Although miniaturisation has been a theme in electronics for more than 50 years, reflecting advances in semiconductor and passive-component manufacturing technologies, there are limits on how small some components can get.
The capacitor is a prime example of a component where small is not always beautiful, as meeting electrical characteristics and safety demands means that a component small enough to solder directly onto a PCB may just not be up to the job.
Capacitors are widely used in filtering and decoupling. For small-signal applications the capacity of the device is determined by the filter topology used and the frequency range over which the filter operates. When dealing with power-rail circuitry, the capacitors need to be able to store very high levels of electrical energy, whilst satisfying safety requirements. This is a prime example of where big is beautiful as the power handling capability of the capacitor and isolation built-in are more important than the size of the component.
Size, especially the minimum distance between power-handling traces on the circuit board or between cables, is a critical parameter in many safety standards, leading to many manufacturers directing improvements in materials and component design to deliver higher levels of capability rather than reducing size. Many of the big and beautiful capacitors are designed for filtering in power systems.
Class-X and Class-Y capacitors
Class-X and Class-Y capacitors are some of the most commonly used large filter capacitors in AC applications and are often known as EMI/RFI suppression capacitors, or AC line filter safety capacitors. Due to safety risks, caused by connection directly to the AC power input across the live and neutral lines (Class-X) or neutral and ground (Class-Y), these connectors are all safety certified and if they fail, it must be in a manner that ensures safety.
Class-Y capacitors are particularly subject to the over-voltages and voltage transients often seen on domestic and industrial AC supplies. Lightning strikes nearby, for instance, can lead to extremely high-voltage transients and there is a high probability of damage to the device. A failure in a Class-Y device, however, could lead to a potentially fatal shock because of the loss of the ground connection and to avoid this, Class-Y capacitors are therefore designed to fail open, which causes a loss of filtering capability but avoids a safety hazard.
When a Class-X capacitor fails it is most likely to cause a short circuit, which in turn will normally engage an over-current protection device such as a fuse or circuit breaker. The capacitor’s failure is therefore unlikely to cause an electric shock to the user.
The degree to which a capacitor is prone to failure depends on its internal construction. In ceramic X and Y capacitors, damage can occur over time that weakens the device until it fails. Metallised-film capacitors, however, such as Vishay’s MKP3381 family or the B32916 series from TDK-Epcos, have self-healing capabilities that avoid short-circuit and related damage. At the point of failure, an insulating region forms so that the capacitor can regain almost complete operational capability, avoiding a short circuit. Metallised film in Y-type devices is also more likely to fail open than ceramic-based devices, improving safety.
There are a number of X and Y capacitor subtypes defined in the international standard EN 60384-14. X-type capacitors are divided into X1, X2 and X3 categories: X1 capacitors are used for applications that need high pulse resilience and X2 and X3 types are used for general-purpose applications.
Y-type capacitors are similarly divided but differ according to the level of insulation: Y1 class capacitors are rated up to 500 V a.c. and a peak test voltage of 8 kV, scaling down to Y4, which are rated to 150 V a.c. with a peak test voltage of 2,5 kV. To improve density, manufacturers such as Kemet build more than one capacitor into the same package, for instance the PHE840M which has a maximum capacitance of 10 µF.
Power factor correction
Larger capacitors are also frequently used to perform power factor correction, which is enforced by law in many territories and can reduce the cost of electricity delivery.
The power factor of a system is the ratio between the true power used by the system (averaged over an AC cycle) and its peak consumption. If linear power converters are used on the front end of a system, they normally exhibit a power factor close to one, creating some issues. These converters are inefficient when compared with more advanced switched-mode front-end power supplies used in many systems; they take power in brief bursts, rather than over the full cycle of an AC wave, resulting in high peaks relative to the true power and a low power factor.
High-voltage capacitors in sizes of 50 µF or more per AC phase can be employed in power factor correction modules, usually installed in parallel to the user equipment, to improve the capacitance density and ease design-in. Suppliers such as TDK-Epcos and Kemet provide capacitors that have terminals for all three phases, each connected to a separated internal capacitance module.
Large capacitors are invaluable for filtering and spike suppression purposes within the system, particularly when dealing with highly inductive loads such as in motor control. Polypropylene film capacitors, similar to those often employed for power factor correction, are effective choices for spike-suppression applications due to their self-healing properties, low losses and high isolation capability. Sizes for these devices can range from several hundred nanofarads to hundreds of microfarads, with products supplied by a wide range of manufacturers, including Kemet, TDK-Epcos, Vishay and Wima.
Motors can draw high currents both when starting up and when brought to a sudden stop. This is a particular problem in battery-powered systems that may struggle to supply the instantaneous current required. The inclusion of large capacitors able to supply large current pulses gives the system the ability to ride through such events and avoid brownouts in other parts of the system. Typically, electrolytic capacitors such as Kemet’s MS/MD family of aluminium electrolytic capacitors, provide the low series resistance and capacity needed to support the system for the milliseconds during the peak current demand.
High-current relays within a system present another source of energy spikes that need to be suppressed: arcing across contacts when the relay is opening or closing. If not controlled, repeated arcs erode the relay contacts until it fails. A capacitor wired in parallel with the relay’s load will suppress the arcs because it can charge and discharge more quickly than the time taken for the relay to complete a switching operation.
As with other high-power filtering and suppression applications, effective capacitors use a metallised film or foil construction. As charging time is a vital consideration, component selection needs to take into account the relationship between capacitance required and the effective series resistance. Furthermore, as many film capacitors provide an ESR (effective series resistance) that is lower than that required for a typical suppression circuit, a low-value resistor is often wired in series.
As the ability to survive large voltage and current spikes – as well as handle high power levels on a long-term basis – is key to many applications, capacitors that are both large in capacitance and in size are essential.
Manufacturers continue to work on improving in-system density through co-packaging as well as on performance, but in these applications small is not beautiful: the design-in decision comes down to capability.