DC-DC converters
DC-DC converters are circuits that convert one DC voltage to another. Some type of DC converter is needed in nearly all electronic devices, and there are DC-DC converters in all imaginable sizes and price ranges. They are often called voltage regulators since they regulate the output voltage by measuring it and adapting their function depending on how hard they are loaded.
Strictly speaking, a DC-DC converter doesn’t have to be a regulator, and many examples below show converters without regulation for the sake of simplicity (the wiring diagrams showing regulated converters are also simplified, however). Unregulated conversion is perhaps most common in unregulated AC-DC converters, which basically consist of a transformer (which converts 240VAC to, for example, 12VAC - approximately), a rectifier and capacitors to smooth the waveform. The actual voltage on such a converter can vary greatly depending on the power outlet.
Voltage regulators are usually divided into two categories: linear and switched converters (switched mode converter). In this guide, we will focus on the switching converters, but to illustrate the differences, it will be useful to first briefly describe linear converters.
Linear voltage regulators
One very common type of linear DC-DC converter are the classic voltage regulators in the 780x series, or adjustable regulators such as the LM317 and LM337 (to regulate positive and negative voltages, respectively). They typically take an input voltage of up to about 40V, and regulate it down to 5V for the 7805, 9V for the 7809, etc. All they require are one or two capacitors on the inputs and outputs to keep the regulator stable. The adjustable regulators work in a similar way, but also need a couple of resistors to set the output voltage.
This is all you need to regulate voltages from 7-35V to 5V. The EK023 kit can be found here.
Linear voltage regulators reduce the voltage by using a transistor (emitter follower) as a variable resistor. The voltage is regulated by varying the voltage on the transistor’s base. A transistor used in this way is usually called a pass transistor. The method is straightforward, but just like with unregulated voltage dividers with resistors, there is a cost: the resistance causes losses in the form of heat. The method is also limited to regulating voltages down, not up.
On the right: a schematic, simplified image of a linear DC-DC converter.
The great advantage of linear converters is their simplicity. For many applications, you basically just need to place the regulator, and you’ll get nice and smooth regulation without having to think too much. With switched regulators, it is never quite that simple. Even if you use a ”ready-made” regulator circuit, many more peripheral components are required (more about them further down), and the regulation loop is more sensitive to their particular values. There are web tools that can basically do all the circuit design work for you, but even when the component choices are made, other factors, such as their their placement, may determine whether the circuit works or not.
Fortunately, there are ready-made modules available for purchase, with the regulator IC and all necessary peripheral components pre-assembled on a board. When using such modules, all the work needed is making sure that the input voltage is within the correct range (with the correct polarity) and that the current draw isn’t too high – much like with the 780x or 3x7 regulators. Unlike them however, these modules often have extra features, such as an enable input, variable output voltages, selectale current limiting, and so on. But above all, they deliver voltages with significantly higher efficiency, often around 90% or more.
Some examples of switched DC-DC modules in Electrokit’s range.
Switched DC-DC converters
In switched DC-DC converters, the transistor is instead used as an electronic switch. The reason is that switches have hardly any resistive losses when they are closed, and when they are open, no current passes through them at all. Some losses occur, especially during the time that the transistor opens or closes (when it passes through the linear band), but these are very low, significantly lower than if they were constantly used linearly.
Instead of ”burning off” energy resistively, the switch is opened and closed to varying degrees to vary how much current passes, after which the resultant pulsed DC voltage is smoothed out with passive components. The method offers several advantages over linear converters. Better efficiency leads to lower heat generation, which results in lower weight and a much more compact circuit.
Another important advantage of switching converters is that they can both convert voltage downward and upward, depending on the configuration (topology). With linear converters, this is simply not possible.
Switched DC-DC converters can raise, lower or invert voltages, and provide stable regulated voltages with control over current draw, protection against overheating and more. They are found in everything from test equipment, mobile phone chargers, power banks, computers, televisions, audio equipment – it is almost easier to list things that don’t have DC-DC converters, than those that do. The size varies from tiny surface-mounted components of a few square millimeters and upwards.
Some common topologies
Switched DC-DC converters can be constructed in many different ways (called different topologies). You usually don’t have to think very much about exactly which topology you need, but in order to find the right converter it may be good to know what the different variants can do and what they are called.
Buck converter (step-down)
The buck converter converts voltage downwards, e.g.: +12V in, +5V out. The input voltage must always be higher than the desired output voltage. The type is also called step-down, and is both useful and common. This is the only type of switched converter with an equivalent linear DC-DC converter (albeit one with lower efficiency).
On the right a simplified buck converter, without regulation.
Boost converter (step-up)
The next topology does the opposite of the buck converter – it increases the voltage. For example, you can take +3V from two AA batteries and convert the voltage up to +5V to drive logic circuits. The input voltage must always be lower than the desired output voltage.
On the right a simplified boost converter, without regulation.
Step-up/step-down (buck-boost converter)
The buck-boost converter is a very useful DC-DC converter, which can either convert voltages up or down, e.g. +3.3 to +12V in, and +5V out. They are useful when the input voltage varies, and might lie both above and below the desired output voltage. One typical example is batteries, whose voltage drops when they are discharged.
On the right a simplified buck-boost converter, without regulation.
With a buck-boost converter, you may for instance operate a microcontroller that requires 3.3V with a lithium battery, even though its voltage drops from 4.2V (when fully charged) to 3.2V (discharged).
Three switched DC-DC converter modules. From left: buck, boost and buck-boost.
Concepts, specifications and parameters
Efficiency
As mentioned, the high efficiency represents one of the big advantages of switched DC-DC converters. The efficiency is usually stated as a percentage, e.g. 90%, where the figure indicates how much of the energy that goes in is usable – the remaining 10% are losses. The losses consist of, among other things, heat generation when the transistor turns on/off, plus the energy consumed by the oscillator and other control circuits.
As mentioned, a high efficiency provides benefits in addition to saving energy. One important aspect is that more efficient converters are both lighter and smaller. The difference is very large between switched and linear converters, but size will also differ between switched converters of different generations.
The efficiency of a regulator is rarely constant, but normally varies depending on the current draw and the difference in input and output voltage. As the power consumption of the control electronics is relatively constant, efficiency is usually lower at low current draws – their power consumption thus forming a larger part of the total current draw.
The diagram below shows how the efficiency varies for the TPS63070 buck-boost converter, for different current draws and input voltages, with a 5V output voltage. Note how low current draw and high input voltages yield lower efficiency, but also that large current draw with lower input voltages give decreasing efficiency.
Efficiency curve for the TPS63070. Source: Texas Instruments
Current draw
Simply indicates how much current a regulator can deliver. For many regulators, this is not just a simple value, but something that depends on external factors. A buck converter can, for example, deliver more current out than it draws in, e.g. 10V 2A in, and 5V 4A out. Here, a higher input voltage may mean that a device can source more output current.
The opposite applies to boost converters: the current out will always lower than the current in (but at a higher voltage). For these regulators, the maximum current is often specified as the one that the switches inside the regulator can handle.
For example, if the internal switches can handle 2A, and you’re switching up from 3V to 9V, the maximum theoretical current output (with zerp losses) will be 2/3 = 0.667A. With lower input voltage or higher output voltage, the current output will be even lower.
Input and output voltage and dropout voltage
Input and output voltages are just that, specified in volts. In cases where the output voltage is adjustable, a datasheet will usually also specify what input voltage is required for a certain output voltage; or for buck converters a minimum difference between input and output voltage – known as the dropout voltage. The dropout voltage works as with linear regulators, but is often lower.
Ripple
Ripple is noise — the leaking remnants of the switching action, and any other noise existing on top of the desired DC voltage. The lower this is, the better. The amount of ripple often depends on several factors, such as the difference between input and output voltage, load and switching frequency (see below).
Switching frequency
The switching frequency of DC-DC converter ICs varies. On many converters, it can be set externally, with e.g. external resistors and/or capacitors. The frequency will affect many things.
Higher frequencies allow the use of smaller (and lighter) passive components, not least the coil (or transformer). A higher frequency can also reduce the amount switching noise that reaches the audible range. This is of great importance in audio electronics, but even electronics that aren’t related to audio/media can emit disturbing noise in the form of a high-frequency whine from vibrating coils. Higher switching frequency also enables faster regulation when a load changes (step response).
Since a large part of the losses in switched-mode converters occur when the transistor switches between on and off, high switching frequencies also cause greater losses. Each switching operation involves some power loss, and with a high frequency this occurs more often. And although high frequencies can be easier to filter out of the audible range, they can also cause conducted and electromagnetic interference (EMI), with higher frequencies sometimes exacerbating the problem.
Components of a switched-mode DC-DC converter
A simplified switched-mode DC-DC converter requires only four components: a transistor, a diode, an inductor (coil), and a capacitor.
The transistor in a DC-DC converter is, as mentioned, the electronic switch – a transistor switch. Typically, the transistor is a MOSFET, but for higher voltages, IGBT transistors may be used. In simplified illustrations such as these, the switching transistor is usually shown with the symbol of a switch.
The inductor is used to store energy. Inductors always strive to have a constant current passing through them. When the switch opens and closes, either no current or a certain current (depending on the load) passes through the converter. The inductor smooths out that current.
The capacitor also stores energy, but in a different way: it tries to make the voltage across the capacitor as smooth as possible.
The diode blocks voltages in the wrong direction. It also prevents the circuit from being open when the transistor switch opens. Without it, the current through the inductor would be abruptly interrupted every time the switch opens, and it wouldn’t be able to do its job of storing energy. Coils (and the electronics they are connected to) don’t fare well when the current is interrupted abruptly – to maintain a constant current, a short spike of very high voltage is generated, often many kilovolts. The diode ensures that current can continue to flow through the coil. The diode is usually a type with low forward voltage (often Schottky diodes), and in this application they are often called freewheeling diodes (freewheeling diode).
These four components handle the actual DC-DC conversion, and depending on how they are connected together (different topologies – more on them below) you get different types of converters. Real-world DC-DC converters have significantly more components than these four, including for protecting the converter chip, protecting the device the converter is driving, to for reducing electromagnetic interference (EMI). Advanced converters may also limit the current in different ways, adapt the switching frequency to varying loads, vary the switching frequency continuously to reduce EMI (spread spectrum), limit large input currents in different ways, etc.
Regulation
In addition to the four components, an oscillator that opens and closes the transistor is also required, as well as electronics which measure the out voltage of the converter. The measuring electronics in turn control the oscillator, thus regulating the voltage. Regulation allows, among other things, DC-DC converters to generate a fixed voltage independent of variations in both load and input voltage. Without any regulation, the converter in the picture above would convert the voltage to a lower one, but to which voltage would be dependent on the load placed on it, and any variation would be significantly more difficult to both predict and control.
Schematic, simplified picture of a switching DC-DC converter with regulation.
Oscillator with PWM
The most common way to control DC-DC converters is by using pulse width modulation (PWM). This means that the oscillator that opens and closes the transistor switch has a fixed frequency, with different amounts of high/low pulses (i.e. different amounts of on/off in the transistor). The greater the amount of time the switch is closed, the more current flows through it (on average). The average is smoothed out by the inductor and capacitor as above.
Error amplifier
The voltage that comes out of the regulator is then measured in an error amplifier, where it is compared with a reference voltage, e.g. 1.25V. The difference between the desired and actual voltage becomes an error signal, which is used (in some way) to control the oscillator’s pulse width. The control gradually reduces the error, until it disappears completely (or at least, as near as possible).