[sdiy] CoolAudio? USB audio streaming/codec test question

[Brian Willoughby]

Another important consideration is this: The audio quality of a codec is only as good as the quality of its clock signal. Thus, I'd be very wary of hooking loose wires between an evaluation board and some processor board since this could easily end up with signal issues. A well-designed PCB with main processor and codec together, routed carefully, would probably perform better than something jumped together.

Brian


On Feb 22, 2022, at 16:51, Brian Willoughby <brianw at audiobanshee.com> wrote:
> I always evaluate with the exact processor that will be used in the product, which means that the codec and processor must be chosen at the same time. Not all synchronous serial interfaces support the exact mode(s) that the codec uses, so these details have to be confirmed early. Once compatibility is determined, I always build the first prototype with the main processor, codec, and analog support circuitry. From there, firmware needs to be developed to set the processor serial interface to the right mode, and then to communicate with the codec to select the desired sample rate and word size.
> 
> There are often separate interfaces for digital audio versus control of settings. In the case of the ADAU-1772, there is an I2C/SPI interface for settings, in addition to the synchronous serial, 1-bit, digital audio input and output. You'll need to decide whether I2C is sufficient or SPI is needed for performance, and make sure the evaluation occurs using the selected interface. Then, for the audio data, you'll need a synchronous serial interface that can share BCLK and LRCLK across both input and output (unless you're not using half the codec for some reason).
> 
> The ADAU-1772 has extra pins for mic bias and digital microphones, so those are added challenges to evaluate unless those features aren't needed.
> 
> If an additional interface like USB is needed, then that might be evaluated separately, in order to keep the number of "moving parts" to a minimum. I would also consider the selection of a main processor that has proven USB firmware, because that can take significant time all on its own.
> 
> I tend to favor the school of rapid prototyping - fail early, fail often - so my evaluation happens at the same time as the first prototype. This requires experience, but prevents unexpected surprises when transitioning from cobbled-together evaluation boards to the actual product. Thankfully, I've never had to go back to the drawing board and start over.
> 
> Brian Willoughby
> 
> On Feb 22, 2022, at 14:40, Barry Klein wrote:
>> Hi guys,
>> I am tasked with checking out an Analog Devices codec (ADAU-1772) and want to interface to the digital audio IO.
>> I see MiniDSP has a USB Audio Streaming kit thing that may be able to do it and CoolAudio has a TCD3020 but they seem gone (what happened?).
>> How do you guys go about validating a codec other than just using the eval board Analog Devices sells?
>> 
>> Barry