Inductors
An inductor is a magnetic energy storage element, and current flowing through the inductor will induce a magnetic field in the core. This magnetic field is the mechanism for the energy storage. Because current in an inductor cannot change instantaneously, when a voltage is applied across an inductor, the current will ramp. Figure 7 illustrates the current waveform in an inductor.
When the switch closes, the full voltage (V) appears across the inductor. The current in the inductor will ramp at a rate of V/L. When the switch is opened, the current will ramp down at the same rate, and a large voltage will be generated as the magnetic field collapses. This magnetic field is the energy storage mechanism. Figure 8 shows a simplified model of an inductor.
In addition to the inductance, there will be a series resistance (DCR) and a shunt capacitance. The DCR is mainly the effect of the wire coil resistance, and it will be important when calculating power loss in the inductor. The shunt capacitance, along with the inductance, can cause the inductor to self resonate. The self-resonant frequency can be calculated from:
A good rule of thumb is to keep the switching frequency one-tenth the self-resonant frequency of the inductor. In most designs this will not be an issue. Power losses within an inductor will cause a temperature rise within the inductor as well as lost efficiency. There are two main categories of power loss in inductors. Designers need to understand both.
Winding resistance (DCR) losses are simply I^2*R losses within the wire conductor, also known as copper losses. The other contributors to power loss in inductors are known as core losses. Core losses are a combination of magnetic hysteresis and eddy currents within the core. They’re much more difficult to calculate, and may not even be available on a datasheet, but will cause power dissipation and temperature rise within the core.
Figure 9 illustrates the inductors function in both buck- and boost-mode power-supply designs. The primary function of the inductor is energy storage, but it also acts as a filter. Inductor value selection begins with determining the maximum ripple current desired.
A good starting point is to use 30% of the dc load current for buck converters and 30% dc input current for boost converters. With this, inductor value can be calculated using the equations in Figure 9. Inductor tolerances can be as much as ±30% out of the box, so be sure to include this in your calculation. Also be sure to choose an inductor with:
where ISat is the inductor’s saturation current. The saturation current is the current at which the inductance will drop by a certain percentage. This percentage will vary by manufacturer, ranging from 10% to 30%. When choosing an inductor, be sure to note the saturation current change over temperature, as your inductor is likely to be operating at high temperature.
Operating at a 10% reduction in inductance is generally acceptable, providing this is a worst-case scenario. Using inductors that are larger than necessary will take up more PCB real estate and are usually more expensive. A higher switching frequency will allow the use of lower-value inductors.
There are two main core materials used in inductors for SMPS: powdered iron and solid ferrite. A powdered-iron core has air gaps within the material that provide for a “soft” saturation curve. Because of the soft response to saturation, inductors using this core material will be better suited to applications that require large, instantaneous currents. Ferrite cored inductors will saturate more quickly, but cost less and will have lower core losses.
Choosing the right value of inductance for your circuit is not a simple calculation, but most designs will work within a fairly wide range of inductance values.
Multilayer chip inductors are relatively new players in the inductor family. They’re available in very small physical sizes (0805) and allow for a very small overall design. The inductance values are currently available up to 4.7 µH, so they generally lend themselves to designs with higher switching frequencies.
The small size also limits the current-handling capacity, approximately 1.5 A, so they aren’t viable for higher-power designs. They’re smaller and offer lower DCR and cost than standard wire-wound inductors, so they may be right for your application.
Shielded Versus Unshielded Inductors
While shielded inductors are more expensive and have a lower saturation current (for the same physical size and value), they greatly reduce electromagnetic interference (EMI). It is almost always worth using the shielded inductors to help avoid any EMI issues with your design.
Diodes
Asynchronous switching power-supply designs employ a passive switch. The switch usually takes the form of a diode. However, because of the diode’s forward voltage drop, asynchronous designs are generally limited to less than 3-A output. Otherwise, the efficiency drop will be too great.
For all but the highest-voltage designs, Schottky diodes are the recommended choice for asynchronous regulators. They are available in breakdown voltages up to approximately 100 V. The lower forward voltage drop of Schottky diodes, compared to silicon diodes, reduces the power dissipation considerably. The effectively zero reverse-recovery time also prevents switching losses in the diode.
Schottky diodes are available with ultra-low forward voltage drop as well. These are only available in breakdown voltages up to approximately 40 V and will cost a bit more, but they will reduce power dissipation in the diode even further.
When selecting a diode, consider the forward voltage drop, breakdown voltage, average forward current, and maximum power dissipation. Choose a device with a forward drop as low as possible, but be sure to use numbers from the data sheet that reflect the forward voltage drop at the current that will be seen in the design.
Often, forward voltage drop will increase greatly with increasing forward current. A higher forward voltage drop will cause greater power dissipation in the device. This, in turn, will decrease converter efficiency and may overheat the diode.
Diodes have a negative forward voltage temperature coefficient, which will be a double-edged sword. As the temperature of the diode rises, the forward voltage drop will decrease, decreasing the power dissipated within the device. But because of this effect, the paralleling of diodes to share current is not recommended, as one diode will tend to dominate and hog all the current in a paralleled system.
The diode’s breakdown voltage should be rated above the voltages in the system. The forward current rating should be greater than the designed rms current for the circuit’s inductor. And of course, the diode needs to be able to dissipate enough power to avoid overheating. Choose a device with a maximum power dissipation specification larger than the design requires. ADIsimPower, Analog Devices’ online power design tool, has a large database of diodes and will strive to choose the best one for your application.
MOSFETs
The “switch” in switching power supplies is generally a MOSFET. Very high-voltage and high-current designs may use an IGBT-type transistor. MOSFETs come in two main varieties: N-channel and P-channel.
N-channel enhancement-mode devices require a positive gate-to-source voltage for turn on, have lower on resistance than P-channel (for the same size), and are less expensive. P-channel devices require a negative gate-to-source voltage for turn on, have higher on resistance, and are more expensive.
Because of the positive gate-to-source voltage requirement, N-channel devices tend to be more difficult to drive, as the gate may need to be driven above the main supply in the system. A simple bootstrap circuit usually handles this, but it adds cost and complexity to the system. P-channel devices, on the other hand, are much easier to drive, and no additional circuitry is required. The consequences for using P-channel MOSFETs are higher cost and higher on resistance.
In choosing a MOSFET, one has to be aware of some key performance parameters, such as RDS, VDS, VGS, CDSS, CGS, CGD, and PMax. MOSFET devices will be rated for maximum current and maximum power dissipation. These ratings must be adhered to. Internal power dissipation comes from two main sources: I^2*RDS and switching losses.
When the MOSFET (switch) is on, the only power dissipation comes from the I^2*RDS loss. When the switch is off, the device dissipates no power. But during transitions, the device will dissipate power. The dissipation during transitions is called switching loss.
Figure 10 shows how the switching loss manifests itself. It is mainly caused by capacitance on the gate, both gate-to-source and gate-to-drain capacitances. These must be charged and discharged to turn on and off the MOSFET. Figure 10 also illustrates the waveforms of the voltage and current.
During turn-on, there’s a period when there’s both voltage across the device and current flowing through the device. This will cause V*I dissipation within the device. Switching losses are greater at higher frequency. This is one of the many tradeoffs in SMPS design. Lower frequency means larger inductors and capacitors and better efficiency. Higher frequency means smaller inductors and smaller capacitors, but more losses.
Summary
When designing a SMPS, the supporting cast of components often takes a back seat to the choice of controller or regulator IC. But the choice of active and passive components will have a huge effect on overall power-supply performance. Efficiency, heat generated, physical size, output power, and cost will all rely in some way on the external components that are selected.
Careful analysis of the required performance is needed to make the best selections. The use of an integrated design tool, such as ADIsimPOWER from Analog Devices, will simplify this process.