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: Limit and Convergence Questions

  1. Feb 27, 2010 #1
    1. The problem statement, all variables and given/known data

    What is the limit as n approaches infinity of: (4n+2^n)/(5n+3^n)


    Does the series from 1 to infinity of (1+n)/sqrt(1+n^6) converge absolutely, conditionally or diverge?

    2. Relevant equations

    L'hopital's rule, convergence tests...

    3. The attempt at a solution

    For the limit, I attempted to use l'hopital's rule but was left stuck with a 2^n*ln2 at the top and a 3^n*ln3 at the bottom.

    For the series question, I had no idea which test to use :S

    Any help would be greatly appreciated!
  2. jcsd
  3. Feb 27, 2010 #2


    User Avatar
    Science Advisor

    It shouldn't be hard to see that [itex]2^n[/itex] and [itex]3^n[/itex] will be much larger than 4n and 5n so, for very large n, this is essentially [itex]\frac{2^n}{3^n}= \left(\frac{2}{3}\right)^n[/itex] which, since [itex]\frac{2}{3}< 1[/itex], goes to 0.

    For a more rigorous proof, use L'Hopital's rule.

    Strictly speaking, L'Hopital's rule only applies to limits of functions, but since [itex]f(x_n)[/itex], as [itex]x_n\to a[/itex] must converge to [itex]\lim_{x\to a} f(x)[/itex], we can apply it to sequences as well.

    Here, the derivative of [itex]4n+ 2^n[/itex] is [itex]4+ 2^n ln(2)[/itex] and the derivative of [itex]5n+ 3^n[/itex] is [itex]4+ 3^nln(3)[/itex]. Since
    [tex]\frac{4+ 2^n ln(2)}{5+ 3^n ln(3)} [/tex]
    is still "infinity over infinity", apply L'Hopital a second time, getting
    [tex]\frac{2^n(ln(2))^2}{3^n (n(3))^2}= \frac{ln(2)}{ln(3)}\left(\frac{2}{3}\right)^n[/tex]
    which clearly converges to 0.

    As for [itex]\sum_{n=0}^\infty (1+n)/\sqrt{1+n^6}[/itex], since it involves only positive numbers, there is no question of "conditional convergence"- it either diverges or converges absolutely.

    You should be able to use the comparison test: show that this is less than some (possibly large) number times [itex]n/\sqrt{n^6}= n/n^3= 1/n^2[/itex] and since the series [itex]\sum 1/n^2[/itex] converges (by the integral test), this series converges absolutely.
    Last edited by a moderator: Feb 27, 2010
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook