Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
?
EE Times-Asia > Amplifiers/Converters
?
?
Amplifiers/Converters??

Design Class D audio amps into portable applications (Part 2)

Posted: 18 Oct 2007 ?? ?Print Version ?Bookmark and Share

Keywords:Class D amplifiers? audio amps? portable applications design?

By Nicholas Holland, Greg Hupp and Andy Liang
Texas Instruments Inc.

Read Part 1

Highly-efficient Class D audio amplifiers are becoming popular for use in cellphones, smart phones, PDAs and similar portable applications in place of Class AB amplifiers. The use of a Class D audio amplifier allows the end product to operate longer on batteries, and run cooler, which solves thermal design problems. Test results showing the advantages of Class D amplifiers quantify the battery life and thermal improvements achieved. Comparisons between Class AB and Class D amplifier architectures highlight areas that require more attention to the board-level design for Class D amplifiers. Proper board layout, component selection and amplifier measurements differ from traditional Class AB amplifiers.

Selecting, placing external parts
The selection and placement of external components is another area to consider when designing and laying out the circuit. These are the extra components (e.g. resistors, capacitors and ferrite beads) required for the proper operation of the audio amplifier. For an audio amplifier in which the gain is set by the input resistors, make sure the input resistor is as close to the input pin as possible. This is because the input pin is a summing node and noise can couple into this node easier than other places in the circuit. Therefore, minimize the trace length connected to this pin to minimize the area for noise coupling. The input resistors must also be closely matched. Any mismatch in the resistors changes the gain slightly between each side of the amplifier. This results in a decrease in the PSRR of the audio amplifier, which allows noise on the power supply to be heard in the speakers. However, the price of highly matched or tight tolerance resistors can be a deterrent to using these components. So a tradeoff must be made and usually a 5-percent tolerance resistor provides an acceptable performance level.

Decoupling capacitors on the power supply should be placed as close as possible to the IC. These "local" decoupling capacitors help to stabilize the power supply, where as larger "bulk" decoupling capacitors can be placed further away from the device. Another design decision for capacitors is the value of the input DC blocking capacitors. The input DC blocking capacitors form a high-pass filter with the input impedance of the amplifier. If incorrect values are used for the capacitance, some of the audio signal will be attenuated before it reaches the amplifier. The frequency can be calculated using the following equation:

Formula1

Where, RIN = input impedance of the amplifier CIN = input DC block capacitor

Ferrite beads should be placed as described earlier near the output pin of the amplifier. Selection of the proper sized ferrite bead is very important to provide proper functionality. The current rating must be sized correctly for the ferrite bead in order to prevent the bead from saturating. If the current rating is too low and the bead saturates, it will look like an AC short. If this happens the ferrite bead filter will function as a capacitor to the switching output, which looks like an AC short above a frequency determined by the capacitance. If the magnitude of the energy that shorts through the capacitor is too high, the amplifier could trip the short-circuit protection and turn off. This same behaviour occurs for capacitors that are placed directly on the outputs of the amplifier pins. Capacitance is sometimes used for ESD protection, but care must be taken since this type of capacitance on the output pins can prevent the amplifier from operating.

Measuring Class D amp performance
Measuring the output power of linear amplifier is generally a straight forward process. Due to their switching operation, the same methods cannot be applied when using Class D amplifiers. For Class D amplifiers, the outputs must be low-pass filtered to allow for accurate output power measurements, due to the limitations of the measuring equipment. This is also true for measuring the Total Harmonic Distortion + Noise (THD+N) with an Audio Precision (AP).

Texas Instrument's Class D audio power amplifier families (TPA2000D and TPA3000D) use a modulation scheme that does not require an output filter for operation. They do, however, require a low-pass filter to make an output power measurement or a THD+N measurement and to minimize EMI.

The 250kHz switching signal is seen as a common-mode voltage across the inputs of the audio measurement instrument. Typically, audio analyzing equipment has very low common-mode rejection at 250kHz, since it is designed to work in the audio band.

Even though most audio analyzing instruments have internal filtering, they still have input amplifiers that cannot respond to the fast rising edges of the PWM signal. Without affecting the audio performance of the amplifier, an RC filter can be used to remove the 250kHz switching component from the PWM output of a Class D amplifier.

Portable devices that use audio power amplifiers typically use speakers with impedances from 3次 to 32次. It is important to understand the relationship between peak output voltage, VO(P), peak-to-peak output voltage VO(PP) and the RMS output voltage VO(RMS). When calculating the power, the output voltage (VO) will be specified as an RMS value and the following equation is used:

Formula 2

where,

Formula 2a

The typical THD+N measurement combines the magnitudes of noise, distortion, and other undesired signals into one measurement and relates it (usually as a percentage) to the fundamental frequency magnitude. Ideally, only the fundamental test frequency of the sine-wave input is present at the output of the audio power amplifier. The THD+N measurement requires notching out the fundamental test frequency and measuring the RMS voltage (which includes unwanted harmonics and noise) across the audio band (which the AP does automatically) and then dividing that measured value by the fundamental test frequency value and expressing it as a percentage.

The low-pass RC filter
The cut-off frequency of the RC filter shown in Figure 8 is chosen to be around 30 kHz, since this is just outside the audio band and provides 20dB/decade of attenuation for higher frequencies.

The cut-off frequency for a first order RC filter is:

Formula 3

By simple component selection, e.g. 100次 for R and 0.047mF for C, the resulting fo is 33.86kHz and the corresponding circuit is shown in Figure 8.

Figure 8

Figure 8: RC low-pass filter for one channel

Note:

It is very important that the signal generator, Class D amplifier, filter box (GND) and oscilloscope or AP all be connected to the same ground in order to remove any common-mode voltage. The power supply ground should be used for this purpose. To simplify measuring the output power and THD+N of a Class D amplifier, Texas Instruments created a simple RC filter box that has two channels and basically includes the low-pass filter that is described in Figure 8.

Figure 9 shows a picture of the top of the box and the inside of the box.

Figure9aFigure9a

Figure 9: Top and inside view of TI RC filter box

The "SHIELD" connector is ground for the entire box. A connection to a quiet ground is required (the power supply) and not at the same ground as the amplifier and the measurement equipment.

This box can be used for both output power measurement with an oscilloscope and THD+N with an AP. This box is designed to allow easy connection to the various measurement devices either during the initial evaluation, prototype development or final test. These measurements can be done without the filter box, by using a simple RC filter as explained in Figure 8. The following sections explain output power and THD+N measurements using the filter box.

Measuring output power
Connect the filter box between the load and the oscilloscope.

Figure 10

Figure 10: Stereo Class D amplifier and RC filter box for output power measurement

Once the test circuit is correctly setup for either a mono or stereo measurement, the output power for each channel is calculated using the peak voltage (read off the oscilloscope) and the known load impedance. It is important to remember that this calculates the average output power of the amplifier.

Note: This circuit assumes differential probes at the oscilloscope. An alternative is to use single-ended probes with the oscilloscope and use the mathematics function to obtain the difference between both signals. The same output voltage measurement can be achieved using an RMS voltmeter.

Example:

The output power calculation for one channel:

If the peak output signal voltage on the oscilloscope is 2V and the load RL is 8次, the RMS voltage will be:

Formula4

Measuring THD+N
The setup is similar to the output power measurement, but an AP is connected to the amplifier's inputs and the outputs of the filter as shown in Figure 11.

White SMD LEDs from Vishay

Figure 11: Stereo Class D amplifier and RC filter box for THD+N measurement

The input bandwidth of the analyzer is usually limited with filters to reduce the out-of-band noise. However, this eliminates higher order harmonics that are needed for accurate THD+N measurements. The filter box attenuates all the non-audio band frequencies and therefore allows accurate reconstruction of the original audio waveform. Higher frequencies are not important for this measurement, because they are beyond the audible threshold of the human ear.

Conclusion
Designing with Class D amplifiers allow designers of portable devices with audio to extend the battery. When designing with Class D amplifiers, it is important to pay attention to layout, component selection and placement to avoid a reduction in performance of the device or system. Finally, understanding the different techniques used to measure the output waveforms of a Class D amplifier is necessary for accurate measurements of output power and THD+N.

Read Part 1




Article Comments - Design Class D audio amps into porta...
Comments:??
*? You can enter [0] more charecters.
*Verify code:
?
?
Webinars

Seminars

Visit Asia Webinars to learn about the latest in technology and get practical design tips.

?
?
Back to Top