Surge Protection Design for a Low Voltage AC Supply Port

A 36 volt AC supply rail does not look like a high risk interface. The voltage is low, the current is modest, and the connected equipment is often a controller, a sensor or a relay board. The mistake is to judge the risk by the rated voltage rather than by the transient that arrives at the port. A lightning induced surge, the switching overvoltage produced when a contactor opens, or a fault current elsewhere on the same network can raise the port voltage into the kilovolt range for a few microseconds, and the devices behind it are the ones that suffer.

That combination of a low nominal voltage and sensitive downstream circuitry is what makes surge protection on these ports a design task rather than a component purchase. The protection has to clamp the transient below what the downstream device can withstand, divert the energy somewhere other than the signal path, and survive repeated events without degrading into a short circuit. Each of those requirements points at a different part of the design.

Why a Low Voltage Port Still Needs Protection

Equipment supplied from a 36 volt AC rail tends to be precision hardware: programmable controllers, instrumentation, medical monitoring equipment, and auxiliary systems in rail vehicles. These devices often contain microcontrollers, analogue front ends and communication interfaces with limited transient tolerance, and their small geometry makes them more vulnerable than the power stage that feeds them. A transient that the supply rides through can still destroy an input circuit.

The disturbance sources are close to the application rather than distant. Contactors and relays on the same network produce fast inductive kickback when they open. Motor starting produces a dip followed by a recovery transient. A short circuit elsewhere raises the potential of the supply conductors for the duration of the fault. Electrostatic discharge adds a direct injection path when a connector is touched. Together, these events justify protection at the port even though the nominal voltage is low.

Coordinating Devices in a Surge Path

Protection is built from devices with different characteristics, and the design work lies in making them cooperate. A metal oxide varistor clamps quickly at a voltage related to its rating, which makes it suitable for limiting the transient that reaches the load, but its energy handling capability is finite and it degrades with each event. A gas discharge tube can divert far more energy, but its breakdown voltage is higher and its response is slower, and it may not extinguish until the current falls below its holding value.

Placing a varistor in series with a gas discharge tube is the usual way to combine the two. The varistor limits the initial rise and takes the first part of the transient; the tube then conducts and carries the bulk of the energy, while the varistor in series limits the current through the tube and helps it extinguish. The combination also reduces the leakage that a varistor alone would present at the working voltage, which matters on a supply where a small standing current is undesirable.

Selecting the Parts

The varistor rating must be chosen from the continuous working voltage rather than from the nominal supply. Its maximum continuous operating voltage has to sit above the highest steady state voltage the port can present, including regulation tolerance and any temporary rise, because a varistor that conducts in normal operation will overheat and eventually fail. The clamping voltage at the surge current then has to be compared with what the downstream circuitry can tolerate, with margin for the impedance of the path between them.

The gas discharge tube is selected for its breakdown voltage, its impulse current capability and its insulation resistance. It must break down above the highest voltage the port can present and below the level that damages the equipment, and its leakage at the working voltage must be low enough not to disturb the circuit. Because the tube is the slowest element, the varistor in series determines when the combination begins to conduct, which is why the pair is specified together rather than as two independent choices.

Layout of the Surge Path

The layout of the protection is part of its function. The protection devices belong at the connector, before the transient has any opportunity to couple into the rest of the board, and the path from the protected node through the device to the earth reference must be short, direct and wide. A long thin trace adds inductance, and the voltage developed across that inductance appears at the load even though the protection device itself clamped correctly.

The earth connection deserves particular attention because it is the path the surge current takes when it leaves the board. It should be a low impedance connection to the chassis or the protective earth, wide enough to carry the transient without a significant voltage rise, and it should not share a path with signal returns. The principles that apply are those of EMI suppression and the connector area practice described for ESD on board edges, with the mechanical aspects of the mounting and earth bonding covered in board outline and mounting design.

Fuse Coordination and Common Mode Paths

A fuse in front of the protection serves overload and short circuit protection, and its rating must be coordinated with the surge devices so that a transient does not open it and a fault does. Differential mode protection across the supply lines handles the surge between conductors, while common mode protection from each line to earth handles the surge that arrives on both conductors together; both paths need a device, and the common mode devices must be able to withstand the full test voltage without degrading. Where the assembly is coated after production, verify that the conformal coating cannot bridge the isolation gap around the protection devices.

Testing and Verification

The protection should be verified against the standard the product is expected to meet rather than against a functional check. Surge immunity is applied to the port with the specified waveform and coupling network, with the equipment operating in its normal configuration, and the criterion is not simply that the unit survives: it must continue to function without reset, data loss or a change of state that the application would notice.

Repeat the test at the extremes of the supply voltage, because the headroom between the working voltage and the clamping voltage changes with the input, and repeat it after the product has been assembled into its enclosure, since the enclosure and the earth bonding are part of the surge path. Where the protection devices are socketed or accessible, check their leakage after the test to confirm that the varistor has not degraded into a state that draws standing current.

FAQ

Is surge protection needed on a 36 volt supply? Yes. The rated voltage describes normal operation, not the transient that arrives at the port. Equipment with sensitive inputs can be damaged by a disturbance that the supply itself tolerates.

Why not use a single varistor for both modes? Because differential and common mode surges follow different paths. A device across the lines does not limit the potential of both lines relative to earth, which is the direction a common mode surge takes.

How many events can a varistor survive? It depends on the energy of each event and on the part. Degradation is cumulative, so the protection should be specified for the number of events expected over the product life, not for a single test.