Speaker Class D Amplifier Q&A

The use of Class D switching amplifier circuit can significantly improve the efficiency of the power amplifier. The Class D power amplifier converts the audio signal into a high-frequency pulse whose width varies with the signal amplitude, controls the power tube to saturate or cut off at the corresponding frequency, and the signal output by the power tube drives the speaker to sound through the low-pass filter. Because the power tube is in the saturated conduction and cut-off state most of the time, the power loss is very small, and its efficiency can reach more than 90%. A typical Class D power amplifier can provide 200W output, with an efficiency of 94% and harmonic distortion of 1%~2.8%.

Since its launch, Class D audio amplifiers have been widely concerned by designers for their high efficiency and small size. In recent years, with the reduction in price and audio quality comparable to that of Class AB audio amplifiers, they have gradually replaced Class AB audio amplifiers in some popular consumer electronics fields, including home theaters, DVD players, desktop audio, portable multimedia, etc., and have won a lot of room for growth.

The fidelity of Class D amplifiers is not as good as that of linear amplifiers, but they can meet the requirements in many occasions. For example, car audio systems only require distortion less than 2% at low power output and less than 5% at full power output. Moreover, the performance of Class D amplifiers will be improved after improvements. In addition, Class D amplifiers do not have crossover distortion.

The concept of Class D switching amplifier originated 50 years ago, but because its operating frequency should be at least 4 to 5 times the upper frequency limit of the audio signal (20kHz), the early circuits using electron tubes and transistors could not fully demonstrate their superiority in terms of power and efficiency. In the 1980s, MOSFETs that met the switching speed and conduction loss requirements appeared. In recent years, integrated pre-drive circuits have appeared, such as Harris's HIP4080, which has promoted the practical development of Class D amplifiers. The MOSFET used in Class D amplifiers is N-channel, because the conduction loss of N-channel MOSFET is only 1/3 of that of P-channel MOSFET of corresponding specifications.

The class D switching amplifier consists of an integrator, a duty cycle modulator, a switch drive circuit and an output filter. It uses a half-bridge driven class D power amplifier. It uses a fixed frequency duty cycle modulator. The square wave signal output by the power tube is mixed with the audio signal and sent to the integrator as a negative feedback signal. The integrator also has a filtering function, and the output correction signal is sent to the duty cycle modulator. The duty cycle modulator consists of a comparator and a triangular wave generator. The correction signal is used to modulate the triangular wave to generate a modulated output to drive the power tube to work. Negative feedback should be taken before the low-pass filter, otherwise the filtered signal and the input signal have a phase difference (the second-order filter may cause a 180° phase difference), which may cause the circuit to self-excite, and a complex phase compensation circuit is required.

The modulation signal driving the power tube is a square wave whose duty cycle changes with the audio input signal. The half-bridge drive circuit drives the two power tubes in opposite phases, one is turned on and the other is turned off. The square wave drive is used to make the MOSFET change its working state as much as possible, reduce the time it is in the linear amplification area, thereby reducing heat loss and improving efficiency. The efficiency of the circuit mainly depends on the switching loss and conduction loss of the power tube. The output filter converts the square wave into an amplified audio signal to drive the speaker to sound.

What is the state of soft distortion?

Soft distortion means that the output of the power amplifier begins to be suppressed before it is about to reach saturation, so that the harmonic content at saturation can be reduced. Class D power amplifiers also have a similar phenomenon. The oscillation you see when it is close to saturation is actually a parallel frequency phenomenon. The actual frequency is far more than 20KHz and will not emit oscillation noise. The Class D power amplifier solution is very efficient, and the influence of power supply voltage on efficiency is very limited. If conditions permit, it is recommended to increase the operating voltage to avoid saturated output of the power amplifier to improve the sound quality.

How to choose between Class D and Class G speaker amplifiers?

Class G amplifiers are the product between Class D and Class AB. They have not gotten rid of the Class AB working mode, but have optimized the power consumption from the power supply side. The power supply part is complicated, and the efficiency and damping coefficient are also much worse. From the indicators and actual performance, Class D amplifiers are much better, and Class G has no advantage.

What is the difference between Class D and Class AB speaker amplifiers?

Class D and Class AB amplifiers, like switching regulators and linear regulators, have completely different operating modes. Class D amplifiers have an efficiency of up to 94%.

What is the approximate efficiency of a Class D amplifier? How is power consumption calculated?

The loss of Class D amplifiers mainly comes from two aspects. On the one hand, it lies in its auxiliary circuits, internal audio signal processing and auxiliary circuits, etc. Another part of the power loss comes from the switching part. Regardless of the size of the amplifier, the auxiliary circuits are not very different, and the main difference is in the switching loss part. Generally speaking, the power consumption of the front-stage analog circuit is roughly around one watt, and the switching efficiency of the back-stage hardware is about 96%. The power consumption calculation of MOSFET mainly comes from switching loss and conduction loss.