• State TI Thinks Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 1 answer
  • Subscribers 23 subscribers
  • Views 2454 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.

CDCM6208: what's the actual model of Vcm-DC ?

Jiaxin Tong
Prodigy 60 points
Part Number: CDCM6208


Hello，

     I find a strange description about the Vcm-DC of LVPECL/CML/LVDS from datasheet of CDCM6208:

LVPECL:

CML:

LVDS:

Can you tell me the actual model of DC couple? I want to know the test DC couple schemetic of CDCM6208's output,Thank you very much.

  • 0
    TI__Guru** 111790 points
    Hi Jiaxin,

    The lead engineer is on vacation, he will get back to you later. Thank you for your patience.
  • 0
    TI__Genius 9920 points
    Hi Jiaxin,
    CDCM6208 adopt CML driver to output LVDS-like and LVPECL-like type.

    "LVDS-like: All outputs Y[7:0] support LVDS-like signaling. The actual output stage uses a CML structure and drives a signal swing identical to LVDS (~350mV)."

    "LVPECL-like: Outputs Y[3:0] support LVPECL-like signaling. The actual output stage uses a CML structure but drives the same signal amplitude and rise time as true emitter coupled logic output stages."

    So VCM-DC test condition is the same as CML driver: "DC coupled with 50 Ω on-chip termination to VDD_Yx_Yy".

    Regards,
    Shawn
  • 0 in reply to Shawn Han
    Prodigy 60 points

    Reply to your : CM-DC test condition is the same as CML driver: "DC coupled with 50 Ω on-chip termination to VDD_Yx_Yy"，

    But According to my calculations I think the real calculation is not the same as what datasheet. says

    So can you give me a detailed process of analysis in the DC couple with LVDS like and LVPECL like?

  • 0 in reply to Jiaxin Tong
    TI__Genius 9920 points

    Jianxin,

    CDCM6208 LVDS-like type is a low power CML (lpcml) driver, LVPECL-like type is a high swing CML (hscml) driver. We recommended AC coupling for LVDS/ LVPECL receiver, then the unmatched Vcm  from CDCM6208 could be ignored.

    If you really need DC couple with load, a IBIS model simulation could help you.

    CDCM6208 IBIS model

    Regards,

    Shawn

  • 0 in reply to Shawn Han
    Prodigy 60 points

    Hello,

        I think the Vcm-DC should be the open-circuit voltage, other than the coupled voltage. Because the DC--couple will add load after the CML output, then I think the Vcm-DC data is wrong.

        The second reason is if Vcm-DC in the DC-coupled is min(Vref-0.13), considering the wave swing, the max voltage is bigger than Vref, Please give me an explaintion.

  • 0 in reply to Jiaxin Tong
    TI__Genius 9920 points

    Hi Jiaxin,

    Standard CML drivers that are built from an open-drain differential pair. The outputs (Output+ and Output–) require pullup resistors to VDD because the NMOS transistor can drive only falling edges efficiently and needs the pullups to help drive rising edges.

     VDD_Yx_Yy-0.13 is for typical value.

    The spec swing Vod is measured by AC coupling,which can't be add on Vcm-DC directly.

    If you are using LVDS or LVPECL receiver, please use AC coupling.

    If you are using CML receiver, AC or DC coupling can be accepted.

    Regards,

    Shawn

  • 0 in reply to Shawn Han
    Prodigy 60 points

    I disagree with you explantion about CML output struction becasue of the DATASHEET Fig 12:

    From Fig 12, CML output matching impedance 50 ohm is in the IC, So your explatnion is wrong.

  • 0 in reply to Jiaxin Tong
    TI__Expert 4400 points

    The VCM-DC is the open circuit voltage and measurement is done at DC where Voh, Vol are measured with high impedance probe and then Vcm-dc is calculated. 

  • 0 in reply to Madhu Balasubramanian
    Prodigy 60 points

    So there are two questions which are needed to explain.

    1) Output Matching impedance is integrated in the IC or not? According to the datasheet Fig12, I think it is in the IC output struction. But considering your pre-reply, you think there is no Output Matching impedance  in the IC. So please give me a confirm.

    2) VCM-DC's measurement method is which one? The first is your explain: open circuit voltage ( with  Output Matching impedance  is Not in the IC)  The second is output struction (with  Output Matching impedance in IC) DC coupled with post-circuit.At the same time, post-circuit must be DC matching . Which one is the actual way in datasheet.

    Thank you for your reply. And I suggest TI's datasheet should be detailed in measurement method considering the majority of user.

  • 0 in reply to Jiaxin Tong
    TI__Expert 4400 points
    The output is CML based and has on-chip 50 ohm impedance. The DC measurement is done with a high impedance probe at the output pins.

    We will add this to the updates needed for the datasheet in our next update cycle.
  • 0 in reply to Madhu Balasubramanian
    TI__Expert 4400 points
    When I mentioned open circuit, there is no impedance matching external to the IC.
  • 0 in reply to Jiaxin Tong
    TI__Genius 9920 points

    You are correct. CDCM6208 output driver had integrated "50 Ω on-chip termination".

    That's why the test condition is "DC coupled with 50 Ω on-chip termination to VDD_Yx_Yy".

    For CML driver, the pull up termination is must have,  which could be on-chip (internal) or external.

    There are some confusion on "Open circuits" and "Open Drain" in this thread. The key is how to understand CML driver and receiver.

    Below 2 application notes could help understand more.

    Interfacing Between LVPECL, VML, CML and LVDS Levels

    DC-Coupling Between Differential LVPECL, LVDS, HSTL, and CML