+ Post New Thread
Results 1 to 8 of 8
  1. #1
    Full Member level 3
    Points: 2,739, Level: 12
    gilbertomaldito's Avatar
    Join Date
    Aug 2006
    Posts
    159
    Helped
    10 / 10
    Points
    2,739
    Level
    12

    Offset in the differential output signals in an LVDS transmitter?

    Is it possible for an offset in the differential output signals to occur in an LVDS transmitter?

    •   AltAdvertisement

        
       

  2. #2
    Full Member level 4
    Points: 3,156, Level: 13
    PaloAlto's Avatar
    Join Date
    Oct 2007
    Location
    Cádiz, Spain
    Posts
    194
    Helped
    36 / 36
    Points
    3,156
    Level
    13

    Re: LVDS transmitter

    I think that the answer is yes, but you should elaborate your question a bit more so I really know what you are saying.



    •   AltAdvertisement

        
       

  3. #3
    Advanced Member level 5
    Points: 39,713, Level: 48

    Join Date
    Mar 2008
    Location
    USA
    Posts
    6,415
    Helped
    1872 / 1872
    Points
    39,713
    Level
    48

    LVDS transmitter

    Yes, but there should be balance related specs that constrain
    such offsets to a fraction of what the receiver can tolerate.



    •   AltAdvertisement

        
       

  4. #4
    Full Member level 3
    Points: 2,739, Level: 12
    gilbertomaldito's Avatar
    Join Date
    Aug 2006
    Posts
    159
    Helped
    10 / 10
    Points
    2,739
    Level
    12

    Re: LVDS transmitter

    Hi,

    The attached file is the typical LVDS transmitter circuit.
    "A" is the ideal LVDS signal at the other end of the transmission line.
    I just want to ask if is it possible that there will be an offset like what happened to "B" and "C"?
    I have doubts if offset will happen because the differential signal depends on the Resistor. Not unless the resistance varies in time, the voltage drop across the resistor still remains the same. But im not sure, perhaps there are other factors that I did not see. What do you think guys? What do you think are the factors that can make the offset possible?

    --andrew



  5. #5
    Newbie level 6
    Points: 783, Level: 6

    Join Date
    Dec 2008
    Posts
    12
    Helped
    0 / 0
    Points
    783
    Level
    6

    Re: LVDS transmitter

    If A delivers a ''01010101" pattern, I think there can't be such big offset as B or C.
    If A delivers prbs pattern, maybe there is offset like B or C.
    There are some relationship with bit rate, loading, and corss point of A and Ab.

    Quote Originally Posted by andrew_matiga
    Hi,

    The attached file is the typical LVDS transmitter circuit.
    "A" is the ideal LVDS signal at the other end of the transmission line.
    I just want to ask if is it possible that there will be an offset like what happened to "B" and "C"?
    I have doubts if offset will happen because the differential signal depends on the Resistor. Not unless the resistance varies in time, the voltage drop across the resistor still remains the same. But im not sure, perhaps there are other factors that I did not see. What do you think guys? What do you think are the factors that can make the offset possible?

    --andrew



    •   AltAdvertisement

        
       

  6. #6
    Super Moderator
    Points: 260,228, Level: 100
    Awards:
    1st Helpful Member

    Join Date
    Jan 2008
    Location
    Bochum, Germany
    Posts
    45,442
    Helped
    13826 / 13826
    Points
    260,228
    Level
    100

    LVDS transmitter

    The received voltage difference is generated by switching the driver current between both polarities. So how do you think an offset can be generated without an additional current source? It can't.



  7. #7
    Advanced Member level 5
    Points: 39,713, Level: 48

    Join Date
    Mar 2008
    Location
    USA
    Posts
    6,415
    Helped
    1872 / 1872
    Points
    39,713
    Level
    48

    LVDS transmitter

    Not all LVDS drivers are built with that specific topology. Then
    there is always the possibility of some defect that shunts
    current asymmetrically or impairs the switch devices' headroom
    required, etc. (pushing back onto the CMRR of the tail current
    sources).

    Imbalance may not be the "normal" case but real product
    always carries a tail of the abnormal, which you have to
    crop.



  8. #8
    Super Moderator
    Points: 260,228, Level: 100
    Awards:
    1st Helpful Member

    Join Date
    Jan 2008
    Location
    Bochum, Germany
    Posts
    45,442
    Helped
    13826 / 13826
    Points
    260,228
    Level
    100

    Re: LVDS transmitter

    Yes. The question can be either answered for the given simplified circuit (obviously there can't be an offset) or a real circuit. The standard LVDS specification doesn't require actually switched current sources, so it can be nearly anything.



--[[ ]]--