TLV320AIC3110: Data link requirements to stream audio to from this CODEC

Hi Craig,

I'm not sure exactly what you're asking for, but let me know if this helps.

All of our Codecs use I2S/TDM as the standard digital serial audio format. Data rate is not really a word used in digital serial audio, but it's analog would be the frequency of the Bit Clock. Bit clock is a function of the sample rate, bit depth, and the number of channels, and it defines how much time is between two bits on the data stream. Usually sample rate is fixed, so the BCLK required is a trade off between how much resolution you want vs how many channels you need. For most of these codec applications, we are working with stereo only (2 channels) and so bit depth is the only thing to consider. 24 bit audio is about as high as you can go in practice. 32 bits is an option, but the dynamic range of our parts (most parts in the industry actually) is not high enough to take advantage of 32 bits. It's mostly just there out of convenience.

So common BCLKs I see often are

48k * 2 * 32 = 3.072Mhz

48k * 2 * 24 = 2.304Mhz

44.1k * 2 * 32 = 2.8224 Mhz

etc.

Let me know if there's anything else I can clarify,
Jeff McPherson

Hi Jeff

Thank you so much for your detailed response, that information will be very useful in me assessing use of this CODEC.

Perhaps it is better for me to explain a potential end use that illustrates what I need to know.

I want to implement a point-to-point intercom. Each unit has a CODEC, MCU, speaker and microphone and a communications interface. Each MCU supports an interface to the CODEC and other to a communications port. The MCU provides support for a communication protocol and implements data error detection/correction and probably compression algorithm to reduce the data throughput. It is voice grade so 300-3,000Hz audio bandwidth.

For a system like this, can I just assume a two-way data rate as would be necessary for a more basic CODEC such as a CVSD type and handle the conversion in firmware.  Is there anything in particular I should be aware of, other than treating it as I would any other digitised voice communications channel?

Cheers

Craig