Convergence of a series, lots of logs

  • Thread starter Thread starter sillyus sodus
  • Start date Start date
  • Tags Tags
    Convergence Series
Join the discussion
Registration is free. Start your own thread to ask a follow-up.
1 reply · 2K views
sillyus sodus
Messages
4
Reaction score
0

Homework Statement


Show that this sum is convergent if and only if k>1 :

[tex]\sum_{n=1}^{\infty}\frac{1}{n\log{n}\left(\log{\log{n}}\right)^k}[/tex]

The Attempt at a Solution


I've applied the Cauchy condensation test, giving:

[tex]\frac{2^n}{2^n\log{2^n}\left(\log{\log{2^n}}\right)^k}[/tex]
[tex]=\frac{1}{n\log{2}\left(\log{n\log{2}}\right)^k}[/tex]

Then I figured I just apply it again?

[tex]\frac{2^n}{2^n\log{2}\left(\log{\left(2^n\log{2}\right)}\right)^k}[/tex]
[tex]=\frac{1}{\log{2}\left(n\log{2}+\log{\log{2}}\right)^k}[/tex]

But then what? I'm aiming for something along the lines of:

[tex]\frac{1}{C\cdot n^k}[/tex]

So then the answer is obvious but the best I can come up with now is that the term inside the brackets will come out like:

[tex]\left(n\log{2}+\log{\log{2}}\right)^k = n^k\log{2}^k + \mathit{terms\ of\ order\ <k}[/tex]

And therefore the n^k term will dominate the convergence behaviour... Am I on the right track here or have I missed some algebraic trick? Am I wrong to apply the test a second time? I don't want an answer here if that's possible, just some kind of hint. Thanks in advance.

PS for some reason my tex has failed somewhere but you should get the idea...
 
Last edited:
Physics news on Phys.org