What is Convergence? — BUders Math Concepts

Simple Definition

An infinite series $\displaystyle\sum_{n=1}^{\infty} a_n$ is built by adding up infinitely many terms. Convergence means the running total gets closer and closer to some finite number $L$ — and stays there. Divergence means it doesn't: the sum grows without bound, oscillates, or otherwise fails to settle.

Formally, we define the $N$-th partial sum $S_N = a_1 + a_2 + \cdots + a_N$ and say the series converges if $\lim_{N \to \infty} S_N = L$ for some finite $L$.

Convergence — formal definition

$$\sum_{n=1}^{\infty} a_n = L \quad \iff \quad \lim_{N \to \infty} S_N = \lim_{N \to \infty} \sum_{n=1}^{N} a_n = L \in \mathbb{R}$$

✅ Convergent

Partial sums approach a finite limit

The series has a definite, finite value. Adding more terms barely changes the total.

$\displaystyle\sum_{n=1}^{\infty} \frac{1}{2^n} = \frac{1}{2} + \frac{1}{4} + \frac{1}{8} + \cdots = 1$

❌ Divergent

Partial sums grow, oscillate, or wander

No finite limit exists. The series cannot be assigned a meaningful finite sum.

$\displaystyle\sum_{n=1}^{\infty} \frac{1}{n} = 1 + \frac{1}{2} + \frac{1}{3} + \cdots = \infty$

Intuition

🏹

Zeno's arrow

To cross a room, you first cross half of it, then half of what remains, then half again — forever. Yet you reach the other side. That's convergence: infinitely many steps, finite total distance. The series $\frac{1}{2} + \frac{1}{4} + \frac{1}{8} + \cdots = 1$ is exactly this.

🪣

The leaky bucket

Adding terms to a divergent series is like pouring water into a bucket with no bottom. The harmonic series $\sum \frac{1}{n}$ looks like it should converge — terms get smaller and smaller — but it doesn't. The terms shrink too slowly to stop the total from growing forever. Slow decay is not enough; the terms must shrink fast enough.

The p-Series Benchmark

The most important reference series. Everything else gets compared to it.

p-Series

$$\sum_{n=1}^{\infty} \frac{1}{n^p} \quad \begin{cases} \text{converges} & \text{if } p > 1 \\ \text{diverges} & \text{if } p \le 1 \end{cases}$$

$p = \frac{1}{2}$

diverges

$p = 1$

diverges (harmonic)

$p = \frac{3}{2}$

converges

$p = 2$

converges to $\pi^2/6$

$p = 4$

converges fast

Convergence Tests

No single test works for every series. Here are the main tools, in roughly the order you should try them:

1 — Divergence Test (try first, always)

$\displaystyle\lim_{n \to \infty} a_n \ne 0 \implies \sum a_n$ diverges

If terms don't go to zero, the series must diverge. But if they do go to zero, this test is inconclusive — you need another test.

2 — Geometric Series

$\displaystyle\sum_{n=0}^{\infty} ar^n = \frac{a}{1-r}$ if $|r| < 1$; diverges if $|r| \ge 1$

Recognise a constant ratio between successive terms. The only test that also gives the exact sum.

3 — Integral Test

If $f$ is continuous, positive, decreasing, and $a_n = f(n)$: $\displaystyle\sum a_n$ and $\displaystyle\int_1^\infty f(x)\,dx$ both converge or both diverge.

Best for terms involving $\ln n$, $n^p$, or other functions that integrate cleanly. Proved the p-series result above.

4 — Comparison Test

If $0 \le a_n \le b_n$: $\sum b_n$ converges $\implies \sum a_n$ converges. $\sum a_n$ diverges $\implies \sum b_n$ diverges.

Compare to a p-series or geometric series you already know. Pick a comparison that has the same dominant behaviour for large $n$.

5 — Limit Comparison Test

$\displaystyle L = \lim_{n\to\infty}\frac{a_n}{b_n},\quad 0 < L < \infty \implies \sum a_n$ and $\sum b_n$ behave the same.

More forgiving than the direct comparison — you only need the ratio to approach a positive finite number. Great for rational functions of $n$.

6 — Ratio Test

$\displaystyle L = \lim_{n\to\infty}\left|\frac{a_{n+1}}{a_n}\right|$: converges if $L < 1$, diverges if $L > 1$, inconclusive if $L = 1$.

Best when terms contain factorials or exponentials. Useless for p-series (always gives $L = 1$).

7 — Root Test

$\displaystyle L = \lim_{n\to\infty}\sqrt[n]{|a_n|}$: converges if $L < 1$, diverges if $L > 1$, inconclusive if $L = 1$.

Best when $a_n$ is itself a power of $n$ (like $\left(\frac{n}{2n+1}\right)^n$). Often equivalent to the Ratio Test but simpler for those forms.

8 — Alternating Series Test

$\sum (-1)^n b_n$ converges if: (i) $b_n > 0$, (ii) $b_n$ is decreasing, (iii) $\lim_{n\to\infty} b_n = 0$.

Only applies to alternating series. Also gives an error bound: the error in stopping at $N$ terms is $\le b_{N+1}$.

Quick Examples

Ratio Test — factorial

$\displaystyle\sum_{n=0}^{\infty} \frac{n!}{3^n}$

$\left|\dfrac{a_{n+1}}{a_n}\right| = \dfrac{(n+1)!}{3^{n+1}} \cdot \dfrac{3^n}{n!} = \dfrac{n+1}{3} \to \infty$

❌ $L = \infty > 1$ — diverges by Ratio Test.

Ratio Test — exponential

$\displaystyle\sum_{n=0}^{\infty} \frac{2^n}{n!}$

$\left|\dfrac{a_{n+1}}{a_n}\right| = \dfrac{2^{n+1}}{(n+1)!} \cdot \dfrac{n!}{2^n} = \dfrac{2}{n+1} \to 0$

✅ $L = 0 < 1$ — converges. (Partial sums approach $e^2$.)

Limit Comparison — rational

$\displaystyle\sum_{n=1}^{\infty} \frac{3n^2+1}{n^4-2}$

Compare to $\dfrac{1}{n^2}$: $\lim \dfrac{a_n}{1/n^2} = \lim \dfrac{3n^4+n^2}{n^4-2} = 3$

✅ $L = 3 > 0$, and $\sum 1/n^2$ converges ($p=2$) — so this converges too.

Alternating Series Test

$\displaystyle\sum_{n=1}^{\infty} \frac{(-1)^n}{\sqrt{n}}$

$b_n = \tfrac{1}{\sqrt{n}} > 0$, decreasing, $\to 0$

✅ All three conditions met — converges by AST. (Conditionally, not absolutely.)

Absolute vs Conditional Convergence

A series $\sum a_n$ is absolutely convergent if $\sum |a_n|$ converges. Absolute convergence implies convergence, and the sum is immune to rearrangement.

A series is conditionally convergent if $\sum a_n$ converges but $\sum |a_n|$ diverges. Rearranging terms of a conditionally convergent series can change its sum to anything — or make it diverge (Riemann rearrangement theorem).

Hierarchy

$$\text{absolutely convergent} \implies \text{convergent} \implies \text{not divergent}$$

Classic example of conditional convergence

$$\sum_{n=1}^{\infty} \frac{(-1)^{n+1}}{n} = 1 - \frac{1}{2} + \frac{1}{3} - \frac{1}{4} + \cdots = \ln 2 \quad \text{(converges)}$$ $$\sum_{n=1}^{\infty} \frac{1}{n} = \infty \quad \text{(diverges — so not absolutely convergent)}$$

Common Mistakes

✗ "Terms go to zero, so the series converges"

The divergence test only works one way. $a_n \to 0$ is necessary but not sufficient. The harmonic series $\sum 1/n$ is the classic counterexample — terms go to zero, but the series diverges.

✗ Using the Ratio Test on p-series

The Ratio Test always gives $L = 1$ for p-series, making it inconclusive. Use the p-series rule directly, or the Integral Test instead.

✗ Applying Comparison Test in the wrong direction

If you want to show divergence, you need a smaller divergent series ($a_n \ge b_n$ where $\sum b_n$ diverges). Showing $a_n \le b_n$ for a divergent $b_n$ tells you nothing about $a_n$.

✗ Forgetting to check all three AST conditions

The Alternating Series Test requires positivity, monotone decrease, and limit zero. Missing even one condition means the test doesn't apply — use another test or show divergence directly.

Related Concepts

Convergence?

Ready to test your knowledge?