• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 97 subscribers
  • Views 1499 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.

AM5728: USB 3.0 PLL spread spectrum setup

Gunter Schmer
Genius 14442 points

Part Number: AM5728

Hi,

A customer has been testing the USB 3.0 SuperSpeed channel on their AM5728 design.  As a result of TX testing, we discovered that the spread spectrum clock does not conform to the USB 3.0 standard.  They found two threads on the E2E Sirara Forum that describe this exact issue, and seems like the driver software (/drivers/phy/phy-ti-pipe3.c) does not properly configure the spread spectrum portion of the DPLL.

 

https://e2e.ti.com/support/arm/sitara_arm/f/791/t/552051?AM572x-USB3-0-Spread-Spectrum-Clocking

https://e2e.ti.com/support/arm/sitara_arm/f/791/t/594454?AM5728-USB3-0-spread-spectrum-clocking

 

During RX testing, we have discovered that the receiver fails for Bit Error Rate (BER) at the 10 MHz, 19% sinusoidal noise mode.  Further, the bus fails to sync at all at 20 Mhz, 33 MHz and 50 MHz.  While investigating the problem, they discovered that the RX equalizer is completely disabled.  Again, this is a mechanism that should be configured by the device driver.

 Given both of these issues, we are concerned that TI has not fully verified the functionality the USB 3.0 PHY.  Is the AM5728 USB 3.0 complaint?  If so, what do we need to resolve these compliance issues?

Thanks,

--Gunter

  • 0
    TI__Mastermind 27890 points
    Hello,
    We are looking into this.
  • 0
    TI__Expert 4005 points

    We run compliance testing per the USB compliance spec and the device passes.   I've attached a report of our USB3 TX compliance testing.   We will work with the Linux driver team to get the correct settings in the driver. 

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/791/2502.Vayu_5F00_USB3_5F00_TX.mht

  • 0 in reply to Y. Yang
    TI__Genius 14442 points

    Hi,

    thanks for the USB3.0 TX report. However the customer is seeing the failures actually on the USB3.0 RX side.

    The customer has a test report they could share internally, which shows that the USB3.0 RX does not appear to be compliant for additive sinusoidal noise, in this case at or above 10MHz.

    Do we have a USB3.0 RX report we can share?

    Thanks,

    --Gunter

  • 0 in reply to Gunter Schmer
    TI__Expert 4005 points

    I've attached a RX compliance report that we performed.  Hope this helps!

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/791/Vayu_5F00_USB3_5F00_RX.mht

  • 0 in reply to Y. Yang
    TI__Genius 14442 points

    Hi Yining,

    thanks for the RX test report. It is really helpful. The customer thinks that possibly the PHY configuration used for our RX compliance test was different. The customer is using Linux with kernel aligned with our TI Processor Linux SDK and the driver code that comes with it.

    What was the SW used for our TX compliance test?

    Is there a known to work USB3 PHY configuration we could compare against from the RX compliance test?

    The suspicion is that the PHY RX Equalizer may not be enabled or configured when used with USB3.

    Regards,

    --Gunter