TAS6424E-Q1: About using an oscillator to supply the MCLK clock

miyuki.fukumoto

Part Number: TAS6424E-Q1
Other Parts Discussed in Thread: TAS6424E

Tool/software:

Hi Team,

We are struggling to achieve both MCLK waveform quality and EMC's in our product, and when referring to answers on the Forum, we found an answer that said it is okay if it is asynchronous with SCLK and FSYNC.
Waveform shaping using a buffer is one item we are considering, but as a more reliable method we have come up with a method to add an oscillator for MCLK next to the IC and only control the enable.
We would like your opinion on whether this is a feasible idea.

With Best Regards,
Fukumoto

1 month ago

0 Shadow He 1 month ago

TI__Mastermind 36190 points

Hi Fukumoto

For TAS6424E device, the phase between MCLK and SCLK is not so important. As long as we could make sure the clock ratio within datasheet spec would be fine.

0 miyuki.fukumoto 1 month ago in reply to Shadow He

Prodigy 60 points

Hi Shadow san

Thank you for Reply!
I think we're talking about this part of the datasheet here, but how accurate does the MCLK need?

Shadow He said:
clock ratio within datasheet spec

Naturally, the center value will be set to the target value(128 × fS, 256 × fS, or 512 × fS), but are there any concerns if it changes differently from the AUDIO clock(SCLK,FSYNC) due to variations or temperature changes?

0 Shadow He 1 month ago in reply to miyuki.fukumoto

TI__Mastermind 36190 points

Hi Fukumoto

The clock itself could have ±10% variation. But we much guarantee the clock ratio to be always correct. If the clock ratio changes more than ±1, will possibly trigger clock fault.

This is indeed the risk of using two oscillator, customer may need to based on oscillator's spec to do some calculation, what would be the maximum variation.