The perfect combination of high bandwidth Internet connection, deep learning algorithm and high-quality audio brings a new model for user interface, that is voice. The ability to interact with home appliances and cloud services through natural language is a great change in the way of human-computer interaction.

Today’s digital assistants seem to have reached the pinnacle of today’s technology, but they are actually just the starting point of a new generation of user interfaces. Speech recognition is a technology that has existed for many years. For example, many software c920446619921ffca4b252220c1fe2a can convert speech into text. The difference lies in the way the software interprets words. So far, software only relay words to the screen, but recently, machines seem to understand intentions and words. This artificial intelligence technology will be popularized in the next few years, which means that it is easier to obtain technology and integrate it into daily necessities.

The challenge for manufacturers in the future is how to add a high level of audio quality to smaller electronic devices. The trend of wearable technology is to develop synchronously with AI and natural language interfaces. Therefore, combining the two requires a higher degree of integration to maintain fidelity.

Audio fidelity

The key part of audio fidelity is the selection of amplifier. Most amplifiers are linear in nature, including   A   Class, B   Class or ab   class   (refer to illustration   one   To compare the three). Although all linear amplifiers can provide good gain with minimum noise, due to the receiver   (human ear)   Very good at distinguishing the difference between the two, so the signal-to-noise ratio   (SNR)   It becomes more important in audio applications. The above amplifiers have their own advantages and disadvantages in power consumption and efficiency, so different bias categories are required.

chart   1a:A   Class amplifier

chart   1b:B   Class amplifier

chart   1c:AB   Class amplifier

In the current emerging applications, the efficiency of audio clarity and power consumption is very important. All linear amplifiers have a certain degree of bias to provide good linearity, but the power efficiency is poor. For example, a   Class a amplifiers use biased single transistors that turn on even if the input signal is zero. B   Class a amplifiers use two transistors, one transistor per half cycle, so theoretically, only one transistor is on at any time. This can improve power efficiency, but it usually causes crossover distortion. and   AB   Class I amplifiers reduce crossover distortion by slightly changing the bias voltage. although   AB   Class is an ideal compromise in most cases, but it still only provides about   80%   The efficiency may not be ideal for some battery powered and portable applications.

D   The topology of class a amplifier is very different from that of normal amplifier. It does not amplify the input nonlinearly, but digitizes the signal and converts it into pulse width modulation (PWM)   Square wave to present the input signal. The above “full open / full close” method can be provided   90%   However, it is still necessary to filter the output to recover the analog component of the signal, which may bring additional design difficulties. Nevertheless, due to a number of design features,   Pam8014 filter free   D   Class I amplifiers still show long-term progress in their field. Its provision   3.2 W   Mono output power and over   90% efficiency, high power supply rejection ratio and differential design, which can greatly reduce noise and RF   The effect of reflection.

Make better   D   Class amplifier

D   Class topology is usually implemented by comparator, which compares the input signal with the reference signal to establish the actual topology   1-bit ADC; As shown in the figure   two   It is shown that when the input signal is greater than the reference signal, the output of the comparator is high, and when the input is lower than the reference signal, the output signal is low.

chart   2:D   The analog signal is converted into a series of pulses by a comparator   PWM   Output, usually through a low-pass filter to recover the analog signal

The output is fed by two or four switches   FET   Composed of switching stages. double   FET   The output stage is usually composed of complementary components   FET (PNP   And NPN)   It is composed and operates as a half bridge. The disadvantage of this design is that even if the output is constant, the current will pass through the load because the output bias is   VDD/2。 The above problems can be overcome by using the full bridge output stage. The full bridge implementation uses four FETs composed of two and a half bridge stages. In this way, bidirectional swing can be generated in the output current, which is called bridging load   (BTL)。 Although this design uses twice the number of   FET, which also provides significant advantages, especially because the offset is applied to both ends of the load, thereby eliminating the losses associated with the high quiescent current of the half bridge design. chart   three   display   PAM8014   D   The block diagram of class a amplifier adopts full bridge   BTL   Design.

chart   3:PAM8014   Take the whole bridge   BTL   Topology based, very small package provides excellent performance

take   BTL   Topology consolidation to   D   Class a Amplifiers need to be carefully designed to achieve effective output power levels in a single integrated package. This is not a problem in audio applications using more advanced speaker technology. In fact, it is more portable and battery powered devices, creating many new potential applications.

No pass filter

If the class D amplifier still needs the output stage to pass through a large and possibly expensive low-pass filter, the advantages of the above system level may be lost. The independent passive components required to construct a low-pass filter (such as an inductor large enough to handle the output power) are likely to occupy more power than the amplifier itself   PCB   Space and significantly increase costs.

When removing the high frequency from the PWM output, a low-pass filter is usually required, which is an integral part of many applications. The additional PCB space and cost required for low-pass filters may prevent class D amplifiers from being used in some applications, but in fact, this may have a positive effect in audio applications.

The loudspeaker is actually composed of coils, which are essentially inductors with series resistance. These two components are the key to realize low-pass filter. In fact, the built-in inductor of the speaker can be used in this way, however   BTL   The design of the output stage must support the design without filter. This is important because otherwise, current may continue to flow through the speaker, which may shorten the service life. In addition, if the output is biased, the dynamic range of the speaker may be limited. PAM8014   D   Class amplifier   BTL   The output stage is designed for filter free operation, so it can avoid all the above defects and bring great benefits to audio applications   D   Advantages of class operation.

Although human ear is a natural high-frequency filter, it is also easy to receive various forms of noise. Therefore, an excellent class D amplifier will implement internal feedback to provide the best total harmonic distortion + noise (THD + n) value; In this case, the pam8014 typically provides 0.14% at an output power of 0.5 w @ 1 kHz.

Other advanced functions of pam8014 include shutdown mode, which can reduce the supply current to 1 µ a at most. Pam8014 also integrates special circuits, which can greatly reduce the “pop and click” that may be generated by transients when opening and closing or exiting from off mode. The special mode can mute the internal amplifier until the reference voltage is stable, so as to achieve the above effect; Once stable, the full operation mode will be restored. Other functions include: undervoltage locking (UVLO). If the supply voltage is lower than 2.0 V, pam8014 will enter the off mode; Short circuit protection (SCP), if the output terminals are short circuited or grounded, this function can protect the device; Overheating protection (OTP), if the grain temperature exceeds 150 ° C, will put the device into shutdown mode.

PAM8014   use   U-WLB1313-9   Wafer level   BGA   encapsulation   (see Fig   4) , each side is smaller than   1.3 mm, even for the smallest device.

chart   4:PAM8014   Wafer level   BGA   Package, each side size is less than   1.3 mm

PAM8014   The fixed gain is   18 dB, efficiency up to   93%, requiring only a minimum of external components, as shown in the figure   five   As shown in.  

chart   5:PAM8014   With only a few external components, it is very suitable for applications with limited space and ultra-low power consumption

These advanced features allow   PAM8014   Ideal for a variety of applications, including smart phones, VoIP, MP4 / MP3   Players and, of course, digital assistants.

Leave a Reply

Your email address will not be published. Required fields are marked *