• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 33 subscribers
  • Views 281 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.

DP83822I: Placement of 50 ohm resistors and if the extra caps are necessary

Dean Hofstetter
Expert 2585 points
Part Number: DP83822I

Tool/software:

The DP83822I datasheet says that the 49.9 ohm resistors and capacitors must be placed close to the device.  I've seen other PHYTER datasheets mention the same thing, so I'm 99% sure this is correct.  However, when I look at the layout on the DP83822 EVM board, those 49.9 ohm resistors are placed right next to the magnetics.  Is the EVM layout correct or the recommendation in the datasheet?  I believe the datasheet is correct, please confirm.

The DP83822I datasheet also has 1uF and 0.1uF caps on the AVD supply that feed the 49.9 ohm resistors.  The EVM doesn't have these caps.  The power supply recommendation (section 10 in the datasheet) already calls out for 1uF and 0.1uF caps (in addition to others).  It also recommends those caps be as close to the DP83822 as possible.  Since AVD is pin 14, and the TP+/TP-/RD+/RD- pins are right next door, I don't see how the caps recommended in the datasheet (section 9.2.1) will make any difference.  Is the EVM layout correct or the recommendation in the datasheet?  In this case, I think the EVM is correct.  Please confirm.

Thanks, Dean

  • +1
    TI__Mastermind 24988 points

    Hi Dean,

    EVM layout places these resistors as close as possible while keeping the MDI traces symmetrical and length-matched. Datasheet recommendation is correct, however please consider MDI trace layout as a higher priority when placing these resistors.

    This layout checklist has details on MDI trace layout recommendation:
    https://www.ti.com/lit/zip/snlr048 

    Dean Hofstetter said:
    The DP83822I datasheet also has 1uF and 0.1uF caps on the AVD supply that feed the 49.9 ohm resistors.

    These decoupling caps are not a strict requirement. If you expect noise on the power supply rails, these caps provide more redundancy to protect MDI signal quality.

    Thank you,

    Evan

  • 0 in reply to Evan Mayhew
    Expert 2585 points

    Thanks for the confirmation that the 49.9 ohm resistors should be placed as close as possible to the PHY, but not in a way that violates MDI trace layout.  That makes sense.  Thanks for also clearing up the extra 1uF and 0.1uF caps around the 49.9 ohm resistors are redundant, optional, your mileage will vary parts.  We are using 50V MLCC caps for the (4) recommended decoupling caps on AVD, and have the optional ferrite bead installed.  With the ferrite bead, the fact AVD (pin 14) is close to the MDI pins, and the 50V caps won't suffer hardly any DC bias derating, I think our design can get rid of these redundant caps.

    There is one other question regarding the AVD rail.  We have the optional ferrite bead in our design, but placement of the bead is up for debate.  Since AVD must connect to pin 14 on the PHY, and the (4) decoupling caps next to the PHY, it would make sense for the ferrite bead to be close to the PHY.  However, AVD also connects to the center taps on the RJ45 (we use a jack with integrated magnetics) and a few caps that are also placed close to the RJ45.  So it would make sense for the ferrite bead to be close to the RJ45.  Finally, we have this recommendation from Pulse, that says to reduce EMI you should keep a 1 inch separation between the RJ45 and PHY.   https://www.pulseelectronics.com/wp-content/uploads/2020/12/Pulse_Layout-Considerations-v7.pdf

    Our board has just over 1 inch separation between the RJ45 and PHY.  Where would you recommend placing the ferrite bead?  Next to the PHY, next to the RJ45, halfway point between PHY and RJ45, use two ferrite beads?

    Thanks, Dean

  • 0 in reply to Dean Hofstetter
    Expert 2585 points

    Do you have any recommendation for placement of the ferrite bead on the AVD rail when the PHY and RJ45 are separated by a significant amount (in our case about 1 inch)?  See above post.

    Thanks, Dean

  • 0 in reply to Dean Hofstetter
    TI__Mastermind 24988 points

    Hi Dean,

    Sorry for the delay here.

    Ferrite bead should be placed closer to the PHY in this case. If significant ESD noise is expected in the application, a second Ferrite bead can be used closer to magnetics as well.

    Thank you,

    Evan