• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 2753 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.

Linux/pcm5122: trouble using clocks for SCK

Keith Grider
Prodigy 50 points

Part Number: PCM5122
Other Parts Discussed in Thread: LM4916, PCM5122

Tool/software: Linux

I am using this DAC with a raspberry pi. I have built a board and have loaded the driver. I initially did not populate the crystals connected to the SCK pin in order to test the board without the clocks. This test was passed as all worked as expected (this is the 3rd iteration of this board so no surprise to this point). I then populated the crystals and powered up the unit and had low output volume (for similar settings) and high distortion. I am using a software driver for the Pi which also uses the 24.576mhz and 22.5792mhz oscillators (hifiberry dac plus pro) with both of them being controlled (enabled/disabled) via GPIO 3 and GPIO 6. I am doing the same on my board. Checking with the scope when listening to an MP3, the GPIO 6 pulls high to enable the 22.5792 crystal and I was able to see the oscillator frequency on this SCK pin. I am controlling the chip with i2c and it is on address 4c. It appears to be correctly identified by the OS (through alsamixer), but the output is distorted and very low even at 100% volume. Can anyone here point me in the proper direction? My supply checks out at 3v3. I am using a pre-built software image, currently, (have tried both Volumio and Rune Audio which are running MPD with ALSA) and have the same issue,

Thanks for any help

Gearhead

  • 0
    TI__Mastermind 46290 points
    Keith,

    The primary Audio Applications Engineer for our organization is out of office this week at a trade show. I apologize in advance for my fumbling around in an attempt to support you while he is out, but he will return on Monday.

    You mentioned that this is the third iteration of the board and that previously everything worked correctly without the crystal oscillators. In the previous boards, what was the SCK source?
  • 0 in reply to Kevin Duke
    Prodigy 50 points

    SCK was connected to gnd in the previous designs.

  • 0 in reply to Keith Grider
    TI__Guru 57363 points
    Hi Keith,

    What are you using for the crystals (part number)? I assume they have high impedance output when they are disabled. If you disable both clocks and source I2S does the sound quality improve?
  • 0 in reply to Paul_Frost
    Prodigy 50 points

    Paul,

    Thanks, but I figured it out. The DAC works fine. I had the analog output connected to a LM4916 and apparently the 5122 overloads the 4919. I was checking it with a set of headphones connected to the LM4919. When I connected the output to a proper amplifier, it worked fine. Switches between the 2 crystals based on file format and it sounds great. Thanks for looking. How could I interface these 2 chips, correctly?

  • 0 in reply to Keith Grider
    TI__Guru 57363 points

    Good to hear Keith!  The LM4916 is meant for very low voltage operation, 0.9 to 2.5V.  The output of the PCM5122 is a 2VRMS output (~5.6Vpeak-peak centered at 0V).  I suspect that your were over driving your inputs.  At what gain do you have the LM4916 configured for?  Assuming you are using it in single ended operation, you should be able to attenuate the output of the PCM to a voltage range that the LM4916 can handle.

    Thanks!

    Paul