• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 3 answers
  • Subscribers 33 subscribers
  • Views 619 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.

Original question:

USB3.0 cable with two SN65LVPE512

SN65LVPE512: USB redriver SN65LVPE512

Manoj Shenoy30
Prodigy 100 points
Part Number: SN65LVPE512
Other Parts Discussed in Thread: TUSB522P

Hi ,

I was referring the datasheet of the re driver SN65LVPE512. In that i have a doubt regarding the control settings for each channel. Do we have to set the same (OS1/2,EQ1/2,DE1/2) strap for the parameters or can I have individual settings. Block diagram says the 2 control signals goes to a single control block. So just want to know the relation between the 1 and 2 signals in controlling.

Regards

Manoj Shenoy

  • 0
    TI__Guru** 119255 points
    it should be separate for CH1 and CH2, you need to look at your layout how you connect CH1 and CH2.
  • 0
    TI__Expert 7940 points
    This device has two channels, each channel configuration is individually controlled. For example, Channel 1 receiver EQ may require more gain than Channel 2 receiver EQ since the trace length may be greater, EQ1 can be set to a different value than EQ2.

    Also, this is an older device, please consider using the TUSB522P for new designs.

    Regards,
    Undrea
  • 0 in reply to Undrea Fields
    Prodigy 100 points
    Hi Undrea,

    Thanks for the response. This I am not using for new design. I am facing some issues with existing design. Hence tuning the gain values.

    Issue : In my current design, the re driver circuit has been designed in add on card. Some of the USB stick(Sony) is working where as some not (SanDisk , Transcend). By default the OS was set to 1241mVpp and with EQ of 15dB and DE of 0dB on both the channels.
    When I changed the OS to maximum of 1324mVpp the performance is improved. But the other 2 USB Sticks are performing intermittently.

    Any suggestions on fine tuning?


    Thanks and Regards
    Manoj Shenoy
  • 0 in reply to Manoj Shenoy30
    TI__Expert 7940 points

    Hi Manoj,

    15dB EQ seems pretty high, do you know the amount of loss from the Host to redriver and from the redriver to USB connector?  I suspect the EQ value would need to be lowered to 7dB on the USB connector channel (This is typically a shorter distance than from a Host to redriver), or even both channels depending on the loss.

    Adding some de-emphasis to the output may also help, with 1324mV setting I would select the -4dB setting.  We can provide better suggestions if the loss is known (Or approximate trace lengths from Host to redriver and redriver to USB connector).

    Regards,

    Undrea

  • 0 in reply to Undrea Fields
    Prodigy 100 points

    Hi Undrea,

    Thanks for the inputs.

    Below is the architecture of my design with the approximate trace lengths.

    I tried with below Combinations. Results are mentioned along

    Comb 1:

    OS1/OS2 - 1324mV

    DE1/DE2- 0dB

    EQ1 -15dB EQ2 - 0dB

    Result : I got a decent performance when tested with 3 different pen drives and a USB HUB

    Comb 2:

    OS1/OS2 - 1324mV

    DE1/DE2- 0dB

    EQ1 -15dB EQ2 - 7dB

    Result : Pen drives are having the best performance but USB hub is failing all the time

    Comb 3:

    OS1/OS2 - 1324mV

    DE1 - 0dB DE2- -4dB

    EQ1 -15dB EQ2 - 7dB

    Result: Pen drives are having marginal pass but USB hub is failing in huge margin

    Comb 1 seems to be the best to get the decent performance. However I would like to improve. Please provide suggestions

    Thanks and Regards

    Manoj Shenoy

  • 0 in reply to Manoj Shenoy30
    TI__Expert 7940 points
    Hi Manoj,

    It seems the issue may be over-EQ for this system. I would suggest the following:
    EQ1 = 0 (7dB)
    OS2 = NC (1241mV)
    DE2 = 0 (-3dB)

    EQ2 = 0 (0dB)
    OS1 = NC (1241mV)
    DE1 = 0 (-3dB) Can also try NC (0dB)

    The loss from Host to redriver @ 2.5GHz would be ~8-11dB (Ballpark), the Host also has EQ so a setting of 7dB on the Host-side(EQ1) should be sufficient (Unless there is much more loss). The distance from redriver to connector is very short so no EQ is needed to compensate for the loss of that short trace.

    Regards,
    Undrea