If your system doesn’t need to be portable, this is the most obvious choice. What comes down the “pipe” is AC and is far too high a voltage to be of immediate use to a digital system. It must be converted to a DC voltage of significantly lower magnitude. There are plenty of solutions for doing this. You can use DC “lab” power supplies, standard PC supplies (probably overkill for your needs), or AC adaptors . The last of these is probably the best choice for most applications.

AC adaptors (also known as plug packs or sometimes power bricks ) are the little black boxes that come with your cell phone and a host of other appliances. They are a cheap, easy, and reliable solution and can be purchased from any good electronics vendor. Typically, they will provide an output voltage somewhere in the range of +5 VDC to +12 VDC and can supply a current of up to a few Amps, depending on the particular plug pack. Choose one that will supply an appropriate voltage and current for your system. One caveat with plug packs is the polarity of the connector. Most plug packs have the positive voltage on the center of the connector jack and ground on the outside. Other plug packs have the exact opposite arrangement! Not knowing which you have could lead to disastrous consequences for your embedded system. As always, check the technical data and, if in doubt, use a multimeter to check the output polarity before connecting it to your system.

A better way is to incorporate a bridge rectifier as part of your design (Figure 5-1). That way, the polarity of the power source makes no difference. The input power is DC, but the polarity of the connection makes no difference. The embedded system uses the output of the rectifier as its power source and has internal voltage regulation. (We’ll discuss regulators shortly.)

Figure 5-1. A bridge rectifier makes an embedded system “polarity-proof”

Batteries are easy to use. The only caveat is to ensure that the battery (or batteries ) you have chosen can supply enough current at the right voltage. With the right choice of battery and a carefully designed system, you can achieve extended operation over very long periods. For example, a small PIC- or AVR-based computer can (depending on application and design) operate for up to two years off a single AA battery. A poorly designed system can drain a battery in minutes. A poorly chosen battery unable to supply sufficient current will result in erratic operation, or may result in the system being unable to start at all. When choosing a battery, consider not just its average current capability but also its peak current. An embedded computer may need only an average supply of 20 mA but may require as much as 100 mA at peak loads. This is especially true of systems using flash memory, which may require high currents during write operations. The battery for such a system must be able to supply not just the continuous load, but also the peak load when required.

There are several ways you can reduce the power consumption in your embedded system. The use of low-power devices is the most obvious place to start. The power consumption of different chips varies considerably, and there are many low-power variants of common devices available. RISC processors often have lower power consumption than comparable CISC processors, so they are often used in preference to CISC in low-power applications. The PIC and AVR microcontrollers can have current draws of less than 5 mA (and as low as 10 nA when in sleep mode). This is considerably less than the 35 mA used by a 68HC11 microcontroller.

Many memory chips and peripherals will automatically enter a low-power mode when they are not in use. Others may be placed in low-power mode by toggling a digital input or by an appropriate software command. The power consumption of some devices can be reduced even further by turning them off when not in use. If the processor is executing code from RAM and outputting data to a serial port, then the power to the ROMs and any other I/O devices may be turned off since they are not in use. Implementing this requires separate power sources for the chips that are to be disabled, switched via software control. Some voltage regulators (discussed in the next section) have shutdown inputs, allowing the subsystem they are powering to be turned off.

Further, some low-power devices (such as sensors) may need very little current, so little that they can be directly powered from the I/O line of a microcontroller. The I/O signal is the power supply for the device. The devices can be turned on or off under software control by toggling the I/O line. Some processors, such as the PIC and AVR, can sink relatively large currents (20 mA) through their I/O pins, and these can be used as ground for some devices (such as LEDs).

A voltage regulator is a semiconductor device that converts an input DC voltage (usually a range of input voltages) to a fixed output DC voltage. They are used to provide a constant supply voltage within a system.

While many components in an embedded system can operate from a wide power-supply range, a fixed operating voltage is necessary for devices such as Analog-Digital Converters (ADCs), since many use the internal power supply as a reference. In other words, the output voltage of a sensor is sampled as a percentage of the voltage supply of the ADC. If the supply is not a known voltage, then any sampling performed by the ADC is meaningless. (We’ll look at ADCs and sampling in Chapter 13.) Therefore, we need a voltage regulator to provide a constant voltage source, and thereby a constant voltage reference. Further, a voltage regulator can assist in removing power-supply noise and can provide a degree of protection and isolation for the embedded system from the external power supply. If your system is operating from a battery, the varying current draw of your computer can combine with the battery’s internal resistance to create a varying supply voltage. The addition of a voltage regulator prevents this from becoming a problem since it provides a constant output. Including a voltage regulator in your design is good practice.

The types of regulators we will look at are termed DC-DC converters . They take an unregulated DC voltage (often over a range of possible voltages) and provide a constant DC voltage output of a fixed value.

There are three types of DC-DC converters: linear regulators , which produce lower voltages than the supply voltage; switching regulators that can step up (boost), step down (buck), or invert the input voltage; and charge pumps , which can also step up, step down, or invert the supply voltage, but with limited current-drive capability. (Note that not all charge pumps provide regulated voltage.)

The conversion process of any regulator is not 100% efficient. The regulator itself uses current (known as >quiescent current ), and this is sourced from the input supply. The greater the quiescent current, the more power (and therefore heat) the regulator must dissipate. In choosing a regulator, select one that can supply the appropriate output voltage and the required current needed by your embedded system, yet has the lowest quiescent current.

Linear regulators are small, cheap, low-noise, and very easy to use. The basic circuit for a linear regulator is shown in Figure 5-2. The inputs and outputs are filtered using decoupling capacitors, but beyond that, no other external components are needed.

Figure 5-2. Example of a linear regulator circuit

As well as helping to smooth the voltages, the capacitors also help remove momentary glitches in the power source known as brownouts . These momentary drops in power are infrequent, but when they occur, they can severely corrupt a computer’s operation.

Switching regulators get their name because they switch a power transistor (MOSFET) at their output. They tend to be more efficient than linear regulators in converting the input voltage to the output voltage. In other words, they waste less power during the conversion process. However, their drawback is that they require more external components (such as an inductor and diode) and therefore take up more space. They also typically cost more and generate far more noise than linear regulators. Unlike linear regulators, they can step up a voltage as well as stepping one down, and they can also invert. For example, a switching regulator can take a supply voltage of 3.6 V from a battery and provide you with a regulated 5 V supply for your embedded system. Alternatively, a switching regulator may take an unregulated 8 V supply and convert this to a regulated -12 V supply. Switching regulators are far more versatile than linear regulators. However, they do require careful design and board layout, so pay careful attention to the directions of the particular component manufacturer. As always, read the datasheets carefully.

Charge pumps, like switching regulators, can step up, step down, or invert voltages. Unlike switching regulators, they require no external inductor. However, due to their limited capacity to supply current, they are not commonly used. The MAX3222 (and similar devices), discussed in Chapter 9, use internal charge pumps to generate the +12 V and -12 V required for RS-232C level shifting.

There are literally thousands of voltage regulators available. Probably the most commonly used are the LM78xx linear regulator series made by several manufacturers such as Fairchild (http://www.fairchildsemi.com), Semelab (http://www.semelab.co.uk), and ST Microelectronics (http://www.st.com). They typically come in a TO-220 package (Figure 5-3) and have a metallic attachment point for a heatsink, such as the one shown in Figure 5-4. The regulator is normally mounted flat against the circuit board, and the pins are bent 90 degrees downward.

The part number designates the output voltage. For example, an LM7805 provides a 5 V regulated output, while an LM7812 gives a regulated 12 V output. They can provide an output current of up to 1 Amp (and as much as 2.2 Amps peak), with a quiescent current of between 5 mA and 8 mA. They also feature overload and short-circuit protection. Table 5-1 lists the regulators, their input voltage ranges and their output voltages.

Figure 5-3. LM78xx

Figure 5-4. Heatsink

The LM78xx is simple to use. Decoupling capacitors (nominally between 10 uF and 47 uF) are required on the input (pin 1) and output (pin 3), as shown in Figure 5-5. Pin 2 is connected to ground.

Figure 5-5. LM78xx circuit

For negative output voltages, use an LM79xx regulator. It’s used in exactly the same way as an LM78xx.

A good small regulator is the Maxim MAX603 (5 V output) or MAX604 (3.3 V output). I tend to use these as the default “workhorse” regulators for many of my small designs. They use far less quiescent current than LM78xx regulators and as such are ideal for low-power or battery-operated systems. They are available in tiny surface-mount SO-8 or in standard DIP packages (discussed in Chapter 6) ande require only two external components.

The MAX603/604 can provide up to 500 mA of current and can operate from an input voltage of between +2.7 V and +11.5 V DC. They have built-in protection in case you inadvertently switch power and ground and consume as little as 15 μA of current for their own use. As such, they are ideal for use in low-power, embedded computers. The schematic for using a voltage regulator such as a MAX603 or a MAX604 is shown in Figure 5-6. In this case, the input supply is a battery, but it could just as easily be a DC plug pack or some other sort of supply.

Figure 5-6. MAX603/MAX604 circuit

A capacitor is required at the output (pin 8 of the regulator) to filter the voltage. This capacitor forms part of the regulation circuit and is required. The device will not work without it. The second support component required is a capacitor on the input (pin 1). This stabilizes the input voltage and can provide an additional current source during peak loads. These capacitors are known as decoupling capacitors and are discussed in detail later in this chapter.

The ground pins (2, 3, 5, 6, and 7) should all be connected to ground, as they act as a small heatsink for the device. When laying out the PCB, place a ground fill under the regulator and connect these five pins directly to it. Pin 4 () places the regulator in shutdown mode. For constant operation, this pin is connected directly to the power source, such that the device is always on. If you’re using this regulator to power a subsystem within your embedded computer, you can use to act as a software-controlled power switch by connecting it to an I/O line of the processor.

A MAX604 gives a regulated 3.3 V output. For a 5 V output, replace the MAX604 with a MAX603. The circuit is otherwise the same.

The MAX1615 is a linear regulator that can operate from a supply range between 4 V and 28 V. It is tiny and is capable of supplying 30 mA of output current at either 3.3 V or 5 V. 30 mA is not much current, but for very small (battery-powered) applications, it may be sufficient. The MAX1615 has a shutdown input, allowing an external system to power it down. One use for this regulator could be as a power source to subsystems within an embedded computer, allowing the host processor to turn them off when not in use. Before turning off a subsystem, ensure that its “absence” won’t adversely affect the functionality of the rest of the system.

The basic schematic for a MAX1615 circuit is shown in Figure 5-7.

Figure 5-7. MAX1615 circuit

The 5/ pin determines the output voltage. By tying it to ground, the output is set to 3.3 V. Connect the pin to the input, and the output voltage is set to 5 V. For continuous operation, tie the shutdown pin () to the input voltage. To allow the regulator to be powered down, connect to a processor I/O line or a simple power switch. (It goes without saying that if you’re driving with an I/O line of the processor, then that processor must have a power source other than this regulator! Otherwise, it could lead to some interesting situations.)

For higher-current situations, the MAX724 can supply up to 5 A from an input supply voltage between 8 V and 40 V. Its output is adjustable from 2.5 V to 35 V, and its quiescent current is 8.5 mA. It comes in a 5-pin TO-220 package, with an attachment point for an external heatsink. The basic circuit for a MAX724 is shown in Figure 5-8. Note that it is a step-down regulator only.

Figure 5-8. MAX724 circuit

The inductor is nominally 50 uH, but can be any value in the range of 5 uH to 200 uH. When the output pin V_SW turns off, the diode provides a path to ground for the inductor current. The inductor chosen should have a high saturation current. If it doesn’t, the effects will be disastrous. The Maxim datasheet provides detailed information on selecting an appropriate inductor.

It is recommended by Maxim that a Schottky diode, such as an MBR745, be used due to the fast switching times required. Both the regulator’s input and output require large decoupling capacitors to filter out ripple. The capacitors must have low ESR (Equivalent Series Resistance) over the expected temperature range and operating lifetime.

The output voltage is set by resistors R1 and R2. The equation for calculating the output voltage is:

VOUT = 2.21 * (R1 + R2) / R2

R2 should be less than 4 kΩ. Maxim recommends choosing 2.21 kΩ for R2, as it simplifies the above equation:

VOUT = (R1 + 2.21k) / 1000

So, the equation to calculate R1 becomes:

R1 = 1000 * VOUT - 2.21k

The regulators I’ve covered so far should be useful in most situations. If you’re designing a machine with special requirements, there is bound to be a regulator that will suit your needs. Just spend some time browsing component manufacturers’ web sites and see what is offered.

Digital systems are inherently analog in operation. Digital signals suffer degradation and noise due to analog effects present in the system. Spurious noise or reflections from nearby electrical machinery or radio transmissions can induce signals within your circuit that can cause false events to occur, or even prevent a digital system from functioning at all. The one way to ensure that your system is immune to electromagnetic interference is to avoid the use of electricity! Unfortunately, the steam-powered microprocessor is not a reality, so if your system is to operate reliably in the real world, you must take electromagnetic effects into account. What follows is not a comprehensive overview of noise and associated problems and solutions, which is far too complex a field to cover properly here. What I will do is provide an introduction. It is worth seeking out information that is more detailed to understand these concepts more thoroughly.

Noise can be a significant problem in digital systems. Noise can disturb signal transmission, leading to corrupted data, or may even cause a program to crash. Problems with an embedded system may or may not be noise-related. Similar problems can also be caused by inadequate power-supply levels, insufficient decoupling capacitors, marginal timing tolerances, and software bugs. However, even a well-conceived system can be disrupted by noise. There may be noise problems inherent in the design. Switching noise from integrated circuits, ringing, and crosstalk (discussed shortly) are all due to aspects of the designed system. Other forms of noise may be due to environmental effects (such as nearby motors or radio emissions). The bad news is that electromagnetic problems are getting worse for digital systems. The environment is bathed in ever-increasing emissions from radio, TV, and cell-phone towers. At the same time, integrated circuits are becoming increasingly sensitive, as designs move towards higher-speed and lower-power operation.

These sources of noise may not be present during the design and test phase of the system but will only manifest themselves once the system is out “in the field” and in service. A crashed embedded system may be due to a hardware problem, a software problem, or the fact that the factory a block away turned on a compressor, which caused a spike on the power supply. Field problems created by noise may only occur very occasionally and can often be very difficult to track down. It is not unusual for some problems to occur only once every few days. Any problem is not satisfactory and must be fixed, but identifying the cause is not always easy. It is better to design the system from the beginning to be as immune to these problems as possible. You have to consider not just what emissions your system may produce, but also how it may be susceptible to external effects. This will not guarantee your system will be problem-free, but every bit of immunity helps.

Electromagnetic interference (EMI) is noise generated by sources external to the embedded system. Some examples of EMI are motors, switches in power consumption, fluorescent lighting, RF emissions, and electrostatic discharges. All can be significant sources of noise. For example, turning on or off a machine with an electric motor can cause a 1000 V spike on the AC power-supply line. An electrostatic discharge (ESD) can send a spike of 35 kV from a finger into an integrated circuit with a current rise time of 4 Amps per second! This can be enough to permanently damage a sensitive chip. Cars are particularly noisy environments. The 12 V supply line to the automotive electronics may be reversed, driven at voltages ranging from 6 V to as much as 24 V, and have ± 400 V transient spikes. All of these can have very adverse effects on the operation of an embedded system.

In any circuit, there is a wire carrying current in and a wire carrying current out. Current flowing through a wire generates a magnetic field around that wire. Such a magnetic field can be a source of EMI. The intensity of the magnetic field felt is inversely proportional to the distance from the source of the field, and the orientation of the field relates directly to the direction of current flow in the wire.