1
$\begingroup$

I have to determine for which values of $\alpha$ the series

$\sum\limits_{n=0}^{\infty} \frac{\log(n+4n^\alpha)-\log(n)}{2n}$

converges. First of all I simplified the expression to

$\sum\limits_{n=0}^{\infty} \frac{\log(1+4n^{\alpha-1})}{2n}$

Since the terms are definitely positive I used the ratio test, so I studied the limit

$\lim\limits_{n \to +\infty} \frac{n\log\left(1+4(n+1)^{\alpha-1}\right)}{(n+1) \log(1+4n^{\alpha -1 })}$

and I found out it equals to zero for any $\alpha$, so it should converge for any $\alpha$. However the solution of the problem is $\alpha<1$. What am I getting wrong?

$\endgroup$
  • $\begingroup$ I can be wrong, but ratio test says that if the fraction of a(n+1) / a(n) from some N is less then q (where 0 < q < 1), then the sum absolutely converges. If you plug in 1 for alpha, the limit will be 1 and we cannot say anything about convergence or divergence. If alpha > 1, limit is infinity, so the answer is alpha < 1 $\endgroup$ – Kirill Korolev May 31 at 17:47
  • $\begingroup$ Stefan, I think you miss 4 in the last limit. $\endgroup$ – richrow May 31 at 18:12
  • $\begingroup$ I corrected it. $\endgroup$ – Stefan Jun 1 at 7:54
2
$\begingroup$

Solution of the problem.

Denote $a_n=\dfrac{\log(n+4n^{\alpha})-\log(n)}{2n}=\dfrac{\log(1+4n^{\alpha-1})}{2n}$. Note that for all $n\in\mathbb{N}$ we have $a_n>0$. Now, consider two cases:

Case 1. $\alpha\geq 1$. In this case we have $\log(1+4n^{\alpha-1})\geq\log 5$ for $n\in\mathbb{N}$, so $a_n\geq \dfrac{\log 5}{2n}$. Hence, series $\sum\limits_{n=1}^{\infty}a_n$ diverges because harmonic series $\sum\limits_{n=1}^{\infty}\frac{1}{n}$ also diverges.

Case 2. $\alpha<1$. In this case $n^{\alpha-1}\rightarrow 0$ when $n\rightarrow 0$, so from the equivalence $\log(1+t)\sim t$ when $t\rightarrow 0$ we obtain $$ a_n=\frac{\log(1+4n^{\alpha-1})}{2n}\sim\frac{4n^{\alpha-1}}{2n}=\frac{2}{n^{2-\alpha}},~ n\rightarrow \infty. $$ Note that $2-\alpha>1$, so series $\sum\limits_{n=1}^{\infty}a_n$ converges because series $\sum\limits_{n=1}^{\infty}\dfrac{1}{n^p}$ converges whenever $p>1$.

Summing up two cases we get that series $\sum\limits_{n=1}^{\infty}a_n$ converges iff $\alpha<1$.

Remark. Notice that $\frac{a_{n+1}}{a_n}\rightarrow 1$ when $n\rightarrow\infty$, so I think that there is a mistake in your calculations.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.