+ Post New Thread
Results 1 to 6 of 6
  1. #1
    Member level 1
    Points: 592, Level: 5

    Join Date
    Sep 2015
    Location
    Denmark
    Posts
    34
    Helped
    0 / 0
    Points
    592
    Level
    5

    gm/Id methodology for biasing transistor

    I am designing a Fully diff Folded cascode OTA and for this I need to design a wide swing bias.
    I read a lot about gm/Id methodology used to bias transistors in strong inversion for having a good design.
    But I could not get get proper rersourse to follow this method. Dose anybody have resources related to gm/Id methodology?

    Any help will be welcommed.
    Thanks in advance.

    Cheers

    •   AltAdvertisement

        
       

  2. #2
    Super Moderator
    Points: 52,015, Level: 55
    Achievements:
    7 years registered
    erikl's Avatar
    Join Date
    Sep 2008
    Location
    Germany
    Posts
    8,101
    Helped
    2674 / 2674
    Points
    52,015
    Level
    55

    Re: gm/Id methodology for biasing transistor

    Quote Originally Posted by predator89 View Post
    ... gm/Id methodology used to bias transistors in strong inversion
    It is correct to use bias transistors in strong inversion, but for strong inversion operation the gm/Id methodology isn't the correct design method: the gm/Id methodology is used for moderate or weak inversion operation.

    Use standard, i.e. strong inversion operation design methods for bias transistors: Use long transistors (L ≧ 10*Lmin) for high output resistance and low supply voltage dependency, and choose the width W for strong inversion operation for the required current and bias voltage.

    Depending on technology size and required current you often end up with W/L < 1 for low bias current.


    1 members found this post helpful.

    •   AltAdvertisement

        
       

  3. #3
    Member level 3
    Points: 835, Level: 6

    Join Date
    Sep 2014
    Location
    Guwahati, INDIA
    Posts
    63
    Helped
    4 / 4
    Points
    835
    Level
    6

    Re: gm/Id methodology for biasing transistor

    Quote Originally Posted by predator89 View Post
    I am designing a Fully diff Folded cascode OTA and for this I need to design a wide swing bias.
    I read a lot about gm/Id methodology used to bias transistors in strong inversion for having a good design.
    But I could not get get proper rersourse to follow this method. Dose anybody have resources related to gm/Id methodology?

    Any help will be welcommed.
    Thanks in advance.


    Cheers


    Boris murman has given beautiful notes to understand Gm/Id method. See below in chapter-5, gm/Id based design
    https://coursework.stanford.edu/acce...eader_2014.pdf





    "Take a step forward and things will start happening automatically"
    ~Rahul



    •   AltAdvertisement

        
       

  4. #4
    Member level 1
    Points: 592, Level: 5

    Join Date
    Sep 2015
    Location
    Denmark
    Posts
    34
    Helped
    0 / 0
    Points
    592
    Level
    5

    Re: gm/Id methodology for biasing transistor

    Quote Originally Posted by Rahul Sharma View Post
    Boris murman has given beautiful notes to understand Gm/Id method. See below in chapter-5, gm/Id based design
    https://coursework.stanford.edu/acce...eader_2014.pdf
    Thanks!! These are very nice notes for gm/Id method.

    But I am stuck at plotting graph for gm/id in Cadence spectre. I am not able to understand how to plot gm during DC sweep. Please guide me with this.
    Last edited by erikl; 26th July 2016 at 18:31. Reason: forgotten end [/QUOTE] added



  5. #5
    Member level 3
    Points: 835, Level: 6

    Join Date
    Sep 2014
    Location
    Guwahati, INDIA
    Posts
    63
    Helped
    4 / 4
    Points
    835
    Level
    6

    Re: gm/Id methodology for biasing transistor

    Quote Originally Posted by predator89 View Post
    Thanks!! These are very nice notes for gm/Id method.

    But I am stuck at plotting graph for gm/id in Cadence spectre. I am not able to understand how to plot gm during DC sweep. Please guide me with this.

    Hi

    See if you are taking to plot the gm/Id graph with respect to Vgs sweep then it is quit obvious to plot using calculator functions.
    -First, you will plot Id Vs Vgs (u will get one plot say plot-1)
    -Second, Differentiate Id with respect to Vgs using calculator (u will get another plot, say plot-2)
    -third, divide plot-2 with plot-1 using calculator and the final result will be nothing but gm/Id Vs Vgs

    ps- This may be the solution if i understood your question right.


    -Rahul
    Last edited by erikl; 26th July 2016 at 18:47. Reason: wrong quote mark removed


    1 members found this post helpful.

    •   AltAdvertisement

        
       

  6. #6
    Member level 1
    Points: 592, Level: 5

    Join Date
    Sep 2015
    Location
    Denmark
    Posts
    34
    Helped
    0 / 0
    Points
    592
    Level
    5

    Re: gm/Id methodology for biasing transistor

    Quote Originally Posted by Rahul Sharma View Post
    See if you are taking to plot the gm/Id graph with respect to Vgs sweep then it is quit obvious to plot using calculator functions.
    -First, you will plot Id Vs Vgs (u will get one plot say plot-1)
    -Second, Differentiate Id with respect to Vgs using calculator (u will get another plot, say plot-2)
    -third, divide plot-2 with plot-1 using calculator and the final result will be nothing but gm/Id Vs Vgs

    -Rahul
    Yes this was my question. I am new to cadence and not used the calculator functions much. I will try this out.

    Thanks a lot!!
    Last edited by erikl; 26th July 2016 at 18:45. Reason: wrong quote mark removed



--[[ ]]--