• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 33 subscribers
  • Views 496 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

SN75LVCP601: M.2 SATA redriver will not connect

Peter Hendriks1
Prodigy 70 points
Part Number: SN75LVCP601

We are trying to use the SN75LVCP601 on a M.2 card in a INTEL NUC computer to communicate with an external SSD.

We are able to connect one of your SN75LVCP601EVM demo boards from the INTEL NUC M.2 port as well as from the internal SATA connector to the SSD, with several different  switch settings, but have not been able to get our board to connect with the SSD.

I have attached a PDF of our schematic. Can you see any reason why this should not work?

Peter Hendriks

M2-SATA-REDRIVER.pdf

  • 0
    TI__Mastermind 43795 points

    Is the SSD a SATA drive or a PCIe drive?  The SN75LVCP601 is for SATA only.

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Prodigy 70 points

    Lee,

    It is a SATA drive.

    As mentioned we have it working with the evaluation board.

    It works with several configurations of the switches. Yet our circuit on the M.2 board has not shown any response yet with several of those setting that work with the evaluation board.

    Peter

  • 0 in reply to Peter Hendriks1
    TI__Mastermind 43795 points

    Hi Peter,

    If the SATA drive is not even coming up a Gen1 SATA it is likely there is some configuration issue with OOB.

    Please attach a schematic of your circuit.  You should check to ensure power to the device and AC coupling on all highspeed signals.

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Prodigy 70 points

    Lee,

    The schematic of our design was posted with my initial posting. At the top of this chain.

    I did check the 3.3V on the chip.

    I did see some comments on a minimum trace-length.

    The traces on my board are really short.

    See the attached PDF.

    M2-SATA-PCB.pdf

  • 0 in reply to Peter Hendriks1
    TI__Mastermind 43795 points

    Hi Peter,

    Does this board have just 2 layers?  I don't see any GND vias at the connectors or under the LVCP601 on the thermal pad.

    Thanks and Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Prodigy 70 points

    Lee,

    This is a 4 layer PCB 0.031" thick.

    Layer 2 is GND plane, Layer 3 is 3.3V

    There is a gnd pad under the chip with 4 vias connected into the internal gnd plane.

    Signal trace are designed for 90 Ohm differential with gnd plane reference and shield vias along the traces.

  • 0 in reply to Peter Hendriks1
    TI__Mastermind 43795 points

    Hi,

    That sounds good.  You could also try to de-populate a LVCP601 and solder a short 100ohm cable segment (or wires) across the high speed pads.  Just to see if the drive will respond without the device in the path.

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Prodigy 70 points

    Lee,

    Removing the chip is a problem for us. We don't have the tools to do that at this moment.

    I can't get into the reasons why we are doing this, but what we try to do is as follows:

    We are trying to boot our NUC from a remote drive.

    This works with a the drive connected with a 6 ft. SATA to eSATA cable from either the SATA port on the NUC or through a M.2 SATA adapter in the M.2 slot on the NUC.

    However in the real application we have another panel to go through with another connector pair. This makes the our boot fail.

    We tried the re-driver demo board and found about a dozen settings that will boot from the drive and when we run a speed test program we get the SATA3 speeds.

    This works with the demo board connected to the SATA port on the NUC and through the M.2 adapter board.

    However when we try our board with the re-driver in the M.2 slot with any of the known working settings from the demo board, all we get is the message "Could not detect a bootable drive"

    As you can see on our board layout, the traces from the DC blocking caps to the chip are really short.

    I saw some previous questions on the forum about trace length. The spec sheet talks about 3" to 48", but it doesn't really give a minimum length needed.

    The trace length from the caps to the chip on the demo board is more like 1". Ours is more like 0.1". Could this be an issue?

    Peter

     

  • 0 in reply to Peter Hendriks1
    TI__Mastermind 43795 points

    Hi Peter,

    The short distance to the capacitors could start to be an issue at 6 Gbps - it would depend on the board stackup.  I do not think it is the root cause of the issue you are seeing - not able to detect the drive.  The SATA drive requires OOB signaling to get to configuration - if this is somehow distorted or disrupted the drive will be undetectable by the host.  Putting wires in place of the SN75LVCP601 is one method to help identify the root cause.  If you have access to an oscilloscope to measure the input and output signals, that can also be very helpful.

    I looked at your schematic and connectors - I believe all the high speed signals are connected correctly.

    Having a longer trace length helps to attenuate any reflections at the SN75LVCP601 and also provide some signal loss so the internal equalization does not over-equalize the input signal and add jitter to it.

    Regards,

    Lee