As agreed in the first article of this series, this month we will discuss the power supply section of the power amplifier. Although relatively simple, it is largely responsible for the performance and reproduction quality of a hi-fi device. Before proceeding, it would be appropriate to define what we mean by the power supply block.
Sketches Credit: Normand Daigle
Any electronic device requires a direct current (DC) source to power its circuits, meaning the current always flows in the same direction, from + to -, according to accepted convention (the physical reality is the opposite, but it does not change the functionality). However, our electrical distribution network operates in alternating current (AC) to take advantage of the transformer effect, which facilitates transportation by minimizing losses. In this case, the polarity of the current fluctuates constantly at a frequency of 50 or 60 Hz (Hz = cycles per second of the sine wave generated by the power plant). The mains outlet to which we connect our devices provides a voltage ranging between 100 and 240 VAC (Volts in alternating current) depending on the country. The first thing to do to meet the demands of the electronic circuits is to reduce the AC voltage to a suitable level for our needs, then convert it to DC. The process is relatively simple, at least for conventional power supplies, but the noise and stability requirements of audio circuits make their design quality particularly important.
Before going further, it is important to define some terms and explain the operation of the components used in power supply circuits. The first item encountered from the mains connection will be the transformer. It essentially consists of two electrically isolated copper wire windings, identified as primary (connected to the mains) and secondary (used to create our DC supply). They are mounted on a metal core through which a magnetic flux generated by the current in the primary flows. This flux induces a voltage in the secondary proportional to the turns ratio of the windings. For example, a primary winding of 100 turns connected to 120 VAC mains would produce 12 VAC on a secondary winding of 10 turns.
The next step is to convert the AC to DC using a component called a diode, which only allows current to flow in one direction. By using four diodes configured in a bridge, we can transpose the negative part of the wave to the positive side, creating a pulsating wave with double frequency and fixed polarity. Left as is, this power supply would produce a 100 or 120 Hz parasitic signal in the loudspeakers. The solution is to add a high-value capacitor (usually several thousand uF) between the two poles to produce a filtering effect. To understand its operation, the capacitor can be seen as a small rechargeable battery that charges during the wave’s rise and partially discharges during the fall, and so on for each half-cycle. The resulting power supply will have a residual ripple (“hum“), the amplitude of which will vary with the current demand, but we are already much closer to a valid DC source.
The popular belief that associates the performance of a power amplifier with its weight is not entirely wrong, although things are changing with newer technologies like switching power supplies. However, let’s focus on more traditional systems for now. The transformer has limitations, and the high currents involved in power amplifiers require minimizing losses and increasing efficiency. Since the magnetic flux intensity is proportional to the current in the primary winding, the core must be designed to avoid saturation. Moreover, the windings have significant numbers of turns, resulting in a non-negligible resistance of the wire used. The solutions to circumvent these issues are to increase the core size and use a larger gauge wire, both choices that will quickly increase the weight and manufacturing cost. In fact, the transformer accounts for at least 50 % of the weight of a well-designed power amplifier.
A quality power supply is crucial for good performance, but the requirements can vary greatly depending on the type of device. For example, in the case of a phono preamplifier, the current demand is minimal, but great voltage stability and total absence of noise are priorities. To improve the performance of the basic power supply circuit we saw earlier, a regulation system will stabilize the voltage and eliminate the residual ripple. Most devices use this type of regulated power supply, which is easily achievable given the low currents involved, on the order of milliamps (mA). However, power amplifiers are a different case, and a different approach is necessary. It will involve higher voltages but, more importantly, very high currents on the order of several amps (A), but these points do not necessarily exclude regulation.
Regulation ensures a very stable voltage from a power supply that may fluctuate due to mains variations or variable current demands. It is achieved through a combination of active and passive components. For a low-power, less demanding circuit, a simple Zener diode may suffice. This component provides a constant voltage at its terminals but requires a series resistor, making high-power use very inefficient. However, a Zener diode can be used as a reference in conjunction with a transistor to increase efficiency and performance. In such cases, the current capacity of the circuit will be limited by the transistor’s characteristics, not the diode’s. For higher performance, an integrated regulator offers a simple and effective solution, as it is a single component containing a complex circuit (IC). However, their use is limited to low and medium power circuits, as most have a current limit of less than 2 A. For higher current needs, a hybrid circuit combining an integrated regulator and a power transistor can do the job. Note that variants using an IC can also be made with discrete components. The performance improvement provided by regulation does come with constraints, namely increased complexity and thermal dissipation. However, this is a small price to pay for quality performance.
We have only discussed simple power supplies so far, which have only two poles, a + and a – which is usually also the ground (ground). This has some disadvantages. A dual or symmetrical power supply solves several problems without significantly increasing circuit complexity. In this case, the ground is the central point, with a positive terminal (identified as V+) and a negative one (V-). One major advantage of such a configuration is the reduction of residual power supply noise, as the ripple of each pole is out of phase with the other, causing partial cancellation in the circuit it powers. Another important point is the possibility of coupling a circuit directly to its load (the speaker in our case) without using coupling components (capacitor or transformer). This aspect may not be immediately obvious but should soon become clear.
What we have seen about regulated power supplies applies to what are known as linear regulators. Although they are highly effective, inexpensive, and relatively simple, they are relatively inefficient. The stability of the output voltage comes at the cost of some energy consumption, resulting from the current passing through the regulator multiplied by the voltage difference between its input and output. For example, a 15 V power supply regulated to 10 V and connected to a 10 Ω load will cause a 5 W dissipation in the regulator, i.e., a current of 1 A (10 V / 10 Ω) multiplied by a 5 V voltage drop (15 V – 10 V), requiring a heat dissipation device to prevent component damage from heat.
An increasingly used alternative for powering electronic circuits, especially interesting for power amplifiers, is switching power supplies. The advent of very fast, high-voltage transistors and the development of the computer industry have enabled the production of highly efficient switching power supplies at reasonable prices, with the added advantage of being universal, accommodating a mains voltage between 90 and 250 VAC. Their very high efficiency, sometimes exceeding 95 %, and their low weight and size make them very attractive for high-power applications. However, they also have drawbacks, the main ones being critical design and the tendency to generate radio frequency interference (RFI), which is particularly difficult to control. However, if well designed, their performance is very impressive, and they represent the unavoidable solution for the future.
To simplify understanding of the circuits, simple power supplies were mentioned here, but in practice, they are almost no longer used because symmetrical versions offer superior performance. But there is also another advantage we mentioned earlier, the possibility of coupling the load directly to the amplification circuit, an important factor in a power amplifier due to the low impedance of the load (the speaker). It is important to note that the speaker should only receive the audio signal, which is an assembly of alternating waves, thus AC. It must always be isolated from the power supply. In a simple 40 V DC power supply circuit, for example, the connection point with the load will be about 20 V DC to allow the audio signal to modulate across the 0 V DC to 40 V DC range. The solution will be to use a capacitor or transformer. These components, both expensive and undesirable, significantly impact performance, and it is preferable not to introduce them in a circuit where optimal performance is sought. With a symmetrical power supply of + 20 V DC and – 20 V DC, the central point corresponds to the ground, so 0 V DC, and the load can be connected without issue.
Finally, a quick overview of tube circuits is warranted given the renewed interest in this technology. In this case, even though a symmetrical power supply is theoretically possible, it would only add complexity without offering any advantage. The reason is that tubes have too high an output impedance to directly drive a load of 8 Ω or less like a speaker. The only solution is to use a transformer, which among other functions, will perform this impedance coupling.
Next month, we will begin analyzing each class of power amplifiers. We will specify the requirements, strengths, and weaknesses of each, and try to link their mode of operation to perceived performance.
To read article one, CLICK HERE
To be continued in September…
Subscribe to our newsletter; info@quebecaudio.com
