The master/slave relationship between dm8168 and aic3106.

SuitJune Young

Some days ago,I remembered that DM8168 acts as slave and aic3x(xx) codec as master.However,when I read DVRRDK 4.0's kernel code,I found that it seems that DM8168 acts as master and aic3x(xx) as slave.Is it so?

If it is the case,why TI makes a such change?(All right,maybe DM8168 is always a audio master).

over 12 years ago

0 ketty Lu over 12 years ago

Intellectual 950 points

In the kernel code, DM8168 is slave and aic3x is master.

0 SuitJune Young over 12 years ago in reply to ketty Lu

Genius 3985 points

Is any TIer here?

I noticed that according to the Linux Kernel code,alsa part,aic3x is a master in DVRRDK 3.0 or 4.0.However,if the aic3x case,even if aic3x is a master,the driver of davinci-mcasp.c,AHCLKX of DM8168 is a outpout pin.Why does a UNMATCHED operation?Slave output clock?

0 ketty Lu over 12 years ago in reply to SuitJune Young

Intellectual 950 points

the AHCLK register value is 0x00008000.

15 HCLKXM Transmit high-frequency clock source bit.
0 External transmit high-frequency clock source from AHCLKX pin.
1 Internal transmit high-frequency clock source from output of programmable high clock divider.

Bit15 HCLKXM is 1, it is internal.

0 SuitJune Young over 12 years ago in reply to ketty Lu

Genius 3985 points

thanks for your reply.

yes,the AHCLKX is output.But aic3x is a master,so bclk/lrck should output from aic3x.

But why aic3x's MCLK is get from DM8168's AHCLKX pin rather than from a external oscillator?

Processors

Processors forum

The master/slave relationship between dm8168 and aic3106.