Class D Audio Amplifier IC FAQ

Part 1 of this FAQ provides an overview of Class D amplifiers and answers questions on how to select an amplifier and design filters for Class D amplifiers.

What is a Class D amplifier?

Class D amplifiers use pulse-width modulation circuits to keep their output transistors operating in either a fully on or fully off state. In other words, at any given time, the instantaneous output voltage is either one supply voltage or the other, ignoring the brief transition period when switching. As a result, the output current is conducted from the device without a noticeable voltage drop.

Ohm's law states that power equals voltage times current. Class D amplifiers keep the voltage portion of this equation close to zero, thereby minimizing power consumption in the output stage. Class D amplifiers have advantages over other technologies, with typical efficiencies of up to 95% and average efficiencies in the 80% range. Class D amplifiers can switch at frequencies above the audio band. Most Class D amplifiers switch at frequencies between 300K Hz and 2M Hz.

Why Use Class D Amplifiers

Because Class D amplifiers are very efficient, they make full use of the limited power from batteries and other power-limited sources. In addition, this higher efficiency eliminates the heat sinking requirements of many amplifiers below 10 watts of output power. Class D amplifiers do not cause heat dissipation to other nearby components and other topologies, thereby reducing the temperature of the environment. In addition, the thermal efficiency of Class D amplifiers allows them to use standard IC packages without special considerations for heat dissipation.

When to Use a Class D Amplifier?

The most important reason why using a Class D amplifier is not suitable for all applications is that the switching of the output will cause electromagnetic interference. In many applications, this electromagnetic interference can be tolerated, so these devices can be considered to meet the electromagnetic compatibility certification, but there are other considerations for designers not to choose a Class D amplifier.

The second thing to consider with Class D amplifiers is that their sound quality is generally not as good as that of Class AB amplifiers and other technologies. Although comparing the two topologies on paper may lead to this conclusion, in some end-use applications, this is often no longer an issue because the distortion of the speaker is the main factor in system distortion.

What is a Half-Bridge/Single-Ended Class-D Amplifier

The half-bridge Class D amplifier has one output per channel. In half-bridge mode (single-ended output amplifier) ​​the speaker is connected to a single output in a dual supply system and grounded. In a single supply system, a large capacitor is used to prevent the VCC/2 DC voltage from appearing through the speaker load. This capacitor is usually several hundred microfarads or more, depending on the system's bass requirements and the speaker's rated impedance.

What is a Full-Bridge/Differential Class-D Amplifier

Half-bridge amplifiers are great for symmetrical dual-supply systems. The cost and size of the required DC blocking capacitors make them impractical for single-supply systems.

A full-bridge Class D amplifier has two outputs per channel. A full-bridge amplifier is also called a bridge-tied load (BTL) amplifier or a differential amplifier. In full-bridge mode, the speaker is connected to both outputs. The output bias of a Class D amplifier is very low, so no DC blocking capacitors are required.

The full-bridge amplifier provides the smallest system solution and is the most common Class D amplifier topology.

Is it a Class D amplifier or a "digital" amplifier?

In many cases, Class D amplifiers are not digital amplifiers. There are several reasons for this. In a basic open-loop digital Class D amplifier, the amplifier's power supply rejection is almost zero. In fact, the amplitude of the power supply is used for volume control. Another problem with pure digital amplifiers is the mismatch in output delay, propagation time, and overshoot. These combine to produce output nonlinearity, which produces harmonic distortion.

Fortunately, analog methods can mitigate the above disadvantages. The vast majority of Class D amplifiers use global feedback and error correction techniques in the analog domain. This allows THD+N in the range of 0.01% and PSRR in excess of 08dB to be common.

Some Class D amplifiers are true digital amplifiers. Dedicated digital circuitry compensates for the nonlinearity of the output limitation. A true Class D amplifier actually takes up only a small portion of the Class D amplifier chip.

What are the advantages of a half-bridge Class D amplifier over a full-bridge amplifier?

Half-bridge Class D amplifiers use smaller chips to achieve the same order of power, thus minimizing the cost per watt relative to the cost of the entire chip. Half-bridge devices require a DC blocking capacitor in a single-supply system, thus negating the overall cost advantage and making the overall solution larger.

In applications with many large LED backlights powered by, say, 24V, a half-bridge amplifier will give a better price/performance ratio, with a corresponding drive load of 8Ω and 8W-10W per output channel.

What other types of amplifiers are there?

Commonly used audio amplifiers are:

* Class A

* Class B

* Class AB

* Class D

* Class G

* Class H

By searching the web, the reader can obtain descriptions of these types of amplifiers. Some chip manufacturers have added the "new" type to describe their special varieties of Class D. The discerning designer will see the advantages of these devices and will not be confused by the latest marketing description.

What about a Class D headphone amplifier?

Class D headphone amplifiers have been around for many years. Their application revolves around the characteristics of the headphones themselves. The low EMI performance of Class D amplifiers requires engineers to control the length of the wires, the type of wires, and the load impedance of the speaker. All of these controls are in vain when the user connects any headphone to a standard headphone jack. In fact, the headphone wire itself is a very good antenna, and many portable devices use it as an FM receiving antenna.

In a self-powered headset, this is not a problem. Many Bluetooth headsets have full-bridge Class D amplifiers to provide the best battery life. In such a system, the headphone cable will be short and the load impedance is known.

Can a Class D amplifier operate on lithium batteries?

The standard operating voltage of Class D amplifiers for portable devices is 3V to 4.2V, which is ideal for use with lithium batteries or lithium polymer batteries. Within this supply range, the available power varies with the battery voltage. For example, when the load is 8Ω, a supply voltage of 3V means a power of 500mW, while a voltage of 4.2V means a power of 1.1W. The actual performance depends on the Class D amplifier.

How to obtain stable output power within a larger power supply range?

The high efficiency of Class D amplifiers makes them ideal for boost-powered systems. Some applications require 1W of output power into an 8Ω load, regardless of battery voltage. In such systems, using Class D amplifiers can meet such performance requirements.

Some Class D amplifiers have a boost converter, such as the LM48510. This provides a switch-mode power supply to the amplifier, so the Class D amplifier can operate when the voltage exceeds 5.5V. Another benefit of this approach is that the boost can be used as an LED flash or lighting.

What is a filterless Class D amplifier?

The filterless Class D amplifier has an adaptive output modulation function, so the tower can be used to connect directly to the speaker without the need for filters in between. The filterless Class D amplifier can be used in applications where the headphone cable is less than 10cm.

If the amplifier is not filterless, a filter is recommended. The pulse width modulation (PWM) waveform will cause high I2R distortion in the speaker's voice coil, which will reduce battery life and may damage the speaker.

Why does my filterless Class D amplifier have a filter?

Many filterless Class D amplifiers, such as the LM4675, still include a filter on their demo boards. This filter is intended to allow the user to test the performance of the Class D amplifier using a typical oscilloscope and audio analyzer. The PWM waveform is much stronger than the sound signal itself, so it can drive the input of these test instruments. The presence of the filter allows the system to be evaluated using standard test equipment.

Do I need a filter for my Class D amplifier?

In applications with short connection lines, the answer to this question is: No. Some Class D amplifiers use spread spectrum clocking to reduce the RF energy appearing at the output. Edge rate limited circuits reduce the actual RF energy that bleeds into the output. Combining spread spectrum and edge rate limiting, such as the LM48310, can achieve the best EMC certification without the need for an output filter.

How to design filters for Class D amplifiers?

L = (0.225 * RL) / fC

C = 0.113 / (RL * fC)

The same formula and/or software can be used for low-pass filters for Class D amplifiers, such as speaker crossovers. In most applications, a second-order Butterworth transfer function will provide the best combination of performance, including sensitivity and cost. For a single-ended amplifier, this formula can be expressed as:

L = (0.225 * RL) / fC

C = 0.113 / (RL * fC)

Where RL is the impedance of the speaker, fC is the ideal cutoff frequency, L and C are the inductance and capacitance of the filter respectively. For example, when the load impedance is 8Ω and the ideal cutoff frequency is 30kHz, the inductance value is 60μH and the capacitance value is 0.47μF.

Unfortunately, 60μH is a non-standard value, so we need to increase this value to the standard 68μH. By working backwards through the inductance equation, the new cutoff frequency is 26.5kHz, so we get a new capacitance value of 0.53μF, which can be approximated by connecting a 0.47μF capacitor in parallel with a 6800pF capacitor.

L1 = L2 = (0.113 * RL) / fC

CTOT = 0.225 / (RL * fC)

For a full-bridge Class D amplifier, the above formula can be modified as follows:

L1 = L2 = (0.113 * RL) / fC

CTOT = 0.225 / (RL * fC)

CTOT = CS1 + CS2 + (2 * CD1)

Where L1 and L2 are the two required inductors and CTOT is the total load capacitance. The load capacitance of a full-bridge Class D amplifier can usually be obtained by the following formula:

CTOT = CS1 + CS2 + (2 * CD1)

Where CS1 and CS2 are parallel capacitors connected to ground, and CD1 is the differential capacitor. For example, for a load impedance of 8Ω and an ideal cutoff frequency of 30kHz, the inductance is 30μH and the capacitance is 0.934μF.

Unfortunately, 30μH is a non-ideal value, so it needs to be modified to the standard 33μH, which requires reversing the inductance formula to get a new cutoff frequency of 27.4kHz, and then a new capacitance of 1.03μF. The required CTOT can be obtained by setting CD1=0.47μF, CS1=0.047μF, and CS2=0.047μF.

The biggest advantage of this method of separating filter capacitors is that it can achieve good electromagnetic compatibility and good audio performance. The larger the CD1 value, the better the filtering performance of the audio frequency band; the smaller the CS1 and CS2, the more it can reduce high-frequency interference during electromagnetic compatibility testing.