1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Differentiation check

  1. Aug 18, 2009 #1
    1. The problem statement, all variables and given/known data

    By using cos and sin subs for tan and sec, find the gradient of:


    2. Relevant equations



    3. The attempt at a solution



    Using the chain rule let:

    [tex]z= \frac{sin2x}{cos2x}+\frac{1}{cosx}[/tex]



    looking at z use the quotient rulee for the first part sin2x/cos2x let:

    [tex]u=sin2x[/tex] [tex]\frac{du}{dx}=2cos2x[/tex]

    [tex]v=cos2x[/tex] [tex]\frac{dv}{dx}=-2sin2x[/tex]





    As [tex]sin^2x+cos^2x=1[/tex] therefore:


    Going back to the now partionally differentated value of z


    if [tex]y=cos^{-1}x[/tex] then [tex]x=cosx[/tex]

    so [tex]\frac{dx}{dy}=-siny[/tex] so [tex]\frac{dy}{dx}=\frac{-1}{sin y}[/tex]

    since - [tex]cos^2y+sin^2y=1[/tex] then [tex]sin^2y=1-cos^2y=1-x^2[/tex]

    meaning [tex]siny=\sqrt{1-x^2}[/tex] and [tex]\frac{d}{dx}{cos^{-1}}=\frac{-1}{\sqrt{1-x^2}}[/tex]

    That menas that[tex]\frac{dz}{dx}=\frac{2}{(cos^22x)}-\frac{1}{\sqrt{1-x^2}}[/tex]

    Back to the fist chain rule:


    Sorry for the mass workings but I've been struggling with this and have now had a brain failure and can't see if I am right or even how to simplify the final differential.

    Is what I have done correct?
  2. jcsd
  3. Aug 18, 2009 #2


    User Avatar
    Science Advisor

    No! 1/(a+ b) is NOT equal to (1/a) + (1/b).

  4. Aug 18, 2009 #3
    Thanks for the input. Have I started off using the method by using the chain rule?

    Do I just leave it as:


  5. Aug 18, 2009 #4
    You have another mistake:
    [tex]\frac{1}{cosx} \neq cos^{-1}x = arccosx[/tex]
    You can look here on how to differentiate something like 1/f(x) here
    Look at the last example on the page :wink:

    One more note: it would have been much easier to differentiate tan2x using u=2x than changing it to sin2x/cos2x, although you did get that part correct.

    That, or put it in the denominator as /(tan2x + secx), or find a common denominator and add them, then take the reciprocal to get rid of the -1.
  6. Aug 22, 2009 #5
    I'm still not 100% with this question.

    I have been re-working on it and have now come up with this answer.


    Does this look right?
  7. Aug 24, 2009 #6
    Where did that last 2 come from in the numerator at the end? If you take that out, you have the correct derivative of ln(tan2x + secx).
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook