Beyond the deceptively simple schematic symbol of the polarized capacitor (see Figure 1) is a sophisticated, vital component of many electronic circuits. This capacitor, often called an electrolytic capacitor or simply “electrolytic” due to its construction, plays an essential role in guaranteeing that the output of a power supply can source the current needed, and at the rated DC supply voltage.
Figure 1: The most-common symbol for polarized capacitor in the a) U.S., and b) Europe; there are many variations.
Electrolytic Capacitor Polarity Explained
Why use such a capacitor and why is it polarized? The primary role of this capacitor is to act as a reserve storage container of electrical energy for the load, even as the output of the power-supply itself—usually an AC/DC supply—has ripples at 60/120 Hz (50/100 Hz in some regions of the world) due to the nature of the power-regulation circuitry.
A 33uF aluminum capacitor from Lelon Electronics.
The capacitor is analogous to a reservoir: the core of the power supply is pumping energy (water) into the reservoir, but not at a steady rate. The load (the users) takes water out at varying rates, sometimes with slow changes and sometimes with sudden, transient increases in demand. They need to do this despite fluctuations in the main supply pipe leading from the water-purification plant. They do not want to see fluctuations in water pressure (voltage) despite changes in the flow rate (current) at the source or the load.
The capacitor is a cushion or buffer of electrical energy and so does two things—it smoothes out the ripples in the output of the basic regulator when the load is constant, and it supplies energy as needed when the load itself varies. For these reasons, large-value electrolytic capacitors used at the output of power supplies are often called “bulk-storage” components, and act as basic filters against unwanted output supply-voltage fluctuations despite changes in regulator input voltage or load demand.
How are Electrolytic Capacitors Made?
In principle, a capacitor is formed by two conductive surfaces separated by a dielectric. This dielectric can be air, paper, ceramic, or a specialized electrolytic chemical film. Most electrolytic capacitors are constructed from two very thin layers of metal foil (aluminum, tantalum, or niobium) with a dielectric oxide layer which is coated onto one layer, and then the entire assembly is rolled up (Figure 2).
Figure 2: Internal construction of an aluminum-based electrolytic capacitor shows the layers separated by a dielectric, and then rolled into a cylindrical housing. (Source: Nichicon Corp.)
The final unit is sealed with a specialized coating which can be plastic, epoxy, metal, or other material to keep moisture out while confining the electrolytic material inside in case of chemical “leakage” or failure of the case (Figure 3).
Figure 3: A completed electrolytic capacitor ready for use; this one is rated 10,000 μF (0.1 F), 15 VDC and is 40 mm high with a diameter of 18 mm. (Source: Kemet Corp.)
Why We Use Electrolytic Capacitor in Power Supply
With a non-chemical dielectric, the resulting capacitor is not polarized, and can be used with AC waveforms; also, it can be inserted either way in the circuit. However, due to the chemical nature of the film and construction used for electrolytic capacitors, there is a polarity of installation and use. Reversing the voltage on such as device will degrade and then damage it.
Given this constraint, why even use polarized electrolytic capacitors at all? The answer is simple: to achieve high capacitive density and related value. Most AC/DC power supplies need capacitance on the order of several hundred to ten thousand microfarads (μF), and this can only be achieved in a component of reasonable size using electrolytic-capacitor construction. Using ceramic or air as the dielectric would require a capacitor volume easily ranging between 100× to 1000× as great.
Cost is a consideration as well—a larger capacitor would require more material, so there will be higher direct cost as well as the higher “cost” of using more PC board space or a larger overall power supply. Supercapacitors might seem to be a better and smaller alternative as they can easily provide ratings of several farads, but they cannot handle the ripple current or charge/discharge nature of a power-supply regulator and its load.
Choosing an Electrolytic Capacitor: Design Parameters
The primary parameter for these bulk-storage devices is their capacitance, of course. Electrolytic capacitor values begin around 1 μF and go into the thousands of μF. If more capacitance is needed than a single component can provide, the capacitors can be used in parallel, of course.
The next parameter the designer must select is the working voltage, usually designated as WVDC (working voltage DC). This is the maximum DC voltage rating at which the capacitor will operate reliably, and is a function of the design and housing. A higher WVDC requires a larger physical-size device to withstand internal arcing and punch-through and is more costly, so the designer must be careful not to over-specify this factor. Most designers use a 2× safety margin on WVDC to accommodate any ripple or transients on the capacitor from the supply; thus, a 25-V WVDC capacitor would be used with a nominal 12-V DC power supply.
Although ideally a capacitor would be just that; in reality, every capacitor has some equivalent series resistance (ESR) and self-inductance. The ESR of a high-quality capacitor is on the order of 0.1 to 1 Ω; the higher the ESR is, the less the capacitor will perform like an ideal device, and it may actually cause the regulator circuit to malfunction. In lower-quality electrolytic capacitors, the ESR will increase over time and temperature, and may even reach tens of ohms, with detrimental consequences. Capacitors also have a small amount of leakage current due to the non-perfect dielectric.
Further, every real component has parasitic inductance, of course; for capacitors, this inductance is on the order of a few millihenries (mH). While this low value is generally not a problem at AC-line frequencies, it can be a problem as the power-supply operating frequency increases, and may cause instability in the circuit and even failure.
Electrolytic Capacitor Tolerance
Electrolytic capacitors also have tolerance ratings, as do all components; tolerance of ±20 percent is common, though some are specified at tighter tolerances. While this may seem like a large tolerance allowance, it’s acceptable in the application.
To support the designer’s performance and stability analysis, most capacitor vendors provide models that include the ESR, inductance, leakage resistance, and any other non-ideal attributes (Figure 4). They may show these at line frequency as well as higher frequencies, and also at different temperatures.
Figure 4: A simplified low-frequency model of an electrolytic capacitor shows the basic capacitor along with the leakage resistance, the equivalent series resistance, and inductance; for RF use, the model would add various internal parasitics as well as parasitic lead inductance and capacitance.
Electrolytic Capacitor Degradation
Electrolytic capacitors are usually expected to perform to specification for many thousands of hours, although they are often used beyond their maximum “to spec” lifetime with acceptable results. (Think of a power supply in a long-running desktop PC which is “on” much of the time.)
In addition to obvious operation outside the established ratings, every electronic component is subject to factors that affect its reliability and operating life, and electrolytic capacitors are no different.
Heat is the most common factor in shortening their life: a capacitor that is rated for 10,000 hours at 25⁰C will need derating as the temperature increases, and may only be rated for 1,000 hours at 85⁰C and even less at 105⁰C. Since most of these capacitors are used with power supplies, which generally run warm and have localized temperature rise above that of the overall enclosure, these bulk storage devices will see shorter life. Vendors do offer capacitors that are rated for long life at higher temperatures to overcome this problem. (Note that elevated non-operating storage temperature is also an issue affecting their life, but that is a different scenario and has a different specification.)
The second factor that shortens the life of electrolytic capacitors is the ripple current they must endure. This current is the unavoidable fluctuation in the output of the voltage regulator that the capacitor is charged with smoothing out. For complex electrochemical reasons, the ripple current degrades the life of the capacitor and its electrolyte; the higher the ripple current, the greater and faster the degradation. The sensitivity to ripple current is a function of the construction and materials used; vendors specify operating life with different ripple-current values.
There is one non-technical factor that designers must also keep in mind, after they have selected the appropriate capacitor and the corresponding vendor model. It’s relatively easy to have substandard, substitute, or outright counterfeit parts work their way into the production and assembly flow. This is because it is relatively easy to make an adequate capacitor that will work well enough, at least for a while. However, the product itself will have a shortened life in the field, but by then it is too late and will become major headache.
Keep in mind that it’s also tempting for the production facility’s purchasing group to substitute a “similar” capacitor for the one specified on the BOM by the designer, but with the same top-level specifications: capacitance, WVDC, and size. Yet it may have different secondary but still important specifications such as ESR or ripple-current tolerance, and the BOM change may affect system performance or reliability. It’s vital for engineers to work with the production supply chain to guarantee the integrity and traceability of the capacitor back to the specified source vendor.
Electrolytic capacitors located between the power-supply regulator and the load may seem mundane and even routine. Nonetheless, they are essential to providing a stable DC rail for the circuit. As a result, designers need to specify and select them based on their primary and secondary parameters and operating situation, and keep less-obvious supply-chain issues in mind as well.