• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 635 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.

PCM1789: Bypass interpolation filter

Vitor Costa
Prodigy 50 points
Part Number: PCM1789
Other Parts Discussed in Thread: PCM1792A, DAC81001, DAC91001, DAC8811, DAC11001A

Greetings!

I'd like to know if it's possible to bypass the PCM1789 oversampling digital filter (ie, not use it).

I was planning to take advantage of its high resolution in a closed loop control application, but the group delay of 19 or 28/fs will severly limit the bandwidth. For instance, considering only the 19/fs delay at 192KHz, maximum BW should be 10KHz. For 28/fs it's even worse at 6.8KHz.

Best regards,

Vitor.

  • +1
    TI__Guru 53973 points

    Hi Vitor,

    You cannot bypass the filter on the PCM1789.  You can on some other devices in the PCM179x family, like the PCM1792A.

    You could also look at some of our precision DACs, like an MDAC (DAC8811) or unbuffered R-2R DAC11001A (DAC81001, DAC91001).

    Thanks,

    Paul

  • 0 in reply to Paul_Frost
    Prodigy 50 points

    Hi Paul,

    That's unfortunate. But thanks for the answer. I'll take a look at them. PCM1789's datasheet specifies resolution as min 16-bit and max 24-bit, while PCM1792A only says 24bit. What does it mean? By the way, neither of them specifies output glitch impulse. What can I expect with respect to that? In the case of PCM1789,should it be worse or better than say DAC8811 (which does specify it)?

    Best regards,

    Vitor.

  • 0 in reply to Vitor Costa
    TI__Guru 53973 points

    Audio data is usually sampled at 16-, 24-, or 32-bit increments.  Most audio DACs can support a dynamic range ratio to justify between 18-22 bits of data.  If you use 16-bit data, your dynamic range is reduced.  The PCM1792A would also be able to support 16-bit data, but then you are basically buying more dynamic range than you are providing data for.  

    These PCM devices are delta sigma DACs, meaning the output is switching at a higher frequency that the audible band.  For that reason, glitch measurements are not as valid, so we do not specify it.  Another thing to consider is that audio DACs are not really a 16, 24, or 32-bit DAC from a precision perspective.  We do not specify linearity for these devices. 

    I think you should look at the DAC11001A (or the lower resolution DAC91001, DAC81001) for precision control loop applications.  These devices have track-and-hold circuits that minimize glitch.  As such it can achieve excellent THD and SNR specifications, as well as linearity and glitch.

    Thanks,

    Paul

  • 0 in reply to Paul_Frost
    Prodigy 50 points

    Thanks! I think I'll keep the PCM1789 for an open loop section, and use use a faster precision dac for the control loop. I have another question: the default protocol is I2S. Is it possible to drive two PCMs in the same bus, with an FPGA as the master? Do I need one bus for each one? I couldn't find anything like a slave select in the datasheet, and I'm not too acquainted with I2S.

    Best regards,

    Vitor.

  • +1 in reply to Vitor Costa
    TI__Guru 53973 points

    This device only accepts 2-channel I2S formats, so if you want to use 2x devices, you will likely need two data lines on the I2S bus.