The EV6550DHAT brings together the Raspberry Pi and the CMX655D Ultra-low Power Voice Codec in a convenient and inexpensive way, enabling easy demonstration of the CMX655D and providing improved low power audio capabilities to the Raspberry Pi. This is achieved using standardised hardware, opensource software and native drivers, creating a platform that is easy to use and that simplifies product development.
One of the major benefits of the hugely-popular Raspberry Pi is its low power, flexibility and accessible GPIOs which encourage developers to expand on its basic functionality. The multi-way GPIO header on the latest versions of the board consists of 40 mixed-function pins, 26 of which support general purpose input/output connections.
The Raspberry Pi HAT (Hardware Attached on Top) is a hardware standard defining an add-on board that mounts directly onto the Raspberry Pi and attaches to the GPIO header. The EV6550DHAT conforms to this standard and integrates the functionality of the CMX655D Ultra-low Power Voice Codec into the Raspberry Pi environment.
The EV6550DHAT enables the CMX655D’s feature set (two matched channels supporting a wide variety of digital MEMS microphones, signal processing and a 1W high-efficiency Class D speaker amplifier) to be easily demonstrated. The most immediate benefit is an improvement to the standard low power audio capabilities of the Raspberry Pi.
EV6550DHAT Raspberry Pi GUI and source code. Version 1.0 – GUI and source code supplied as three split 7-zip compressed files ready for download. Once all three files are downloaded they must be renamed from EV6550DHAT_00x.7z to EV6550DHAT.7z.00x and then decompressed and recombined using 7-zip to access the GUI and source files.
Q: Why when I look at the CMX655D PA output with an oscilloscope I do not see the expected analogue audio waveform.
A: Assuming that the CMX655D overload current and thermal protection circuits have not been triggered then the signal measured across the AOUTN and AOUTP pins will not be a true analogue representation of the sampled audio. This is because the PA is a Class D design where the audio is represented by a modulated PWM signal and as with all Class D amplifiers additional filtering and loading of the output is necessary for correct operation. Typical methods of measurement use high impedance probes or signal inputs and so do not appropriately load the amplifier. Therefore to achieve a meaningful measurement of the sampled audio the following intermediate output components are advised.