Post Go back to editing

LVDS clock circuit for AD9780, AD9781 and AD9783

I'd like to generate a programmable clock input into the AD9780/81/83 DAC using an LVDS clock source from my FPGA.  I noticed that the DAC sampling clock inputs (CLKP/N) have some special requirements and I wanted to share a possible solution with Engineerzone.

I'm using an LVDS-to-LVPECL buffer (part number MAX9377) to translate my source clock signal to a standard LVPECL differential signal. Then, instead of the (VCC-2V) termination for a standard PECL transmission line, I'm using a simple resistor divider as described in the datasheet to generate a 400mV common mode voltage.  I've attached my circuit diagram to this post.  Can anybody comment on the suitability of this circuit?  Will it work?

Many thanks.

Parents
  • Okay, thanks for your feedback.

    As an aside, I'm not sure why the DAC clock input into the AD9780 is so

    contrived.  Sticking to a standard LVDS clock input (as per all the

    other input signals) would have made the IC much more user friendly!

    In any case, I'll talk to Maxim and see what they recommend ...

  • Hi  . We have hit the same roadblock as you have w.r.t the clock driving circuit for AD9783. Your proposed solution seems quite neat, apt for my system as I intend to use LVDS clock out from my FPGA.

    But I am not sure if the voltages would be compatible (as the clock receiver on AD9783 is 1.8V powered not 3.3V). Instead of me reinventing the wheel, itt would be really helpful if you can let me know whether this circuit worked for you or not ?

    Did you have to add small resistors to ground on the outputs of the MAX9377 (before the ac coupling cap) to provide dc for falling edge currents (similar to the image shown below) with Vterm set to 400mV? 

    Thanks in advance !!

Reply
  • Hi  . We have hit the same roadblock as you have w.r.t the clock driving circuit for AD9783. Your proposed solution seems quite neat, apt for my system as I intend to use LVDS clock out from my FPGA.

    But I am not sure if the voltages would be compatible (as the clock receiver on AD9783 is 1.8V powered not 3.3V). Instead of me reinventing the wheel, itt would be really helpful if you can let me know whether this circuit worked for you or not ?

    Did you have to add small resistors to ground on the outputs of the MAX9377 (before the ac coupling cap) to provide dc for falling edge currents (similar to the image shown below) with Vterm set to 400mV? 

    Thanks in advance !!

Children
No Data