Announcement

Collapse
No announcement yet.

About getting a SATA redriver or not

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • About getting a SATA redriver or not

    Hi everyone,

    The attached screenshot shows a SATA redriver.
    It is located between the mSATA drive and the PCH.
    Pins 11,12 and 14,15 are the differential pairs coming from/to (bidirectional) the mSATA drive.
    Pins 5,4 and 2,1 are the differential pairs coming from/to (bidirectional) the PCH.

    You have noticed that pins 11,5 - 12,4 ... are on a vertical line.

    I would like to create 2 variants: the first one is defined using the SATA redriver, the second one is defined removing everything and adding jumpers between pins 11,5 - 12,4 - 14,2 - 15,1 giving 4 jumpers.
    Is it possible to do it with the shortest length ?

    I guess not, the solution would be to place the jumpers beside the chip and not in the middle.




  • #2
    Perhaps you can bridge it, but you'll have to be strategic about component placement and routing to eliminate stubs on the active circuits. For each path 2 resistor foot prints at the chip inlet and 2 resistor footprints at the outlet. Make the trace length from the resistor to the IC half the bypass path length. I haven't done this but I suspect it will work. The dual resistor steers the signal to the chip pin or to the bypass path. On the outlet the dual resistor steers the signal from the output pin to the output path or from the bypass to the output path.

    Comment


    • #3
      We often re-route signals by using resistors or capacitors - maybe that can help:https://youtu.be/2LWXKOEnIJE

      Comment


      • #4
        robertferanec , I watched the video and I fully agree with the process.
        The only question is how do you place resistors on top of each other ?
        It induces collision. Doesn't it ?

        Comment


        • #5
          robertferanec I think the best way is to place the resistors on top of each other.
          How would you deactivate the collision for those both components only ?
          Thank you.

          Comment


          • #6
            You can download the finished project here: https://www.imx6rex.com/application/...lopment-board/

            But basically, this is the rule:
            Click image for larger version

Name:	component placement exceptions.jpg
Views:	95
Size:	79.7 KB
ID:	19130

            Comment


            • #7
              I took a look at the imx6rex poject and more particularly the base board.
              You only use Thru Vias.
              Is it a choice to optimize the budget ?
              When do you decide to use blind or buried vias on a project ?

              Comment


              • #8
                Hello,

                normally blind or buried via are used when you have problem with the space. So for don't have a through hole via that occupy spce in all the layers.
                Yes it is a cost optimize.


                Comment


                • #9
                  Hello,
                  Thanks for reply.
                  For sure you'll use blind or buried vias when you face a problem with space but I think there is another question to think about.
                  When you use through vias, you make a hole in all layers and maybe in the layers you don't need.
                  It is a good question for EMI ...

                  Comment


                  • #10
                    We often use uvias and buried vias when the PCB has to be small. Big PCBs with uvias and blind vias are super expensive.

                    Comment

                    Working...
                    X