ADAU1701 CLOCK DRIVER

Is there an analog device  clock driver to drive 8 ADAU1701 I can use that is similar to the one shown  ?

Thanks in advance

     Hello izhmel,

     The DS1080CL includes some interesting features, but may not be ideal for your application.  It's a clock generator, not a clock driver -- so its output is only specified at a 15 pF load, and the data sheet gives no output impedance specification.  Thus it's not clear how well it would do driving eight -1701s.  You would definitely want to turn OFF its spread-spectrum feature, as such intentional clock variation would cause audible effects in the DSPs.  If you're looking to minimize measured EMI, I'm afraid this will have to come instead from best-practices design, shielding, etc.  Please see https://ez.analog.com/message/57200#57200 for an EMI example.

     One of our prototypes is working fine with an ordinary 12.288 MHz packaged oscillator, buffered with a 74LVC2G34, to drive five DSPs.  A 50 ohm series resistor placed near the buffer output reduces the ringing that would otherwise occur -- see, for example, https://ez.analog.com/message/130241#130241

     Finally, the 18 inch trace length among your chips scares me.  This may introduce issues not only with MCLK but also with any I2S you might be running between your -1701s.  Our chips are tight together as shown.  Perhaps another engineer having experience with the longer distance may suggest how to manage such issues.

     Best regards.

     Bob

What would be a good clock driver for 8- ADAU1701 ?

can you show a schematics driving the four  ADSU-1701 with the 74LVC2G34 ?

     Hello izhmel,

     Where I work I do the SigmaStudio programming, my friend Mike designs the hardware.  What little PC layout I've done I couldn't stand, but his layouts are art.  He graciously allowed me to display the relevant portion of our schematic below:

     It uses two SN74LVC2G34DBVR dual driver packages -- half of one drives MCLK from the packaged crystal oscillator, the other buffers BCLK and LRCLK from the first ADAU1701 to the other five.  A total of six ADAU1701s are served this way.  Clocks are connected to the -1701s according to the second option shown at:

https://ez.analog.com/message/134475#134475

     Best regards,

     Bob

Thanks you saved me money that would have been wasted..

Thanks

How much ADAU1701 can I drive using the PKG oscillator and the buffer circuit, and  is pin 31 on the ADAU1701 a NC pin  using this master clock circuit ?

Thanks in advance

    Hi,

    Since this scheme uses the packaged oscillator and not the -1701s' XTAL oscillators, pin 31 may be left unconnected:

    I can say for sure this works for six 1701s, since that's what we did.  The 1701's digital inputs present a 5 pF load, and the SN74LVC2G34DBVR is specified with a 50 pF load.  Perhaps this means you could do ten, but there's other factors such as trace lengths, etc.

     Best regards,

     Bob

Hello Hopeton,

I would also backup Bob's answer that trace lengths are a big factor. Then there are problems when you get reflections from many stubs in a transmission line. A stub mainly being each 1701 you connect to. Then if you plan on connecting each one with cables from board to board, I seem to remember that you were doing that. Then you may want to use an individual buffer to get to each board.

Dave T