• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 21 subscribers
  • Views 190 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.

TXV0106: Testing Issue

Jeff YANG
Mastermind 7865 points
Part Number: TXV0106

Tool/software:

Hi,

We created a small board to stack on top of the original daughter card to validate the TXV0106 (as shown in the image below). The circuit is straightforward, as below figure, we have use 4 TXV0106, I will pass full schematic via privacy message.

We are currently facing an issue:
When the TXCLK on side A (input) of the TXV0106 is 125MHz, the TXCLK on side B (output) (refer to the attached TXCLK.BMP) is also 125MHz. However, the small board's GND (GND.BMP) and POWER ( VCCO_XPI.BMP) become abnormal, as shown in the attached files (VCC_XPIO.bmp and GND.bmp), which in turn affects all other signals (e.g., RXCLK_OUTPUT.BMP).

When the TXCLK is reduced to 2.5MHz, all signals return to normal (refer to RXCLK_Nor.BMP).
Or, if I disconnect the TXCLK output on side B of the small board (i.e., not connected to the PHY on the daughter card, but TXCLK still outputs), the issue does not occur.

1.VCC_XPIO

2. TXCLK

3. RXCLK_OUTPUT

4. RXCLK_Nor

5.GND

Do you have any suggestions on how to resolve this problem?

Thanks!

Jeff

  • 0
    Guru 316960 points

    Switching at 125 MHz generates many current spikes on the supply lines. I'd guess you have to add decoupling capacitors to the small board.

    The ringing on the data lines might be a measurement artifact. For such high frequencies, you should use a ground spring instead of a ground clip for the oscilloscope probe; this requires adding nearby ground contacts to the small board.

  • 0 in reply to Clemens Ladisch
    Mastermind 7865 points

    Hi Clemens,

    On my original small board, each TXV0106 already has two 1uF MLCC capacitors on both the A side and B side power supplies. Today, I added an additional 10uF MLCC as an experiment, but the results remained the same. I also changed the measurement method to use a ground spring, but the outcome was still identical.

    Thanks!

    Jeff

  • 0 in reply to Jeff YANG
    Guru 316960 points

    How exactly are you measuring these voltages? What is the reference for the GND waveform?

    I suspect that the waveforms would look better when measured against the TXV's GND pin, and that these problems are caused by the small board's power rails not having a low-impedance connection to the rest of the circuit.

  • 0 in reply to Clemens Ladisch
    TI__Mastermind 22840 points

    Since this thread is almost a month old with no response, I will be closing it.

    Regards,

    Josh