How to select and use an audio codec and microcontroller for embedded audio feedback files
To get started, developers can experiment with the
Figure 3: The Maxim Integrated MAX9867 audio codec can drive stereo headphones and select between digital, microphone and line inputs. Image source: Maxim Integrated
Figure 5: The MAX9867EVKIT+ eval kit for the MAX9867 connects to a PC over a USB cable and features RCA inputs, headphone outputs, and fiberoptic transmit and receive modules. Image source: Maxim Integrated
Conclusion Embedded system users have become accustomed to quality audio to the point that it is now expected instead of buzzers and beeps for alarms, alerts, and other user audio feedback. This puts the onus upon development teams to implement MP3 playback capabilities in their systems. This can at first appear to be a complex endeavor. However, by using the right audio codec alongside a microcontroller, and by following some design best practices, developers can balance the cost and complexity associated with audio applications.
MAX9867EVKIT+ evaluation kit for Maxim Integrated’s MAX9867 (Figure 5). The kit comprises the board and associated software and comes configured to send and receive audio data using the Sony/Philips digital interface (S/PDIF), though it can also be set to use I2S. It has two RCA input jacks, two 3.5 millimeter (mm) analog output headphone jacks, and fiberoptic receive and transmit modules. The software is Windows compatible, and when connected to a PC over a USB cable it opens into a graphical user interface (GUI) through which the developer can experiment with the MAX9867’s settings (Figure 6).
monitoring the power rails during testing. It is not a bad idea during pc board development to leave extra footprints on the board to allow different capacitance values to be tried in order to tune the output circuitry. Tips and tricks for selecting and using an audio codec Audio codecs can dramatically simplify the embedded software and provide an application with great sounding audio quality. Audio codecs can be tricky if a developer has not worked with them before. To successfully leverage an audio codec, there are several ‘tips and tricks’ teams should keep in mind such as: ■ Use the direct memory access controller (DMA) feature within a microcontroller to feed the audio codec with minimal CPU intervention. This will help to ensure that the codec is not ‘starved’ for data ■ When audio is not being played,
and the method decided on for playback, but a generalized diagram is shown in Figure 4. There are several points in this diagram that are worth discussing. First, there needs to be some method that is used to store the audio playback files. There are two options for this; store the files internally in the microcontroller flash memory or store them externally in flash memory. The choice will depend on how large the audio file(s) are and how large the internal flash memory is on the microcontroller. Developers also need to consider what the audio playback format will be. The most common is to use an MP3. In this case, the selected microcontroller needs to have a software stack that supports MP3 decoding. This allows the MP3 file to be opened and then pushed using a dynamic memory access (DMA) controller out via the I2S interface. Even the I2S port can be configured for master/slave and several other modes, so this needs to be carefully examined to ensure that the data is transferred to the codec at the correct rate.
As mentioned earlier, an external audio amplifier may or may not be needed depending on the application. A typical codec outputs around 1 to 1.5 watts, which is useful to drive a small speaker. To drive a 3 watt or larger speaker, it will be necessary to use external amplifier. Again, the most widely used are Class-D. The amplifier does not necessarily need to have variable gain either. The audio codec can adjust volume control digitally to provide a wide range of output power. One area that is often overlooked is bulk capacitance. When audio is playing, it can pull heavily on the power rails. If there is not enough capacitance on the board, the output quality can be dramatically affected and can take on a twangy sound along with several other unwanted noises. This can be detected by carefully
use the codecs mute feature to prevent low-level output noise from
reaching the speaker ■ When disabling or enabling playback, use an audio codec’s soft mute feature to prevent speaker popping and other unwanted noise ■ Use a terminal application to output the codec registers after the codec has been initialized. This can be especially useful
References
1. https://en.wikipedia.org/wiki/ Audio_codec
when attempting to debug issues or tune the speaker
output circuitry and enclosure ■ Leverage the internal digital filter mechanisms included in a codec. The digital filters allow a developer to equalize the output, filter out unwanted high and low frequencies, and maximize the quality of the sound system ■ Do not forget that tuning the sound will only be a useful endeavor when the circuit board and speaker are installed in the
Figure 4: A generalized connection block diagram for an audio playback system in a typical embedded application shows that there needs to be storage for audio files, which can be on the microcontroller or on external memory. Image source: Beningo Embedded Group
enclosure, as the enclosure and mounting make a huge difference
Figure 6: Using the Windows-based GUI, users can experiment with a wide range of MAX9867 settings, starting with Clock and Digital Audio (selected tab), all the way to Registers 1 and Registers 2 (right). (Image source: Maxim Integrated)
we get technical
48
49
Powered by FlippingBook