Plasma cooling technology moves air with no moving parts. The mechanism, the measured heat transfer numbers, where it beats a fan, and where it does not.
Every electronic device that does useful work eventually has to move heat somewhere else. For sixty years the answer has been a fan: a motor, a bearing, and a set of blades pushing air across a heatsink. Plasma cooling technology answers the same question without any of those parts.
Plasma cooling technology is a method of moving air across a hot surface by ionising the air itself and accelerating the resulting ions with an electric field. The moving ions collide with neutral air molecules and drag them along, producing a bulk airflow known as ionic wind or electrohydrodynamic (EHD) flow. There is no motor, no bearing, no impeller and no moving surface of any kind. The only thing that moves is the air.
The practical consequence is that a plasma cooler has no mechanical wear mechanism. A fan fails because a bearing degrades. A plasma actuator has nothing that can seize, unbalance or rattle.
YPlasma builds its actuators on dielectric barrier discharge (DBD). The geometry is deliberately simple: two electrodes separated by a thin dielectric layer, one exposed to the air and one encapsulated. When a high-voltage AC waveform is applied across them, the air immediately above the dielectric breaks down into a thin, non-thermal plasma sheet.
Two properties of DBD matter for electronics. First, the discharge is non-thermal: electrons are energetic, but the gas stays close to ambient temperature, so the actuator does not add heat to the thing it is cooling. Second, the dielectric barrier is self-limiting, which suppresses the transition to an arc and keeps the discharge distributed rather than concentrated in a single filament.
The actuator itself can be built at a thickness of roughly 200 micrometres, which means the cooling element occupies space a fan physically cannot.
The claim worth making is not "plasma replaces all cooling". It is narrower and better supported.
Go, Garimella, Fisher and Mongia at Purdue superimposed an ionic wind on an existing 0.3 m/s bulk airflow over a heated surface. The ionic wind produced roughly a factor-of-two increase in the average heat transfer coefficient and about 20 K of additional cooling on a plate already cooled by that bulk flow, while drawing under 100 mW (67 mW at 15 µA) [1].
That number is the honest headline for plasma heat dissipation. Ionic wind is extremely effective at *local* enhancement, in exactly the boundary-layer-limited regions where a fan's bulk airflow has already given up. It is far less useful as a replacement for moving large volumes of air across a room.
The electrical-to-fluid energy conversion efficiency of a corona-based ionic wind device sits in the region of 1 to 2 per cent [1]. Stated bluntly: as a bulk air mover, plasma is inefficient. As a targeted boundary-layer disruptor placed millimetres from the hot spot, it wins, because the alternative in that space is no airflow at all.
The broader literature on ionic wind in thermal management, including cooling of plates, tubes, channels, power chips and heat exchangers, is surveyed by Pramanik et al. [2], and the underlying plasma actuator flow physics is set out by Moreau [3].
| Approach | Moving parts | Thickness | Fails by | Best at |
|---|---|---|---|---|
| DBD plasma actuator | None | ~0.2 mm | Dielectric ageing | Sealed, thin, silent, boundary-layer-limited spaces |
| Axial or blower fan | Motor, bearing, blades | 5–40 mm | Bearing wear, dust ingestion | Moving large air volumes cheaply |
| Piezoelectric or MEMS cooler | Vibrating membrane | 1–4 mm | Membrane fatigue | Thin consumer devices, moderate flow |
| Thermoelectric (Peltier) | None | 3–5 mm | Solder fatigue, thermal cycling | Sub-ambient spot cooling, poor efficiency |
| Heat pipe or vapour chamber | None (passive) | 0.3–3 mm | Wick dry-out | Spreading heat, not rejecting it |
The last row matters and is often skipped. A vapour chamber spreads heat brilliantly and rejects none of it. Something eventually has to hand the heat to the air. That is the job plasma cooling competes for.
The cases where a DBD actuator is the right answer share a common shape: the fan is either impossible or is the component that fails first.
Anyone evaluating plasma cooling technology should weigh four real limitations.
High-voltage drive. The actuator needs a kilovolt-class AC supply. That supply is small, but it is a component that has to be designed in, certified, and kept away from sensitive analogue lines.
Ozone. Any discharge in air produces some ozone. The quantity depends heavily on waveform, geometry and duty cycle, and DBD designs can be optimised to keep it well below occupational limits, but the question is legitimate and deserves a measured answer rather than a dismissal.
Bulk flow ceiling. Ionic wind produces metres per second, not tens of metres per second. If the thermal problem needs a large volumetric flow rate, a fan is still the correct engineering choice.
Dielectric lifetime. The barrier is under continuous electrical stress. Material selection and encapsulation determine the operating life, and this is where the engineering work in a commercial actuator concentrates.
Is plasma cooling the same as ionic wind cooling? Effectively yes. Ionic wind is the physical phenomenon; plasma cooling is the application of it to thermal management. YPlasma uses dielectric barrier discharge to generate that ionic wind.
How much heat can a plasma cooler remove? It removes heat by convection, so the figure depends on the surface area and the temperature difference rather than on the actuator alone. The useful measure is the enhancement: roughly a doubling of the local heat transfer coefficient in published work, for well under a watt of electrical input [1].
Does plasma cooling replace fans entirely? In sealed, thin, silent or dust-exposed applications, yes. For moving large air volumes at low cost, no. Treat it as the tool for the places a fan cannot go.
Is a plasma cooler safe to put next to electronics? The discharge is non-thermal and confined to a thin layer above the dielectric surface. The engineering work sits in the high-voltage supply design and in electromagnetic compatibility, both of which are standard, solvable problems.
How thin can a plasma cooling element be? YPlasma's actuators are built at around 200 micrometres, which is thinner than the vapour chamber in most laptops.