Single Point Ground and Surge Protection

Every conductor connected to a piece of equipment is a potential path for noise — and for surges. A well-designed ground system gives every circuit a single, unambiguous reference, which eliminates the noise loops that plague systems grounded at many points. A well-designed surge protection scheme diverts the transients that arrive on power, signal and data cables before they reach sensitive components. This final lesson in the Interference and Noise module brings the two together.

The ideas apply at every scale: inside an audio amplifier, on a mixed-signal circuit board, across a rack of equipment, and at the point where power, network and cable services enter a building. In each case, the aim is the same — one reference point, short connections to it, and no path for a noise current or a surge to flow through the equipment between two different grounds.

Single-point entry: every service and cable that enters the building is bonded to one main earthing point, so during a surge they all rise together and no large voltage appears across the equipment connected to them.

View Larger

Why Single-Point Ground?

In a system grounded at many points, each circuit or piece of equipment connects to ground through whatever conductor is nearest — a mains safety earth, a screw into a rack, a cable shield bonded to a different piece of equipment at each end. The result is a network of closed conductive loops, and the “ground” is not at the same voltage everywhere.

Any changing magnetic field threading those loops — from a transformer, a switching supply, or a motor — induces a circulating current. Any current that one circuit returns through a shared ground conductor creates a voltage drop that the next circuit sees as noise on its reference. Either way, the noise is not arriving through the signal input; it is injected directly into the signal path through the ground wiring.

A single-point ground breaks those loops and separates those currents. Every circuit or chassis connects to one common point through its own conductor. No two circuits share a return conductor, and no loop is formed.

The benefit can be dramatic. Moving an audio amplifier or a measurement system from haphazard grounding to a proper star ground routinely removes hum and noise that no amount of filtering would cure.

Star Grounding in Circuits

Audio amplifiers and linear power supplies

The classic application is the audio power amplifier. The large charging pulses that flow from the rectifier into the reservoir capacitors, and the large speaker currents, must not flow through the ground conductors that the sensitive input stage uses as its reference. The standard arrangement makes the negative terminal of the reservoir capacitors the star point. The rectifier return, the speaker return, the output stage decoupling, the input ground and the chassis connection each run to that point by a separate conductor. Hum that disappears when the input is shorted but remains when the input ground is moved is almost always a star-ground problem.

Mixed-signal circuit boards

On a board with both analog and digital circuits, the digital circuits return large, fast current pulses. If those currents flow under a sensitive analog section, the voltage drops they cause appear in the analog measurements. Modern practice is usually a single, unbroken ground plane with the board laid out so that digital return currents stay in the digital area and analog currents in the analog area, with the ADC placed at the boundary. Where separate analog and digital grounds are used, they are joined at exactly one point, normally at or under the ADC. The PCB layout lessons in the next module cover this in detail.

Bench and test setups

On the workbench, the same principle applies to wiring a test setup: bring each instrument's ground lead to one point on the circuit under test, rather than connecting the scope ground at one place, the generator ground at another and the power supply return somewhere else. Each separate connection carries part of the circuit's current and adds its own voltage drop to the measurement.

Single-Point Earthing for Systems and Buildings

Equipment racks and systems

In a rack or a multi-unit system, each unit should be bonded to the rack by short, wide connections, the rack bonded to one earth point, and signal cables between units kept short and run alongside the bonding path. Where long cables connect equipment in different places — different rooms, or different buildings — their ground potentials can differ by volts during a fault or a surge. Isolated interfaces such as optical fiber, isolated RS-485 and transformer-coupled Ethernet avoid connecting the two grounds through the signal cable.

Building service entrances

At building scale, the same idea is called single-point entry. Electrical codes require that the earthing of every service entering a building — mains power, telephone, cable television, satellite dish and antenna cables — be bonded to the main earthing point of the electrical system, close to where the services enter. Cable TV and satellite coax pass through a grounding block bonded to it; telephone and network lines pass through primary protectors bonded to it.

The reason is surge protection. During a nearby lightning strike, the ground potential around the building can rise by thousands of volts relative to distant earth. If every service is bonded to one point, they all rise together and no large voltage appears between them. If the cable TV is earthed to a separate rod at the other end of the building, a large voltage can appear between the coax and the mains earth — right across the television, computer or network equipment connected to both.

Surges and Surge Protection Components

Gas discharge tubes, MOVs and TVS diodes compared. Staged protection uses a high-energy device at the connector, a series impedance, and a fast, tight-clamping TVS diode at the circuit.

View Larger

The figure compares the three main surge-protection components. Each has strengths and weaknesses, and good protection designs often combine them.

Where surges come from

A surge is a short, high-energy overvoltage. The main sources are:

• Lightning: A direct strike is rare, but nearby strikes induce large surges in power, telephone and data lines, and raise the local ground potential.
• Switching transients: Utility switching, and large inductive loads such as motors, compressors and welders switching off, produce surges of several hundred to a few thousand volts on mains wiring.
• Electrostatic discharge (ESD): A person walking across a carpet can charge to 10–15 kV. The discharge into a connector lasts only nanoseconds but can destroy unprotected semiconductor inputs.
• Automotive transients: Disconnecting a charging battery while the alternator is running produces a “load dump” surge of tens of volts lasting hundreds of milliseconds on a 12 V system.

Standards define test waveforms so that protection can be specified and compared: the 1.2/50 µs voltage and 8/20 µs current surge of IEC 61000-4-5, the ESD pulse of IEC 61000-4-2, and the automotive pulses of ISO 7637 and ISO 16750.

Gas discharge tubes (GDTs)

A GDT is a sealed ceramic tube containing an inert gas between two or three electrodes. When the voltage exceeds its breakdown voltage (typically 75 to several thousand volts), the gas ionizes and the tube becomes a near short circuit, carrying thousands of amperes. GDTs have very low capacitance (around 1 pF), so they suit high-frequency and data lines, but they are slow to fire, and once conducting on a DC or mains circuit they may not turn off again without help.

Metal oxide varistors (MOVs)

An MOV is a zinc-oxide ceramic disc whose resistance falls sharply above its rated voltage. MOVs absorb large surge energies cheaply and are the main component in mains surge protectors. Their drawbacks are high capacitance (hundreds of picofarads to nanofarads), a clamping voltage well above their rated voltage, and gradual wear: each large surge degrades them slightly, and an MOV that is repeatedly stressed can eventually fail short and overheat, which is why quality protectors include a thermal fuse.

Transient voltage suppressor (TVS) diodes

A TVS diode is a silicon avalanche diode designed to absorb surge current. It responds in picoseconds, clamps tightly, and does not wear out with use, but it handles less energy than an MOV or GDT of similar size. TVS diodes and TVS arrays are the standard protection for low-voltage DC inputs, data lines and connectors on circuit boards; low-capacitance versions protect USB, HDMI and Ethernet lines.

The components are often used in stages: a GDT or MOV at the cable entry absorbs most of the energy, a small series resistor or inductor limits the current that gets past, and a TVS diode at the protected circuit clamps what remains to a safe level. The series impedance between the stages is what makes them share the surge correctly.

Critical installation requirement

A surge protector works by diverting surge current to ground, so the connection to ground must be short and direct. A protector connected through a meter of wire has an inductance of about 1 µH; a surge rising at 1 kA/µs develops 1,000 V across that wire alone, added to the protector's own clamping voltage. On a circuit board, place TVS diodes right at the connector, with a short, wide path to the ground plane or chassis.

Mains Surge Protection

Surges arrive through the mains wiring as well as signal cables. A nearby lightning strike or a large switching event on the distribution network can send a transient through the building that damages any equipment connected to it.

Whole-building surge protection

A surge protective device (SPD) at the main electrical panel — Type 1 or Type 2 in the United States, or Type 1/Class I or Type 2/Class II under IEC 61643 — clamps surges arriving on the utility conductors before they reach any outlet. Installed by an electrician, it is the first and most important layer of protection for all the electrical equipment in the building. Its earth connection must be short and go directly to the main earthing point.

Point-of-use protection

A quality surge-protected power strip at the workbench or equipment rack adds a second layer (Type 3). Look for UL 1449 listing or the IEC 61643 equivalent, a low voltage protection rating (VPR, around 400–600 V for 120 V circuits), a substantial energy rating in joules, and protection on all three conductor pairs. A power strip that also has connections for coax, Ethernet or telephone lines gives every cable to the protected equipment the same ground reference, which is often its most valuable feature.

Cheap strips often contain a single small MOV that degrades and eventually stops protecting, while the indicator light may still glow. Replace surge strips after a major surge event or after several years of service.

Uninterruptible power supplies (UPS)

A UPS adds surge suppression, filtering and battery backup. Online (double-conversion) UPS units regenerate the output from the battery inverter all the time and give the best isolation from mains disturbances; line-interactive and standby units are cheaper and pass the mains through most of the time, relying on their own surge suppression.

Protecting Your Own Circuits

Equipment you design needs its own protection at every point where a cable connects to the outside world.

• DC power inputs: A TVS diode across the input, a fuse or PTC resettable fuse ahead of it, and reverse-polarity protection (a series diode, or a MOSFET for lower loss). In vehicles, the TVS must be rated for load-dump energy.
• Signal inputs: A series resistor to limit current, followed by clamp diodes to the supply rails or a TVS array. Check that the protection does not overload the source or add too much capacitance for the signal's bandwidth.
• Data connectors: Low-capacitance ESD protection arrays at USB, HDMI and Ethernet connectors, placed as close to the connector as possible.
• Long cables to outdoor sensors: Treat any cable that runs outdoors or between buildings as a surge path. Use isolated interfaces where possible, primary protection at the building entry, and TVS protection at the board.
• Static on outdoor conductors: Long outdoor wires and metal structures can charge up from wind-blown dust, snow and nearby storms. A high-value bleed resistor (for example 1 MΩ) from an otherwise floating conductor to ground lets this charge drain away before it builds up to a damaging voltage.

Thunderstorms and Outdoor Equipment

No practical protection system can survive a direct lightning strike to the equipment itself. Good protection greatly reduces damage from nearby strikes and induced surges, but common sense still applies when storms are near.

Best practice

• Unplug sensitive or valuable equipment that is connected to outdoor cables, and disconnect the outdoor cables themselves, when a storm is approaching. Disconnection is the only complete protection.
• Do not work on outdoor wiring, antennas, solar arrays or roof-mounted equipment during or immediately before a storm.
• Do not handle equipment that is connected to long outdoor cables during a storm.
• After a storm, check surge protector indicators and replace any that show a fault.

Design Checklist