The corona effect is a localized electrical discharge that occurs when high voltage ionizes the air immediately surrounding a conductor, creating a faint glow, hissing noise, and ozone without causing a full sparkover. If you are designing high-voltage power supplies, flying high-altitude drones, or wiring Tesla coils, corona is the invisible thief stealing your efficiency and eating your insulation. It changes a clean DC or AC signal into a source of broadband radio frequency interference (RFI) and generates ozone gas, which rapidly degrades silicone and rubber components. Before we look at how to stop it, we need to understand the exact physics of why air decides to become a conductor.
The Core Mechanism: Ionization Without Breakdown
Air is an excellent insulator right up until the electric field gradient exceeds its dielectric strength. At standard temperature and pressure (STP), dry air breaks down at roughly 3.0 kV/mm (30 kV/cm). When the voltage gradient at the surface of a conductor exceeds this threshold, stray electrons in the air are accelerated fast enough to knock other electrons off nitrogen and oxygen molecules. This creates an electron avalanche.
However, unlike a full arc or spark, this avalanche doesn't bridge the entire gap to the return electrode. The electric field strength drops off exponentially as you move away from the conductor's surface. The ionization is intense right at the wire, but a few millimeters out, the field falls below the sustaining threshold. The result is a localized, self-limiting plasma sheath—the corona discharge.
- Corona Discharge: Happens in the fluid/gas (usually air) surrounding a conductor. It is localized and does not bridge the gap.
- Partial Discharge (PD):strongstrong>Micro-sparks that occur inside solid or liquid dielectrics (like voids inside epoxy potting or cable insulation). PD is a precursor to catastrophic insulation failure.
- Arcing/Sparkover: A complete, low-impedance plasma channel that bridges the entire gap between two electrodes, resulting in a massive current surge and a dead short.
While corona won't instantly short your circuit like an arc, it is not benign. The UV light and ionized oxygen combine to form ozone ($O_3$). Ozone is highly corrosive; it turns flexible silicone wire brittle, cracks heat-shrink tubing, and oxidizes copper and silver contacts. Furthermore, the rapid avalanche of electrons generates broadband electromagnetic interference (EMI), which can easily desensitize nearby RF receivers or inject noise into high-gain audio preamps.
The Math in Practice: Calculating Onset Voltage
To predict whether a conductor will trigger corona, we calculate the electric field gradient ($E$) at its surface. For a cylindrical wire suspended above a ground plane, the maximum surface gradient is approximated by:
$E = \frac{V}{r \ln(R/r)}$
Where:
$V$ = Voltage (in volts)
$r$ = Radius of the conductor (in mm)
$R$ = Distance to the ground plane or return electrode (in mm)
Let's run a worked numeric example to see why wire gauge matters immensely in high-voltage (HV) design. Suppose we are routing a 15 kV DC line suspended 50 mm above a grounded chassis.
Scenario A: Using 12 AWG Wire
- Radius ($r$): 12 AWG solid copper has a diameter of ~2.05 mm, so $r = 1.03$ mm.
- Distance ($R$): 50 mm.
- Calculation: $E = 15,000 / [1.03 \times \ln(50 / 1.03)]$
- $E = 15,000 / [1.03 \times 3.88] = 15,000 / 3.996$
- Result: 3,753 V/mm (3.75 kV/mm)
Because 3.75 kV/mm exceeds the 3.0 kV/mm air breakdown threshold, corona will form on this 12 AWG wire. You will hear a hiss and smell ozone.
Scenario B: Using 2/0 AWG Wire
- Radius ($r$): 2/0 AWG has a diameter of ~9.27 mm, so $r = 4.63$ mm.
- Distance ($R$): 50 mm.
- Calculation: $E = 15,000 / [4.63 \times \ln(50 / 4.63)]$
- $E = 15,000 / [4.63 \times 2.378] = 15,000 / 11.01$
- Result: 1,362 V/mm (1.36 kV/mm)
Because 1.36 kV/mm is well below the 3.0 kV/mm threshold, no corona will form. This mathematical reality is exactly why high-voltage transmission lines use thick, bundled conductors rather than thin wires, and why sharp points on HV terminals are a massive liability.
Where You Meet This in Practice
You don't need to be working on 500 kV transmission lines to encounter corona. According to research published by the Electric Power Research Institute (EPRI), corona losses scale non-linearly with voltage and are highly sensitive to surface irregularities. Here is where it actively ruins projects:
- High-Voltage PCBs (Flyback & Ignition Coils): Sharp 90-degree trace corners concentrate the electric field. Even at 5 kV, a sharp corner with a 0.1 mm radius will easily exceed 3 kV/mm, initiating corona that eventually carbonizes the FR4 substrate, creating a permanent conductive track.
- High-Altitude UAVs and Drones: Air density drops with altitude. According to Paschen's Law, the dielectric strength of air decreases as pressure drops (up to a point). A 10 kV wiring harness that is perfectly safe at sea level will violently spark and corona at 30,000 feet. This is a primary failure mode in high-altitude pseudo-satellites (HAPS).
- Solder Joints on HV Multipliers: A standard rosin-core solder joint often leaves a microscopic sharp spike of solder. At 20 kV, that tiny spike acts as a localized field concentrator, bleeding current into the air and generating enough RFI to scramble nearby microcontroller ADCs.
Decision Tree: Mitigating Corona in High-Voltage Designs
Stopping corona requires either lowering the voltage gradient (by increasing the radius or distance) or increasing the dielectric strength of the surrounding medium. Use the IPC-2221 standard for baseline PCB clearances, then apply this decision matrix to finalize your physical design.
| Design Scenario | Condition / Trigger | Required Action | Concrete Pick / Value |
|---|---|---|---|
| High-Voltage PCB Traces | Voltage > 2 kV DC/AC peak | Increase clearance per IPC-2221 B2/B3 tables; eliminate sharp corners; apply high-dielectric conformal coating to displace air. | Use MG Chemicals 422C Silicone Conformal Coating (dielectric strength ~14 kV/mm). |
| Free-Air HV Wiring | Surface gradient > 3 kV/mm (calculated via formula above) | Increase conductor radius or encapsulate the termination points to eliminate air gaps. | Apply 3M Scotchcast 2130 Electrically Insulating Resin over all lugs and sharp joints. |
| HV Terminal Hardware | Bare metal electrodes > 30 kV in free air | Install a toroidal ring to artificially increase the surface radius and distribute the field gradient evenly. | Install a 150mm OD Aluminum Corona Ring (minimum) on the terminal. |
| High-Altitude Wiring | Operating altitude > 15,000 ft (4,500 m) | Derate air dielectric strength by 40%; encapsulate entire HV harness to remove air from the equation entirely. | Use Loctite EA 9460 structural epoxy potting for all HV connectors. |
Frequently Asked Questions
Can you hear and smell the corona effect?
Yes. Corona on a 60 Hz AC line produces a distinct 120 Hz hissing or crackling hum due to the physical expansion of air during the ionization pulses. The accompanying smell is ozone ($O_3$), which has a sharp, metallic scent often compared to chlorine or a running photocopier. If you smell ozone near an HV bench supply, you have active corona.
Is corona discharge the same thing as St. Elmo's Fire?
Yes. St. Elmo's Fire is simply the natural, atmospheric manifestation of the corona effect. It occurs when the ambient electric field during a thunderstorm exceeds the breakdown threshold of air around sharp, grounded objects like ship masts, airplane wings, or lightning rods, resulting in the same localized, non-arcing plasma glow.
Does corona always lead to catastrophic failure?
Not immediately, but it guarantees long-term degradation. While the discharge itself won't trip a breaker like a dead short, the ozone it produces will chemically attack organic insulation. Silicone jacketing will develop micro-cracks within weeks, allowing moisture ingress, which then lowers the surface resistance and eventually triggers a full arc-over. Treat corona as a ticking clock for insulation failure.
How does humidity affect corona onset?
High humidity actually lowers the corona inception voltage slightly. Water vapor molecules are polar and more easily ionized than dry nitrogen or oxygen. Furthermore, condensation on a conductor creates microscopic water droplets. Because a droplet has a very small radius of curvature, it acts as a massive field concentrator, triggering localized corona at voltages far below the theoretical dry-air calculation.
