• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 660 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Dac 37J82 output analog path

David Schick
Prodigy 30 points
Other Parts Discussed in Thread: DAC37J82

Hello,

We have the DAC37J82 eval board and are attempting to make it work with our Altera FPGA using the JESD IP core.  We can send data to the DAC over the JESD port, but we seem to be unable to bypass the internal complex mixer block.   For example, we can power down the internal mixer and internal NCO and we send a square wave through the DAC.  What we see on the scope is a series of impulse at the correct frequency, but no square wave.    When we turn on the mixer we see what appears to be the correct tone, but instead of an amplitude modulated signal we just see glitches.  Are you supposed to be able to bypass all mixer processing and use the DAC37J82 as a plain analog DAC without any complex mixing?

Thanks!

  • 0
    TI__Guru* 78744 points
    David,
    The complex mixer block is optional path. The data path may be used without any mixing at all. When mixer block is used along with the NCO, be sure the NCO is initialized correctly. The initialization depends on the syncsel_NCO and syncsel_mixer initialization signal selection setting in config 0x1F.

    Also, keep in mind that the eval board has transformer coupled output with a bandpass behavior of 3dB cuttoff from 10MHz to 1GHz. Any signal bandwidth below or above the bandpass range may not be reconstructed perfectly. A squarewave waveform requires the fundamental, and also 3x and 5x of the fundamental bandwidth for about 90% of good rise/fall time. I believe the reason that you see impulses is due to the bandpass behavior not passing DC components (i.e. steady hold of DC for squarewave). You may want to adjust the output to DC coupled with resistive termination.

    See app note for detail:
    www.ti.com/.../slua647a.pdf
    -Kang
  • 0 in reply to Kang Hsia
    Prodigy 30 points
    We were able to get a better signal when we moved the test signal to greater than 10 MHz. I do have a follow-up question about the interpolator block on the DAC. Does it work independently for the I and Q signals? Or does the interpolator require complex data (I and Q samples) in order to work properly? I ask because when I am sending 15 MHz out the Q signal (without the mixer turned on) and then I turn on the interpolator, I see spurious signals everywhere.
  • 0 in reply to David Schick
    Prodigy 30 points

    We were able to fix our problem - it turns out that you must reprogram the DAC output sample rate using the DAC windows GUI tool in the "Overview Screen" - not in only in the "Digital Block #1" screen. That GUI tool reprograms the DAC output frequency on the LMK clock chip appropriately.

  • 0 in reply to David Schick
    TI__Guru* 78744 points
    David,

    Good to hear.

    For your information, the interpolation FIR filter does *not* require complex data. The data among channels can be completely independent since each channel has their own FIR filters. The only block requires complex data is the complex mixer block.

    -Kang