and can make placement clean. It also ties several protection channels into one package and often one return structure. That shared return is useful when the lines enter together and discharge toward the same node. It can be a problem if unrelated signals with different references are forced to share a transient path.

Arrays also tempt designers to route lines to the package because it is convenient, rather than placing the package where the event enters. That is backwards. The ESD current should hit the protector first, then the protected trace should continue into the product. If the line wanders across the board before reaching the array, the discharge has already passed sensitive areas. For service connectors, buttons, display tails and external sensor headers, the first centimeter of routing can decide more than the device model.

The shared capacitance and leakage of an array may be harmless on slow lines and unacceptable on higher-speed or high-impedance nodes. A wake pin, reset pin or analog sense line can be disturbed by leakage, while a faster digital interface can be disturbed by capacitance and mismatch. The protection plan has to group lines by electrical behavior, not only by connector pin count. A neat array that groups unrelated lines may look tidy and still create a subtle failure.

An eFuse protects power in a way a passive fuse cannot

TPS2640 as an eFuse with overvoltage protection brings active control to the power input. An eFuse can limit current, disconnect during overvoltage, control inrush and report fault conditions to the host, depending on configuration. That makes it useful when the product should survive wrong adapters, hot-plug, cable faults or field wiring mistakes without relying only on a disposable fuse. It also makes the power input a managed interface rather than a bare rail.

An eFuse does not replace every protector around it. A fast surge may still need a TVS at the connector because the eFuse cannot absorb every transient by itself. A fuse or upstream protection may still be required by safety or installation practice. Thermal behavior also matters: current limiting turns electrical abuse into heat, and the board must give that heat somewhere to go. If the product is sealed outdoors, the temperature rise during a fault may be as relevant as the electrical threshold.

The value of an eFuse is diagnostic as well as protective. A passive fuse may leave a dead product. An eFuse can let firmware count faults, distinguish inrush from sustained overload, retry under control, or tell the installer that the supply is outside range. That only works if the fault pin, enable pin and reset strategy are wired into the product behavior. Active protection should not be a hidden component. It should become part of how the device explains power trouble.

A protected product gives energy a planned exit

Outdoor protection is not a list of stronger parts. It is a map of bad energy. A powered interface needs a TVS whose working voltage and clamp behavior fit the rail. A field input may need a higher-energy suppressor and a fuse that clears the long fault after the transient. CAN and vehicle-facing lines need fast, balanced, low-capacitance protection at the connector. Multi-line arrays save space only when the lines share a sane return. An eFuse adds control and diagnostics to the power input, but still needs the right fast front-end protection. The product survives because every outside conductor has a planned entrance, a planned clamp, a planned return and a planned failure state.