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.

PCM3003: Design Questions - SYSCLK tolerance, Power-On Reset Procedure, Capacitor Recommendation

Part Number: PCM3003


Hello,

We are working on a design with this device, and have come up with a couple questions during the process:

  • Are tantalum capacitors still recommended for this device, or are ceramic acceptable? (All capacitors >1uF are recommended to be tantalum, though this is an older product and may be a dated recommendation)
  • We are going to be operating at 48kHz sampling rate with a targeted 12.2880MHz SYSCLK provided with a 256fs divider. What is the tolerance on SYSCLK for the device to operate 48kHz / 256fs? (None appears to be provided in the datasheet, and exactly 12.2880 is impractical)
  • Power-On reset: Datasheet states 3 Clock cycles must be complete prior to Vdd > 2.2V. This implies that if Vdd sequencing cannot be controlled by the master, the poweron reset timing cannot be met. If this is the case, is an external reset procedure required on poweron?
  • Section "System Clock" states the following:
    • "When power is supplied to the part, the system clock, bit clock (BCKIN), and word clock (LCRIN) also must be supplied simultaneously. Failure to supply the audio clocks results in power-dissipation increase of up to three times normal dissipation and can degrade long-term reliability if the maximum power-dissipation limit is exceeded"
    • Which power supply is this in reference to? (Just Vdd, or VCC1/2?)
    • Do these lines simply need to be high as the supply comes up (such as through a pullup), or do they need to actively be applying clock pulses to prevent this high (possibly damaging) power dissipation?

Thank you

  • Hi Christian,

    I can comment on the first two points right now but I will need to do some digging on the other questions as this is an older device.

    - For the capacitor recommendation, ceramic is just fine. As I mentioned above, this in an older device and ceramic properties have gotten much better. I can say that on our EVMs now, we use ceramic capacitors. 

    - I don't think we have a tolerance value but as long as MCLK is generally close to 12.288Mhz, that should be fine. 

    I will do some digging on the other questions and get back to you some time early next week.

    Regards,

    Aaron Estrada

  • Thank you Aaron, I figured that was the case with the capacitors.

    Any update on your findings regarding the POR and other power-on behavior?

    Thanks

  • Hi Christian,

    For the POR, yes, an external RESET would be required if the POR requirements can not be met. 

    Regarding the system clock power dissipation, I believe both analog and digital supplies can be affected since the System clock, ADC and DAC are involved here. I would ensure that the actual clock is supplied and not just pulled up.

    Regards,

    Aaron Estrada

  • Hi Aaron,

    I have some quick follow-up questions. Rereading the "System Clock" section, I'm still a bit unclear of the actual requirement outlined here, and how to verify if we are meeting it.

    • "Supplied simultaneously"
      • Does this just refer to synchronization between the first rising/falling edge of each clock line?
      • What is the acceptable deviation to prevent excessive power dissipation? LRCLK and BCLK are significantly slower than SYSCLK
      • At what point must they be applied simultaneously to meet the requirement?
    • "When power is supplied to the part" and "Failure to supply the audio clocks results in power dissipation increase of up to three times normal dissipation" 
      • Does this refer to all the time the part is active? Or just when power is initially supplied (on power-up)?
      • Does this indicate we must be supply all clock lines constantly with power applied, even when not reading / writing to the ADC/DAC?

    Thanks

  • Hi Christian,

    For the simultaneous application of the clocks, it is my understanding that LRCLK and BCLK edges should be synchronized.

    Regarding the power dissipation, I believe that until the clock are applied, there is a possibility of an increase in power consumption. I would keep the clock lines supplied when power is supplied even when not reading/writing to the device. 

    Regards,

    Aaron Estrada