Power supplies
An article introducing the basics of power supplies: the first step toward a good amplifier.
All power amplifiers, A/V receivers and integrated amplifiers need a power supply, an electronic circuit whose job is to convert the alternating current that reaches us through our wall sockets into direct current. But each device requires different current delivery (power) values, so every power supply is different. In the case of power amplifiers, moreover, the power supply is a critical and important point, since it determines how well the amplifier fulfills its purpose: to deliver a current that is not only capable of driving a loudspeaker, but of doing so with the least possible distortion.
Although, as we will see, converting alternating current into direct current can be simple, unfortunately there are always significant variations in the electricity supply, such as voltage drops and transient spikes. In fact, the best way to obtain reliable direct current is to use batteries or energy accumulators, although they discharge over time and this is a tremendously expensive solution. The most common approach is to use circuits that convert the current available at our sockets which, as we know, is alternating, with a frequency of 50 Hz (60 Hz in most of the Americas).
Broadly speaking, a power supply consists of five distinct stages: transformation, rectification, filtering, regulation and voltage division.
The transformation stage is responsible for changing the input voltage (in our case, 220 V) to the value required by the component's circuits. In other words, if necessary, those 220 V AC are changed into new values of, for example, 15 V. As many different output voltages as desired can be obtained, which is common, since there are several electronic circuits, each with its own supply.
THE TRANSFORMER
The best-known element of a power supply is the transformer, precisely because of its heavy, bulky size. Most of the electronic systems we use work with voltages of up to 24 V, so we use the familiar transformers to "step down" those 220 V to the desired 24 V. In our "domestic" environment we have single-phase power, so single-phase transformers are used. They are based on the principle that electrical energy can be transferred efficiently by magnetic induction from one coil to another, through a varying magnetic flux, as long as the coils are located on the same magnetic circuit. This is the image we have of a transformer. It uses a number of laminations of ferromagnetic material forming a loop (as if they were rings).
A coil is connected to each side. At the terminals of each coil we obtain a different voltage. Depending on the number of "turns" of the coils, we obtain one voltage value or another on the secondary. Since the transfer takes place magnetically, we also manage to isolate one current from the other. This electrical isolation usually eliminates some of the interference inherent to the primary current (the one at the socket in our home). A transformer with a 1:1 ratio indicates that the primary and secondary windings are identical, so we have the same voltage at both sets of terminals. But we can eliminate some annoying interference (noise). In this case, we have used a transformer as a power filter. This transformer only works with alternating current. If we apply direct current to the primary winding, we will never get any voltage on the secondary. The three most important parameters of a transformer are the secondary voltage (expressed in effective volts, V RMS), its power rating (in VA or W) and its regulation factor.
On paper, the power in both windings should be exactly the same. If power is the product of voltage and current, reducing one increases the other. In practice, because of losses in the magnetic circuit itself (due to hysteresis and Focault currents) and in the windings (due to the Joule effect), the power figures differ noticeably. The ratio between the power of the secondary and that of the primary is known as efficiency. The lower the result, the better the transformer.
One aspect we must bear in mind is that during the rectification stage there will be a notable voltage loss (a diode bridge has a drop of about 600 mV), so it is normal to look for a secondary output voltage somewhat higher than the one desired. If we want several different voltages from a single transformer, all we need to do is include as many "secondary" windings as we wish or, as you can read at the end of this article, use a voltage divider. However, the sum of the currents of each circuit in the secondary section will never be greater than (on paper, equal to) the current in the primary circuit.
A brief aside on a regular feature of high-end components: the toroidal transformer. In terms of operating principle, it is no different from any other type of transformer. The difference lies in the core, which gives it its characteristic shape. With this design, tolerance and efficiency values are better than with the traditional ring-shaped design.
Curiously, they weigh less and are smaller than conventional designs, although they are usually used for high current delivery, so in the end we think of them as heavy by nature (but there is one exception: noise, which we explain below). One of their best-known defects is that their laminations tend to vibrate, producing a characteristic sound that is unwelcome in high fidelity. Reducing these vibrations means higher manufacturing costs and greater weight.
RECTIFICATION
The next step is to convert AC into DC. The most common method is to connect a rectifier diode in series with the load. A diode is an electronic component that only lets currents of positive potential pass through it. If we think of the AC signal as a sine wave, wherever we connect a diode we will only obtain the waveforms that appear during the positive half-cycle (that is, half of them). A standardized design is the single-phase full-wave bridge, or Graetz bridge.
The signal obtained after this bridge can be considered direct current, but it is also pulsating. In fact, what we have now achieved is to "flip" the negative half-cycles as if they were positive, "drawing" a signal that varies over time but always stays in the positive zone. This current needs to be filtered to obtain a value that is as linear as possible.
FILTERING
To convert that rippling signal into something as close to direct current as possible, a basic design uses a capacitor in parallel. A capacitor is an electronic component capable of storing a certain amount of energy so that we can draw on it whenever we want. The aim is to store voltage and release it just when the voltage coming out of the rectifier bridge falls toward 0 V. The lower the voltage at the capacitor's input, the more of its stored energy it must deliver. And so on, throughout the entire sequence.
Here is another aspect that differentiates power supplies. It is very difficult to obtain a completely linear direct voltage, which means that achieving maximum perfection requires more precise and complex, and therefore more expensive, electronic circuits. The variations in the resulting waveform after filtering (measured by its effectiveness) may not be very critical in some applications, but they are significant in the case of power amplifiers and amplifiers.
Another option is to use capacitors with greater capacitance and delivery to make up for small deficiencies in the primary current supply. In the event of a voltage drop, the (potential) energy stored as a reserve in one of the electrolytic capacitors should provide extra delivery that prevents the momentary lack of voltage from being passed on to the output stage.
REGULATION
To achieve our goal, regulation comes into play, a stage that is not always present in some power supplies. This process aims to turn that slightly variable input direct voltage, fluctuating between two extreme values, into an almost pure output direct voltage. There are a great many alternatives for this stage, all the more so with the proliferation of transistors, Zener diodes and integrated circuits.
In the latter case, 78xx and 79xx ICs are commonly used in basic designs that, at moderate prices, regulate the voltage efficiently. A combination of the two allows voltages of +/- 12 V without the complicated or expensive use of numerous electronic components. Some designers still turn to discrete transistors for even more precise regulation, but this is expensive, owing to the use of electronic components that must be soldered and designed to perfection. Once again we come back to the usual maxim: the best regulation always comes with a higher price.
Also at this stage, the system needs to be protected against unexpected current spikes, not only to prevent the regulation circuit from being damaged, but so that the final direct current is as stable as possible.
VOLTAGE DIVISION
In this last process, often not at all critical, we divide the final direct voltage into as many values and supplies as we need, since not all the circuits that make up an amplifier or power amplifier necessarily use the same voltage.



