ΣSigmaStudio Ask AI Русский Log in

MCLKI and OUTPUT_LRCLK/BCLK synchronous?

SigmaDSP & SigmaStudio ChristianH May 12, 2014 00:00 4 replies Original on forum ↗
Subscribe
ChristianH Question
May 12, 2014 00:00

Hi,

I was looking to some app notes and diskussions but found no final answer to this question.

I use an ADAU1701 with 3 external ADCs and 1 external DAC which will be connected via I2S to the SDATA pins.

I think I will use an oscillator with 12.288MHz to feed the MCLKI of the 1701 and also the master clock inputs of the external converters.

ADAU1701 should be the I2S master.

I'm not sure if the generated OUTPUT_LRCLK and OUTPUT_BCLK signals on the 1701 are synchronous to the oscillator clock feeding the MCLKI pin.

I would guess it but I want to be sure.

Can anybody confim this?

Otherwise I think I have to use one ADC as master and distribute this clocks to the other converters and the 1701.

Thanks!

Christian

Audio Signal Processors

i2s

SigmaDSP Audio Processors

adau1701

ChristianH Reply
May 15, 2014 00:00

Hi Bob,

that was my assumption, that the clocks are synchronous.

But in my opinion the description in the datasheet is not 100% clear.

The first paragraph says the clocks need to be synchronous - this could mean that you need an external clock master that feed MCLK, BCLK and LRCLK synchronous to the ADAU1701.

I was missing the statement like "an external master clock on MCLKI generates synchronous BCLK and LRCLK on I2S/TDM outputs if in master mode...".

Thanks

Christian

KJBob Reply
May 15, 2014 00:00

Hello Christian,

Triggering at the 1701's BCLK output (top), I obtained a stable MCLK trace at the crystal (bottom). Thus the two are synchronized.

The data sheet hints at this indirectly. Since any input I2S clocks must sync with MCLK (there's no sample rate converter), and since input from an external slave ADC is handled using the I2S output clocks, the output clocks must also be synchronous with MCLK.

Best regards,

Bob

d71ccb9fcb1014083c9b079197af303a.jpg
d71ccb9fcb1014083c9b079197af303a.jpg
d71ccb9fcb1014083c9b079197af303a.jpg-560x375.jpg
d71ccb9fcb1014083c9b079197af303a.jpg-560x375.jpg
af725a3704b0f0a06450aaec162984c6.jpg
af725a3704b0f0a06450aaec162984c6.jpg
af725a3704b0f0a06450aaec162984c6.jpg-560x265.jpg
af725a3704b0f0a06450aaec162984c6.jpg-560x265.jpg
EEtwidget Reply
May 15, 2014 00:00

just to add my $0.02 ...

Im also investigating ADAU1701 I2S with a CODEC. I plain to run the ADAU1701 as a slave. (short on IO)

12.288MHz OSC on the CODEC, it has a MCLK OUT which I will connect to the ADAU1701 MCLKI.

ADAU1701 CODEC

------------------ --------------

INPUT_BCLK --------------------- OUTPUT_BCLK

INPUT_LRCLK --------------------- OUTPUT_LRCLK

INPUT_MCLK ---------------------- OUTPUT_MCLK

If the ADAU1701 is the master, OUTPUT_BCLK/OUTPUT_LRCLK needs to be connected to ADAU1701 INPUT_BCLK/INPUT_LRCLK respectively. I am using all MP so slave was the way to go.

ChristianH Reply
May 16, 2014 00:00

Hi,

ok, so my conclusion is that if I feed MCLKI then the generated BLCK and LRCLK signals are synchronous to the MCLKI signal (ADAU1701 is the I2S master).

Thanks!

Christian