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.

TLV320AIC3106: Power Supply Sequencing, Down and Dirty Method

Part Number: TLV320AIC3106
Other Parts Discussed in Thread: TLV757P

Hello, 

I'm building a TLV320AIC3106 module for my project and I'm somewhat confused about the power supply requirements. We have a lot of things to consider. 

  • IOVDD (1.8v)
  • DVDD (1.8v)
  • DRVDD (3.3v)
  • AVDD_DAC (3.3v)
  • AVDD_ADC (3.3v)

I'm trying to build this with the least possible amount of components for simple, yet effective operation. Sure, I'm trying to get as close to the theoretical SNR as possible within reason. I plan on using two LDOs.

  • TLV75733PDBVR (3.3v)
  • TLV75718PDBVR (1.8v)

The datasheet makes mention about power supply startup sequencing. Page 93 wants it in this order. 

  1. IOVDD (1.8V)
  2. AVDD, DRVDD (3.3V)
  3. DVDD (1.8V)

I kinda figured that the 1.8v LDO will naturally power up faster, and then the 3.3v LDO will power up afterwards naturally. Maybe not. Both LDOs will be connected directly to VIN. I will not cascade the power,  for example VIN -> 3.3v -> 1.8v, so that the 1.8v has a better chance of powering first. I have studied the TLV320AIC3106EVM to try and get a handle on the power supply that is built for that module. But I can't get a grasp on exactly how the power up sequencing works on that board. If we need to power DVDD last, how can I delay that startup while still using the same LDO? How many LDOs do I "really" need for powering this part? What's the recommended schematic? Can I just hook up 3.3 and 1.8v to it and call it a day!?? 

Thanks for your help!

Jay

  • Hello Jay,

    I want to mention that our devices are very tolerant when it comes to power supply sequencing and following the recommended sequencing ensures the most robust operation. While we recommend the sequence be followed, it is not a hard requirement. 

    If you would like to delay the start up, you can possibly use a bigger cap on the output of the LDO. I believe something similar is done on the EVM. We usually see customers with 3 separate LDOs and they are enabled in the sequence you mentioned above. 

    Regards,

    Aaron

  • Aaron Estrada51 said:
    We usually see customers with 3 separate LDOs and they are enabled in the sequence you mentioned above

    Thank you for your input. This confirms my theories... So the recommendation is 3 power supplies. And the recommendation is either to find a way to delay the startup of each LDO using the right spec'd capacitor, or using the enable pin on each LDO one at a time. 

    Do you expect any special care for each LDO? I've seen ferrite beads between the power input and the LDO for example. Or is running the input power supply through an LDO and then carefully placing and routing the traces enough?

    What do you think about having 3 TLV757P?

    And, just a thought. This sequence would probably do exactly what the datasheet requires. It would delay the DVDD until the very end. 

    1. VIN-> IOVDD (1.8V)
    2. VIN -> AVDD, DRVDD (3.3V)
    3. 3.3V -> DVDD (1.8V)

  • Hello Jay,

    Running the supply straight through the LDO is fine.... this is what we typically see and this is done on the EVM. The data sheet also shows that the min time from supply to supply is spec'd at 0ms. So technically, according to the data sheet, they can be powered up at the same time. 

    Regards,

    Aaron

  • Hello,

    In addition to Aaron's comments I wanted to add that we've seen some customers use load switch IC's (or discrete MOS based switches) from a single LDO output if some of the supply voltages are the same between the rails.  Which way you go probably depends on which vector out of cheaper, smaller, efficient, easier is most important.