Short spikes, big consequences: Why transients threaten your electronics

Transients, often called overvoltage surges, are short-lived, violent variations in voltage or current that last shorter than a sine wave. Although they often last only microseconds, the energy content is often high enough to instantly destroy sensitive electronics or surreptitiously age them.

In practice, we see many organisations wrongly blaming transients on 'bad luck' or external factors such as lightning, while the cause often lies in their own installation. Correct diagnosis is essential for operational reliability.

In brief

What is it: A very short, fast pulse (impulsive or oscillatory) that disturbs the normal sine wave form.

The risk: Direct damage to printed circuit boards (breakdown), unexplained resets of PLCs and accelerated ageing of insulation.

The solution: high sample-rate measurement, source addressing, adequate surge protection (SPD) and filters.

Is this relevant to your situation?

Not every installation is equally sensitive to transients. The relevance increases with the presence of more power electronics and process-critical control.

  • Technical Managers in industry: You see variable speed drives falling into failure or I/O boards that need to be replaced regularly with no apparent cause.
  • Installation managers in data centres: Continuity is sacred. A single spike can mean data corruption or server failure, even if the UPS is in place.
  • Engineers in the maritime sector: Due to 'islanding' (self-generation), switching actions of thrusters or pumps are directly felt on the main rail, putting navigation or communication equipment at risk.
  • Facility managers in hospitals: Medical imaging equipment (MRI/CT) is extremely sensitive to voltage quality. Transients lead to poor images or outages.

The technical definition: Impulsive vs. Oscillatory

In the world of Power Quality, we define a transient as a "sudden change in the state of the system". Unlike harmonics (which are continuous) or dips (which last several periods), a transient is an 'event'. It is over before you blink, but the impact is like a sledgehammer blow.

Imagine a water pipe. If you close a tap in one go, you hear a loud thump in the pipes ('water hammer'). The pressure briefly rises extremely high. Exactly this happens in electrical cables when switching large currents.

We distinguish two main types (according to IEEE 1159 and IEC standards):

  1. Impulsive transients: These are sudden spikes in one direction (positive or negative). They have a very steep flank (fast rise time).
    • Most common cause: Lightning strike (direct or indirect) or electrostatic discharge (ESD).
    • Characteristic: A lot of energy in a very short time.
  2. Oscillatory transients: In this, the voltage oscillates very rapidly (high-frequency) around its rated value and then dampens out. The voltage briefly becomes very high, then low again, and so on.
    • Most common cause: Switching on or off of inductive or capacitive loads, such as capacitor banks or large transformers. Resonating of the line network also plays a role here.

Nuance: The difference with 'Voltage Swells' A common mistake is confusing a transient with a swell (voltage swell). A swell lasts relatively long (e.g. 100 milliseconds to a minute) and has a frequency of 50Hz. A transient lasts micro to milliseconds and contains frequencies from kHz to MHz. For the solution, this distinction is crucial: a voltage regulator solves a swell but is too slow for a transient.

The silent wrecking ball for your installation

The impact of transients is often underestimated because the damage is not always immediately visible. We see three degrees of impact within electrical installations:

  1. Direct destruction (Catastrophic) When there is a very high peak (e.g. due to lightning or a severe switching fault), the insulation of components fails. Semiconductors in frequency converters, server power supplies or PLC inputs burn out immediately. Result: immediate downtime and high replacement costs.
  2. Software and Data Failure (Operational) Transients can 'confuse' logic gates in digital systems. For a moment, a 0 is seen as a 1. This leads to corrupt data, crashed processors or unexplained operating system resets.
    • Example: A robotic arm in a production line randomly stops three times a week. After a reset, everything works again. Often, this is not a software fault, but a Power Quality problem.
  3. Degradation (Creeping) This is the most insidious form. Lower, repetitive transients (e.g. every time a heavy motor switches) 'tap' against the insulation of cables and windings every time. This creates hairline cracks in the insulation (dielectric stress). Over time, the component 'spontaneously' fails during normal operation. Operating reliability is slowly eroded.

What causes transients (Look beyond lightning)

To solve transients, you need to locate the source. Although lightning is the most well-known cause, most causes lie within one's own walls.

External Causes (approx. 20%):

  • Lightning strikes: Directly on the line or via induction in close proximity. This causes enormous energy impulses.
  • Switching in the high-voltage grid: When the grid operator (DSO/TSO) switches or a fault is switched off elsewhere, this can enter your installation via the transfer point (Point of Connection).

Internal Causes (approx. 80%): Most pollution you create yourself.

  • Switching of inductive loads: Switching off motors, transformers or coils creates a counter-EMK (electromotive force) that tries to maintain current, resulting in a voltage spike.
  • Capacitor banks: Switching on a capacitor bank for reactive current compensation almost always causes an oscillatory transient. If this frequency resonates with the installation, the voltage can rise dangerously high.
  • Power electronics: Modern inverters, welding equipment and thyristor controls continuously 'chop up' voltage. This switching behaviour (commutation) continuously causes small transients (notches) on the sine wave.
  • Short-circuits and fuse actions: The blowing of a fuse is an abrupt power interruption that generates substantial spikes.

Case study: A manufacturing facility was suffering from faulty power supplies of LED lighting in its office spaces. HyTEPS measurements showed that whenever the large refrigeration compressors in the hall next door switched off, a peak of 800V occurred on the low-voltage grid. The LED drivers were only specified to 500V. Cause: internal inductive kickback. Solution: damping at the source (the compressors).

Measurement is knowledge: Why your standard meter is failing

The insidious thing about transients is their speed. A transient often lasts only a few microseconds (millionths of a second).

A standard multimeter or building management system often measures at an interval of seconds or minutes. For a transient, that's an eternity. You see "230V" on your screen, while in reality a peak of 600V passes hundreds of times per second.

Symptoms in practice:

  • Equipment fails without the fuse.
  • Printed circuit boards show black burn marks at the entrance.
  • Protection devices (SPDs) regularly trip or need to be replaced frequently.
  • Audible crackling or tapping in switchgear cabinets (overloading).

How do you measure it? To capture transients, you need advanced Power Quality Analyzers that support continuous waveform recording with a very high sample rate (e.g. MHz range). HyTEPS engineers use equipment that not only measures averages, but captures every microsecond of the sine wave. This is the only way to see the shape, frequency and amplitude of the peak, which is crucial for finding the source.

Strategy for security and optimisation

Blind installation of a surge protection device (SPD) is often not enough, especially with internal, repetitive transients. We use a three-step approach:

1. Solve at the source (Elimination) If transients arise internally, try to attenuate them there.

  • RC-Snubbers: Place these over the coils of contactors or relays. They attenuate energy immediately on switching off.
  • VFD settings: Adjust the switching frequency or ramp angle in variable speed drives, or install line reactors/chokes.
  • Synchronised switching: For capacitor banks, you can use switches that switch exactly at voltage zero crossing to minimise inrush currents.

2. Isolate the path (Impedance & Earthing) Ensure that faults cannot spread easily.

  • Separation: keep sensitive data cabling physically separated from heavy power cables.
  • EMC guidelines: Ensure correct earthing and potential equalisation. A poor earth causes filters and shielding to fail.
  • Isolation transformer: A transformer can act as a buffer for high frequencies.

3. Protecting the victim (Mitigation) As a last resort, or for external events (lightning), apply protection.

  • TVSS / SPD (Surge Protection Devices): These 'cut off' the voltage spike (clamping) and direct the energy to earth. Note: There are different types (Type 1 for coarse, Type 2 for distribution, Type 3 for fine). A Type 3 at the server makes no sense if there is no Type 1/2 at the main connection.

Common mistakes in voltage dips

Blaming the grid operator directly: Although many dips come from outside, the grid operator is not always liable. Standard EN 50160 only gives indicative values for dips and does not set a hard limit on the number of dips per year, as they often occur due to force majeure (weather, third parties).

Focusing only on average voltage: Many meters measure averages over 10 minutes. A dip often lasts milliseconds and is completely missed by simple meters. You need sophisticated Power Quality meters that record 'events'.

Symptom management: Replacing a fuse or resetting a machine does not solve the problem. Without diagnosis, the risk of recurrence remains.

Confusion with 'Notching': Notching (notches in the sine wave) looks like a dip, but is a repetitive phenomenon caused by thyristors in DC drives. This requires a different solution (filters) than an occasional voltage dip.

Avoid these 5 pitfalls

Focusing only on lightning: And forgetting that the lift motor or welding robot causes much more damage internally.

Wrong SPD selection: Placing an SPD with too low a 'clamping voltage' can cause it to wear out too quickly, or too high, thereby failing to protect equipment.

Poor earthing: The most expensive surge protector will not work if it cannot dissipate its energy to a low-impedance earth.

Measuring at the wrong time: Measuring a week when production is down gives a false sense of security. You should measure during worst-case scenarios (start-up, switchover).

Symptom management: Always replacing faulty cards without asking why they break down.

Checklist: Is your installation protected?

  1. Inventory: Do I have critical electronics connected to the same grid as heavy, switching loads?
  2. Visual inspection: are SPDs present? Is the indication still green (safe) or red (faulty)? Are earth connections intact?
  3. Incident analysis: Is there a pattern in outages? (For example: always at 07:00 on start-up?)
  4. Power Quality Measurement: in case of doubt or unexplained faults, have a baseline measurement carried out with high-speed analysers.

When to call in a specialist? Are you experiencing unexplained failure of controllers, frequent damage to circuit boards or are you about to commission a new production line with lots of power electronics? Don't wait for downtime to strike.

Need help with diagnostics?

Do you suspect transients are disrupting your processes? Our engineers will be happy to help you with an installation analysis and a concrete improvement plan. Speak to an engineer to discuss your situation.

HyTEPS

Beemdstraat 3

5653 MA Eindhoven