Class D amplifier modulates the audio signal with reference triangular wave or sawtooth wave, and generates an amplified signal usually in the form of pulse width modulation (PWM) switch output. A typical switching frequency is greater than 10 times the highest frequency of interest in the input signal. In most class D amplifiers, a feedback path with error signal is also used to improve total harmonic distortion and noise (THD + n) performance, power rejection ratio (PSRR) and other performance characteristics.
The limitation of class D amplifier in practical application is obvious. High frequency energy appears in the switching frequency and its harmonic and square wave spectrum components. Until recently, class D amplifiers needed a low-pass filter (usually a 2-pole Butterworth LC filter) to filter out high-frequency square waves with high current and leave only audio signals. In the new class D amplifier, a filter free implementation method uses the loudspeaker itself as the low-pass filter element. These newer “filter free” class D amplifiers have become very popular in portable designs. Unfortunately, because the electromagnetic interference intensity generated by this method exceeds the requirements allowed by the traditional class D amplifier with filter, they may not be allowed to be used.
The trend of portable design intensifies the problem of electromagnetic interference. As products become smaller and smaller and components, leads and wires are closer and closer, proper PCB layout becomes more and more difficult. Due to the limitation of area, it is almost impossible to use the filter again. The current development trend requires higher and higher volume of devices, so higher power and current are needed to produce greater radiation. In addition, the integration of multiple wireless communication functions in a single platform, such as Bluetooth, Wi Fi, wireless local area network, etc., makes electromagnetic interference a prominent problem. Electromagnetic interference is not only a problem that needs to be considered directly inside the product, but also RF radiation interference with other external systems. Most consumer systems need to pass some FCC tests, which are mainly aimed at the interference of unintentional radiation of products to other RF spectrum devices.
Many methods can be used to reduce electromagnetic interference, such as slowing down the edge of square wave, but this will weaken the ability to accurately sample the incoming analog audio signal and reduce the efficiency. Therefore, it is at the cost of increasing thd + n. Using LC filter can greatly reduce electromagnetic interference, but LC filter is very large and expensive, and its size and cost increase with the increase of output power. Once the length of the lead reaches a quarter of the wavelength of the transmitted signal, the PCB lead and wire will essentially produce antenna effect and generate a lot of radiation. Therefore, the length of the lead is usually as short as possible.
Other methods include passing PCB leads transmitting high-frequency signals between ground planes, and using insulated elements and ring inductors. For class D systems without filters, the lead and cable length connecting the amplifier output to the speaker is likely to be the largest RF radiation source. For example, traditional methods such as placing a ferrite bead in series with the loudspeaker near the amplifier can effectively reduce the radiation. Ferrite magnetic beads can act as RF chokes and attenuate high-frequency signal components. However, ferrite beads are only effective in a narrow frequency range, and it may be difficult to provide sufficient attenuation in the whole bandwidth of output noise. If PCB layout and filter can not reduce electromagnetic interference to an acceptable level, shielding measures can be adopted. Power supply is another possible source of electromagnetic interference. Class D amplifiers draw current in the form of large amplitude narrow pulses associated with the edge of the output switch on the power line. Electromagnetic interference related to power supply can be reduced by proper layout and bypass technology.
Although the method of “afterwards” reducing electromagnetic interference is effective, the best method is to make the amplifier produce less interference at the beginning. Compared with the previous class D topology, spread spectrum devices provide this possibility. Spread spectrum technology is not developed recently. It has been used for more than half a century. It was first used in the fields of communication system and military radar. In the past decade, spread spectrum modulation technology has been widely used in other fields, especially in clock circuits. When it is used in class D amplifiers, spread spectrum technology will bring similar advantages.
The spread spectrum modulator can adjust the switching frequency of the output bridge in a frequency band near the central switching frequency (for example, frequency expansion with ± 30% near the 300kHz central frequency). As long as the frequency change is always random, various methods from simple scanning to uncorrelated hopping of carrier frequency can be used as practical frequency conversion methods. Spread spectrum modulation scheme has some key advantages: while maintaining high efficiency and low THD + N, it can reduce radiated noise and electromagnetic interference without reducing the total energy. The peak energy after spread spectrum is reduced, but the total energy remains unchanged, but is distributed in a wider frequency band. The bandwidth of the noise becomes larger, but the noise peak at any frequency point is smaller than that produced by the South fixed frequency device.
Spread spectrum modulation technology has two main advantages: the lower radiated noise peak brings the improvement of electromagnetic interference performance, and can reduce or even eliminate the need for the common electromagnetic interference filter in class D applications. As shown in the figure on page 61, a class D audio amplifier with spread spectrum technology is introduced.
Class D amplifier radiation standard in FCC and CE standards, which are applicable to any digital consumer device not used for transmission. All consumer electronic products must pass these two certifications before they are listed in the United States and Europe. The preliminary EMI test of the product with 2-inch speaker cable and no filter element shows that the product has excellent EMI performance during the FCCB limit test.
Spread spectrum modulation technology brings significant advantages to the application of class D audio amplifier. It reduces RF radiation, simplifies high-cost strategies to reduce electromagnetic interference such as using LC filter, and greatly reduces the obstacles faced by the traditional class D topology in the field of portable design. Relevant applications that can benefit from this include any portable device that needs to comply with FCC / EC rules or other EMI related rules such as mil-std461. In addition, any portable device that needs to reduce system noise can benefit from it.
Electromagnetic interference is an important issue to be considered in system level design. System designers need to start with the unit modules and components used in the design, use all available tools to create a high-performance product, and use devices with spread spectrum modulation characteristics can effectively reduce the electromagnetic interference in portable system design.
Responsible editor: GT