I2C or SPI is just control and signal integrity may not be very important as long as the interface remains functional at its highest frequency but isolating I2S is going to introduce a lot of jitter ...
In this case the purpose of isolation would be to eliminate noise from MCU seeping through ground or power supply. Good isolators do not introduce lot of jitter and in DS DACs only MCK jitter is critical. And in a proper design MCK is local to the DAC and does not need to go through isolators.
Does it make sence in isolation I2S inside DAC if we have isolated USB from PC + linear supply?
While I'm waiting for stm32 board and screen with encoder, started to do a first step. It will be a long way I think.
While I'm waiting for stm32 board and screen with encoder, started to do a first step. It will be a long way I think.
USB isolation does not mitigate noise coming from the chip used for I2S generation (e.g. XMOS). Most commercial chinese DACs do not have I2S (or even I2C/GPIO) isolation but e.g. JLSounds I2SoverUSB has I2S isolation. I have I2S+I2C+GPIO isolators in my USB-I2S boards.