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.

PCM2704: Problem with HID Device Descriptor PCM2704

Part Number: PCM2704

Hi, i am doing a big project where i am use this USB DAC. At first all goes correct (on the PCB Prototype), but at sometimes when i plug the device to the computer it gives a Bad HID descriptor device, its strange and i cannot understand the problem. If i replace the PCM for a complete new one... it works again, and this its a cycle, sometimes it fails in a small period of time, others in a bigger one.

Should i have anything wrong on the schematic?, NOTE: The 1Mega resistor of the crystal isnt populated¡.

Any advice will be appreciate, regards, Carlos.

  • Hi Carlos,

    My suspicion is that the crystal resonator is not oscillating correctly.  Some crystals require exact load capacitance.  You have a comment on your schematic that you are using  22pF or 30pF.  The caps should be about C22=C23 = (2×CL) - Cstray.  It is difficult to estimate Cstray, but it usually 2-10pF depending on the board routing and how clean the board is.  Sometimes putting your figure on the crystal is enough to start it to oscillate, so if probing the crystal changes the performance, then I think that is the cause.  

    You can try a few different C values, as well as adding the 1MΩ resistor back to the design.

    Thanks,

    Paul

  • I still have the same problem, now i test with both capacitors value. Maybe i should change the crystal, i gonna see it under the oscilloscope.

    And try to analize the usb with the Saleae.... but this problem its kinda strange, maybe the usb lanes burn out the integrated if it doesnt have the tvs protector diodes?.

    His use its as internal USB-DAC

  • The crystal isnt working, i test with 22 and 30 pF and it dont work properly. I test the IC with a external clock, and works perfect.... but why the crystal isnt oscillating?..

     This is the Datasheet, someboy can help me?.

    Regards, Carlos.

  • There can be a few reasons why a crystal does not oscillate, though usually, it is due to the wrong cap load.  First, I would check that your soldering is well flowed and the board is very clean.  Excessive flux residue can create additional current leakage paths that might be affecting the total impedance of the design.  You could try removing the capacitors entirely, then try a few new values at 10pF increments.  While your crystal does not specify the need for series resistance, I have seen that make a difference in performance before.  Try adding 100Ω in series with the crystal.

    Thanks,

    Paul

  • Thanks, Paul finally i solve that with 22pf and 1Mega resistance.

    Kinda strange this behaivour but, after an extensive test i can say that its working perfect.

    Now i have another problem, on windows the HID functions works perfect, but under linux i cannot VOL up/down or mute with the IC pins. Any idea?, i will be working to see the linux drivers those days.

    Regards, Carlos.

  • Very cool Carlos.  As you can see, some of these oscillators require very specific load caps and resistances.  

    I do not know about the HID functions not working on Linux.  TI does not really offer any support on the Linux drivers, but there are many resources in the open-source community that might be able to help with that.  The folks at kernal.org and roon have some support threads for the device.

    Thanks,
    Paul

  • Thanks Paul, after read the alsa mail-list i see that the snd_usb_audio doesnt have support to hardware control volume... so i must write my own driver or use another method to work with.

    Thanks for your help.