Post Go back to editing

AD9102 SPI clock polarity

Hello,

I'm designing a waveform generator with the new AD9102 device, and discovered that the data sheet does not clearly specify which edge of the SPI clock is active.  There is a qualitative timing diagram in the "SPI PORT" section of the data sheet that seems to show that when writing to the device, data is transferred on the rising edge of the clock.  Is this correct?  If so, it is opposite from the convention for many Analog Devices DACs (including the excellent new AD5689R, for which I completed and tested a little circuit card just this morning).  However, it appears that a lot of DDS chips use the rising edge of the clock, so I guess this might make sense for the AD9102/AD9106 as well.

  -Ed

Parents
  • Thanks, Larry,

    I had already gotten the chip to work by assuming this, but it's good to get confirmation that this wasn't just accidental.  To date, this remarkable chip seems to be performing as expected with one minor exception.  I can't get the built-in random noise generator to do anything except to randomly jump between two discrete values, rather than randomly selecting among all 2^14 output values as I had expected.  Is there something that needs to be configured other than setting the AD9102 for continuous waveform generation, selecting the pseudorandom sequence with the WAV_CONFIG register (0x27), and setting the gains to reasonable values?

    My actual requirements need only the arbitrary waveform generator and the DDS, not this built-in pseudorandom noise feature.  It's just one of those things that seems like it ought to work, and it would be nice to know it's available.

      -Ed

Reply
  • Thanks, Larry,

    I had already gotten the chip to work by assuming this, but it's good to get confirmation that this wasn't just accidental.  To date, this remarkable chip seems to be performing as expected with one minor exception.  I can't get the built-in random noise generator to do anything except to randomly jump between two discrete values, rather than randomly selecting among all 2^14 output values as I had expected.  Is there something that needs to be configured other than setting the AD9102 for continuous waveform generation, selecting the pseudorandom sequence with the WAV_CONFIG register (0x27), and setting the gains to reasonable values?

    My actual requirements need only the arbitrary waveform generator and the DDS, not this built-in pseudorandom noise feature.  It's just one of those things that seems like it ought to work, and it would be nice to know it's available.

      -Ed

Children
No Data