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.

TUSB564: There is abnormality in the output LFPS signal.

Part Number: TUSB564

Tool/software:

Hi all,

① When the test item related to LFPS of USB3.2 Tx Compliance Test is performed, it fails with “5G LFPS Peak-Peak Differential Output Voltage.”

② When checking the LFPS signal input to TUSB564 and the LFPS signal output from TUSB564 with an oscilloscope, there is no abnormality in the input LFPS signal, but there is abnormality in the output LFPS signal (the

amplitude of the first stage decreases).

(The waveform is probed by the terminals of C1/C9 and U2 of EVM. A personal computer is connected to P1 and a USB3 camera of our company is connected to P2.)

ch1: LFPS signal input to TUSB564, ch2: LFPS signal output from TUSB564

■About EVM settings
① Operating mode of TUSB564
GPIO mode (short pins 3 and 6 of JMP7 so that I2C_EN is directly connected to GND)
② EN
Short pins 2 and 1 of J4 so that they are pulled up by 1064_VCC
③ CTL0/CTL1
USB3.1 mode (CTL0=H, CTL1=L)
Short pins 2 and 1 of J2, and short pins 2 and 3 of J3

■Others
This may not be related to this issue, but
As shown below, the LFPS signal is output from the redriver approximately 8 ms after the LFPS signal is input to the redriver.

The device is different, but the contents are similar to the following.

(+) TUSB1046A-DCI: Compliance Test Fail - Interface forum - Interface - TI E2E support forums

The LFPS amplitude of the host output is not sufficient, the AC coupling value may be noy appropriate, etc.

I would appreciate any comments.

Best Regards,

Ryusuke

  • Ryusuke-san

    For the failed 5G LFPS Peak-Peak Differential Output Voltage, what is the measured LFPS output voltage?

    Can you please insert the test fixture into the EVM and then start the LFPS test, would that help?

    Thanks

    David

  • Hi David,

    Vpp values are as follows:

    a: 329 mV

    b: 938 mV

    c: 1016 mV

    d: 861mV (Host's output)

    I ran the LFPS Test with the test fixture connected to the EVM.

    However, the results do not improve.

    I would appreciate any comments.

    Best Regards,

    Ryusuke

  • Ryusuke-san

    The 220nF on the TX1 and TX2 are the correct AC coupling capacitance, you can also use 100nF AC coupling capacitor.

    Location (a) looks like the TUSB564 common mode voltage is not settling yet before the host sends out the LFPS. Is there a way you can delay the host from sending out the LFPS?

    Can you send out the actual compliance report? With LFPS amplitude of min 0.8V and max 1.2V, I would expect 1046mV would pass the test.

    Thanks

    David

  • Hi David,

    Thank you for the confirmation.

    Does this device have a requirement for the LFPS signal?
    For example, the Vpp requirement of TUSB564 is actually higher than 800-1200 mV defined in the USB3.2 standard.
    Based on the E2E below and the test conditions in the data sheet, I think the Vpp of LFPS needs to be 1 V.

    I will check whether the LFPS signal to be transmitted can be delayed and whether to send a compliance report.

    Best Regards,

    Ryusuke

  • Ryusuke-san

    The re-driver, such as TUSB564, has certain amount of DC gain which will boost the LFPS signal amplitude. 

    With the LFPS spec of min of 0.8V and max of 1.2V, we typically recommend the customer to set the LFPS signal amplitude from the host to be around 1V. With the 1V and the small amount of TUSB564 DC gain, the LPFS signal amplitude out of the TUSB564 will be slightly higher than 1V, but still meet the spec and able to pass the compliance.

    Thanks

    David 

  • Hi David,

    Thank you for your reply.

    About the LFPS signal

    Do you have any information on the upper and lower limits of 1.0V?

    (within 1.0V±10%, etc.)

    Best Regards,

    Ryusuke

  • Ryusuke-san

    Are you asking for the min and max of the TUSB564 DC gain?

    Thanks

    David

  • Hi David,

    When using the TUSB564 to transmit the LFPS signal,

    since the device has a certain amount of DC gain, it is recommended to set the LFPS signal amplitude from the host to about 1.0V.

    I would like to know if there is an upper or lower limit of 1.0V for the LFPS signal requirement from the host.

    (Example: within 1.0 V±10%, etc.)

    Best Regards,

    Ryusuke

  • Ryusuke-san

    There is no upper or lower limit with the input LFPS amplitude of 1V. With the LFPS spec of min of 0.8V and max of 1.2V, setting the input LFPS amplitude around 1V will still meet the spec even with the slight 1V variation and the TUSB564 DC gain.

    Thanks

    David