Free CAT library

Advanced Logarithmic Progressions, Series & Inequalities

Master the linearization of geometric progressions into arithmetic progressions, the rapid Vieta product-of-roots shortcut for logarithmic quadratics, telescoping logarithmic series summations, and strict monotonic inequality direction flips across bases in (0, 1) and variable bases.

1. The Story & Mental Model: The Bridge Across Algebraic Worlds

The Great Linearization Engine

In algebra, addition and multiplication represent two distinct geometric universes.

  • Addition builds straight lines: y=mx+cy = mx + c. Each step is identical and steady.

  • Multiplication bends space into exponential curves: y=a⋅rxy = a \cdot r^x. Each step accelerates violently.

The logarithm is the great dimensional bridge connecting these two worlds.

When numbers multiply along a Geometric Progression (a,ar,ar2,ar3,…a, ar, ar^2, ar^3, \dots), taking their logarithm flattens the explosive exponential curve into a uniform, rhythmic march along an Arithmetic Progression:

log⁡a,log⁡a+log⁡r,log⁡a+2log⁡r,log⁡a+3log⁡r,…\log a, \quad \log a + \log r, \quad \log a + 2\log r, \quad \log a + 3\log r, \dots

By converting multiplication into addition (log⁡(xy)=log⁡x+log⁡y\log(xy) = \log x + \log y), logarithms allow mathematicians to apply elementary linear formulas (such as the sum of an AP, middle-term arithmetic averages, and linear symmetry) to intractable multiplicative problems.

The Inversion Mirror of Logarithmic Inequalities

When solving logarithmic equations, the base acts merely as a reference frame. But in inequalities, the base dictates the very geometry of the number line:

  1. The Expanding Base (a>1a > 1):

    The curve y=log⁡axy = \log_a x rises continuously from left to right (strictly increasing). Larger inputs produce larger outputs:

    x2>x1>0  ⟺  log⁡ax2>log⁡ax1x_2 > x_1 > 0 \iff \log_a x_2 > \log_a x_1

    The inequality sign preserves its original direction.

  2. The Contracting Base (0<a<10 < a < 1):

    When the base is a fraction between 00 and 11 (such as 0.50.5 or 1/31/3), each multiplication makes the number smaller. To reach a larger fraction, you must raise the base to a smaller power! The curve y=log⁡axy = \log_a x plunges downward (strictly decreasing). Larger inputs produce smaller outputs:

    x2>x1>0  ⟺  log⁡ax2<log⁡ax1x_2 > x_1 > 0 \iff \log_a x_2 < \log_a x_1

    The inequality sign flips in reverse!

Failing to see whether the base is an expander (>1> 1) or a contractor (0<a<10 < a < 1) is the most treacherous trap in advanced algebra.


2. The CAT Question Blueprint: How Exam Setters Disguise Advanced Logarithms

Blueprint 1: The GP-to-AP Conversion Disguise

Exam setters construct problems where variables appear trapped in geometric relationships:

If 1log⁡ax,1log⁡bx,1log⁡cx are in Arithmetic Progression, what progression do a,b,c form?\text{If } \frac{1}{\log_a x}, \frac{1}{\log_b x}, \frac{1}{\log_c x} \text{ are in Arithmetic Progression, what progression do } a, b, c \text{ form?}

Candidates who attempt algebraic expansions get lost in fractions.

A top percentiler applies the reciprocal inversion rule:

log⁡xa,log⁡xb,log⁡xc are in AP\log_x a, \log_x b, \log_x c \text{ are in AP}

By the definition of an AP:

2log⁡xb=log⁡xa+log⁡xc  ⟹  log⁡x(b2)=log⁡x(ac)  ⟹  b2=ac2\log_x b = \log_x a + \log_x c \implies \log_x(b^2) = \log_x(ac) \implies b^2 = ac

Therefore, a,b,ca, b, c form a Geometric Progression (GP)! The entire question resolves in two lines.

Blueprint 2: The Log-Quadratic Product of Roots Shortcut

Exam setters ask for the product of all real roots of an equation such as:

xlog⁡2x−3=14x^{\log_2 x - 3} = \frac{1}{4}

Unprepared candidates solve for x1x_1 and x2x_2 individually, which takes minutes of scratch work.

A master candidate sets y=log⁡2xy = \log_2 x. The equation becomes:

y(y−3)=−2  ⟹  y2−3y+2=0y(y - 3) = -2 \implies y^2 - 3y + 2 = 0

Let the roots in yy be y1y_1 and y2y_2. By Vieta's formulas:

y1+y2=−(−31)=3y_1 + y_2 = -\left(\frac{-3}{1}\right) = 3

Because y1=log⁡2x1y_1 = \log_2 x_1 and y2=log⁡2x2y_2 = \log_2 x_2:

log⁡2(x1x2)=log⁡2x1+log⁡2x2=y1+y2=3\log_2(x_1 x_2) = \log_2 x_1 + \log_2 x_2 = y_1 + y_2 = 3

x1x2=23=8x_1 x_2 = 2^3 = 8

The product of the roots is obtained from the sum of the roots in yy in under 1515 seconds without ever computing the roots themselves!

Blueprint 3: The Fractional-Base Inequality Double Ambush

Exam setters present inequalities with bases between 00 and 11:

log⁡0.5(x2−5x+6)≥−1\log_{0.5}(x^2 - 5x + 6) \ge -1

They set a double ambush:

  1. Ambush 1 (The Flip): Students forget to flip ≥\ge to ≤\le when eliminating the base 0.50.5.

  2. Ambush 2 (The Domain Floor): Students flip the sign (x2−5x+6≤2x^2 - 5x + 6 \le 2), but forget that the argument must remain strictly positive (x2−5x+6>0x^2 - 5x + 6 > 0).

Both traps are designed into the distractor options on exam day.

Blueprint 4: The Telescoping Rational Log Series

Exam setters present summations of the form:

S=∑k=2nlog⁡(1−1k2)S = \sum_{k=2}^n \log\left(1 - \frac{1}{k^2}\right)

Factoring 1−1k2=k2−1k2=(k−1)(k+1)k⋅k1 - \frac{1}{k^2} = \frac{k^2 - 1}{k^2} = \frac{(k - 1)(k + 1)}{k \cdot k} decomposes the product into two staggered telescoping cascades that cancel almost all intermediate terms.


3. Core Concepts, Formulae & Derivations

A. Theorem 1: Equivalence of Logarithmic AP and Multiplicative GP

Let x1,x2,x3,…,xnx_1, x_2, x_3, \dots, x_n be a sequence of positive real numbers.

{xk} is in GP with common ratio r  ⟺  {log⁡bxk} is in AP with common difference log⁡br\{x_k\} \text{ is in GP with common ratio } r \iff \{\log_b x_k\} \text{ is in AP with common difference } \log_b r

Rigorous Derivation:

Let {xk}\{x_k\} be a GP with first term a>0a > 0 and common ratio r>0r > 0. The kthk^{\text{th}} term is:

xk=a⋅rk−1x_k = a \cdot r^{k-1}

Take the logarithm with arbitrary base b>0,b≠1b > 0, b \ne 1:

log⁡bxk=log⁡b(a⋅rk−1)=log⁡ba+log⁡b(rk−1)=log⁡ba+(k−1)log⁡br\log_b x_k = \log_b(a \cdot r^{k-1}) = \log_b a + \log_b(r^{k-1}) = \log_b a + (k - 1) \log_b r

This matches the standard explicit formula for an Arithmetic Progression Tk=A+(k−1)DT_k = A + (k - 1)D, where:

  • First term: A=log⁡baA = \log_b a

  • Common difference: D=log⁡br■D = \log_b r \quad \blacksquare

B. Theorem 2: The Log-Quadratic Vieta Root-Product Shortcut

Consider a general quadratic equation in logarithmic form:

A(log⁡bx)2+B(log⁡bx)+C=0(A≠0)A(\log_b x)^2 + B(\log_b x) + C = 0 \quad (A \ne 0)

Let the real roots in xx be x1x_1 and x2x_2.

Then the product of the roots is:

x1⋅x2=b−B/Ax_1 \cdot x_2 = b^{-B/A}

Rigorous Derivation:

Substitute y=log⁡bxy = \log_b x. The equation becomes:

Ay2+By+C=0A y^2 + B y + C = 0

Let y1y_1 and y2y_2 be the roots of this quadratic. By Vieta's formulas:

y1+y2=−BAy_1 + y_2 = -\frac{B}{A}

Since y1=log⁡bx1y_1 = \log_b x_1 and y2=log⁡bx2y_2 = \log_b x_2:

log⁡bx1+log⁡bx2=−BA\log_b x_1 + \log_b x_2 = -\frac{B}{A}

Apply the product operational law on the left-hand side:

log⁡b(x1⋅x2)=−BA\log_b(x_1 \cdot x_2) = -\frac{B}{A}

Convert to exponential form:

x1⋅x2=b−B/A■x_1 \cdot x_2 = b^{-B/A} \quad \blacksquare

(Crucial Caution: Do NOT confuse y1⋅y2=C/Ay_1 \cdot y_2 = C/A with the product of xx. In logarithms, addition of powers corresponds to multiplication of numbers!)

C. Theorem 3: Logarithmic Inequalities Across Base Regimes

Let f(x)f(x) and g(x)g(x) be algebraic expressions.

\hline \textbf{Base Regime} & \textbf{Given Inequality} & \textbf{Equivalent Algebraic System} \\[4pt] \hline \mathbf{a > 1} & \log_a f(x) < \log_a g(x) & 0 < f(x) < g(x) \\[4pt] (\text{Strictly Increasing}) & \log_a f(x) > \log_a g(x) & f(x) > g(x) > 0 \\[4pt] \hline \mathbf{0 < a < 1} & \log_a f(x) < \log_a g(x) & f(x) > g(x) > 0 \; \text{\textbf{(FLIP)}} \\[4pt] (\text{Strictly Decreasing}) & \log_a f(x) > \log_a g(x) & 0 < f(x) < g(x) \; \text{\textbf{(FLIP)}} \\[4pt] \hline \end{array}$$ #### D. Theorem 4: Telescoping Logarithmic Series Summations 1. **Linear Telescoping Series:** $$\sum_{k=1}^n \log\left(1 + \frac{1}{k}\right) = \sum_{k=1}^n \log\left(\frac{k+1}{k}\right) = \log\left(\prod_{k=1}^n \frac{k+1}{k}\right) = \log(n + 1)$$ 2. **Quadratic Difference-of-Squares Telescoping Series:** $$\sum_{k=2}^n \log\left(1 - \frac{1}{k^2}\right) = \log\left(\prod_{k=2}^n \frac{(k-1)(k+1)}{k \cdot k}\right) = \log\left(\frac{n + 1}{2n}\right)$$ --- ### 4. Common Pitfalls & Speed Shortcuts #### Pitfall 1: Forgetting the Inequality Sign Flip for Fractional Bases - **The Blunder:** Solving $\log_{0.2}(x - 4) > 1$ by writing $x - 4 > (0.2)^1 \implies x > 4.2$. - **The Reality:** Because base $0.2 \in (0, 1)$, the inequality **flips**: $$x - 4 < (0.2)^1 = 0.2 \implies x < 4.2$$ And intersecting with domain $x - 4 > 0 \implies x > 4$, the true solution is $x \in (4, 4.2)$! #### Pitfall 2: Using the Product of $y$ Instead of the Sum of $y$ for $x_1 x_2$ - When solving $A y^2 + B y + C = 0$, students remember "product of roots is $C/A$" and write $x_1 x_2 = b^{C/A}$. - **Remember:** $\log(x_1 x_2) = \log x_1 + \log x_2 = y_1 + y_2$. The product of numbers in original space corresponds to the **sum of exponents**! Thus $x_1 x_2 = b^{-B/A}$. #### Speed Shortcut 1: The Vieta Product Formula in One Line Whenever asked for the product of roots of an equation of the form $x^{\log_b x + p} = q$: $$\text{Product of roots } = b^{-p}$$ You never need to compute the discriminant, factorize the quadratic, or find the individual roots! #### Speed Shortcut 2: Factoring Symmetric GP Multipliers If three numbers $x, y, z$ are in GP, always represent them symmetrically as: $$x = \frac{a}{r}, \quad y = a, \quad z = ar$$ Taking logarithms immediately yields: $$\log x = \log a - \log r, \quad \log y = \log a, \quad \log z = \log a + \log r$$ The middle term is $\log a$, and the terms form a symmetric AP with common difference $\log r$.

Formulas

GP to AP Linearization Theorem

x1,x2,…,xn∈GP  ⟺  log⁡bx1,log⁡bx2,…,log⁡bxn∈APx_1, x_2, \dots, x_n \in \text{GP} \iff \log_b x_1, \log_b x_2, \dots, \log_b x_n \in \text{AP}

Common ratio r in GP becomes common difference log_b r in AP
Vieta Root-Product Shortcut

A(log⁡bx)2+B(log⁡bx)+C=0  ⟹  x1⋅x2=b−B/AA(\log_b x)^2 + B(\log_b x) + C = 0 \implies x_1 \cdot x_2 = b^{-B/A}

Product of original roots equals base raised to negative (B/A)
Logarithmic Inequality (Base a > 1)

log⁡af(x)<log⁡ag(x)  ⟺  0<f(x)<g(x)\log_a f(x) < \log_a g(x) \iff 0 < f(x) < g(x)

Inequality direction is preserved for expander bases greater than 1
Logarithmic Inequality (Base 0 < a < 1)

log⁡af(x)<log⁡ag(x)  ⟺  f(x)>g(x)>0\log_a f(x) < \log_a g(x) \iff f(x) > g(x) > 0

Inequality direction flips for contractor bases between 0 and 1
Difference-of-Squares Telescoping Series

∑k=2nlog⁡(1−1k2)=log⁡(n+12n)\sum_{k=2}^n \log\left(1 - \frac{1}{k^2}\right) = \log\left(\frac{n+1}{2n}\right)

Factorization into (k-1)(k+1)/k^2 creates a two-way canceling cascade

Traps and tips

  • Inequality Flip for Base in (0, 1): Always flip the inequality sign whenever the base of the logarithm is strictly between 0 and 1.
  • Domain Intersection is Mandatory: In inequalities, never forget to intersect the algebraic solution with argument > 0; omitting this creates unbounded invalid intervals.
  • Vieta Sum Gives Product: In quadratic substitution y = log_b x, use y_1 + y_2 = -B/A to find the product x_1 * x_2 = b^{-B/A}.
  • Variable Base Splitting: When the base of an inequality contains a variable x, you MUST split into two distinct branches: x > 1 and 0 < x < 1.
  • Symmetric Representation in GP: Set GP terms as a/r, a, ar so that their logarithms form the symmetric AP (log a - log r, log a, log a + log r).

Solved examples

Example 5.1: Logarithms Transforming GP into Cyclic AP Symmetries

If the pthp^{\text{th}}, qthq^{\text{th}}, and rthr^{\text{th}} terms of a Geometric Progression (GP) of positive numbers are uu, vv, and ww respectively, prove and evaluate the exact value of:

E=(q−r)log⁡u+(r−p)log⁡v+(p−q)log⁡wE = (q - r) \log u + (r - p) \log v + (p - q) \log w

Show solution

Step 1: Express the terms of the GP using standard parameters.

Let the GP have first term a>0a > 0 and common ratio R>0R > 0.

The nthn^{\text{th}} term of the GP is given by Tn=aRn−1T_n = a R^{n-1}.

Therefore:

  • u=Tp=aRp−1u = T_p = a R^{p-1}

  • v=Tq=aRq−1v = T_q = a R^{q-1}

  • w=Tr=aRr−1w = T_r = a R^{r-1}


Step 2: Take logarithms to transform terms into an Arithmetic Progression.

Using the product and power operational laws:

log⁡u=log⁡(aRp−1)=log⁡a+(p−1)log⁡R\log u = \log(a R^{p-1}) = \log a + (p - 1) \log R

log⁡v=log⁡(aRq−1)=log⁡a+(q−1)log⁡R\log v = \log(a R^{q-1}) = \log a + (q - 1) \log R

log⁡w=log⁡(aRr−1)=log⁡a+(r−1)log⁡R\log w = \log(a R^{r-1}) = \log a + (r - 1) \log R


Step 3: Substitute the logarithmic expressions into EE.

E=(q−r)[log⁡a+(p−1)log⁡R]+(r−p)[log⁡a+(q−1)log⁡R]+(p−q)[log⁡a+(r−1)log⁡R]E = (q - r)[\log a + (p - 1) \log R] + (r - p)[\log a + (q - 1) \log R] + (p - q)[\log a + (r - 1) \log R]


Step 4: Collect like terms for log⁡a\log a and log⁡R\log R.

Coefficient of log⁡a\log a:

(q−r)+(r−p)+(p−q)=q−r+r−p+p−q=0(q - r) + (r - p) + (p - q) = q - r + r - p + p - q = 0

Coefficient of log⁡R\log R:

(q−r)(p−1)+(r−p)(q−1)+(p−q)(r−1)(q - r)(p - 1) + (r - p)(q - 1) + (p - q)(r - 1)

Expand each product:

  • (q−r)(p−1)=qp−q−rp+r(q - r)(p - 1) = qp - q - rp + r

  • (r−p)(q−1)=rq−r−pq+p(r - p)(q - 1) = rq - r - pq + p

  • (p−q)(r−1)=pr−p−qr+q(p - q)(r - 1) = pr - p - qr + q

Summing all three expansions:

[qp−q−rp+r]+[rq−r−pq+p]+[pr−p−qr+q][qp - q - rp + r] + [rq - r - pq + p] + [pr - p - qr + q]

Notice pairwise cancellations:

  • qp−pq=0qp - pq = 0

  • −rp+pr=0-rp + pr = 0

  • rq−qr=0rq - qr = 0

  • −q+q=0-q + q = 0

  • r−r=0r - r = 0

  • p−p=0p - p = 0

The coefficient of log⁡R\log R is identically 00.


Step 5: Final Evaluation.

E=0⋅log⁡a+0⋅log⁡R=0E = 0 \cdot \log a + 0 \cdot \log R = 0

The exact value of EE is 00.

Takeaway: Logarithms map geometric terms T n = a R^n-1 into linear AP expressions log a + (n-1) log R; cyclic weighted sums across indices cancel identically to 0.

Example 5.2: Product of Roots Shortcut in a Logarithmic Quadratic

Find the product of all real roots of the equation:

xlog⁡2x−3=14x^{\log_2 x - 3} = \frac{1}{4}

Show solution

Step 1: State Domain Constraint.

The presence of log⁡2x\log_2 x requires strictly x>0x > 0.


Step 2: Take log⁡2\log_2 of both sides.

log⁡2(xlog⁡2x−3)=log⁡2(14)\log_2\left(x^{\log_2 x - 3}\right) = \log_2\left(\frac{1}{4}\right)

Apply the power rule on the LHS and evaluate the RHS:

(log⁡2x−3)⋅log⁡2x=log⁡2(2−2)=−2(\log_2 x - 3) \cdot \log_2 x = \log_2(2^{-2}) = -2


Step 3: Transform into a quadratic in y=log⁡2xy = \log_2 x.

Let y=log⁡2xy = \log_2 x:

y(y−3)=−2y(y - 3) = -2

y2−3y+2=0y^2 - 3y + 2 = 0


Step 4: Apply the Vieta Root-Product Shortcut.

Let the roots of the quadratic in yy be y1y_1 and y2y_2, corresponding to x1x_1 and x2x_2 in original space.

By Vieta's formulas, the sum of the roots in yy is:

y1+y2=−−31=3y_1 + y_2 = -\frac{-3}{1} = 3

Since y1=log⁡2x1y_1 = \log_2 x_1 and y2=log⁡2x2y_2 = \log_2 x_2:

log⁡2x1+log⁡2x2=3\log_2 x_1 + \log_2 x_2 = 3

log⁡2(x1⋅x2)=3\log_2(x_1 \cdot x_2) = 3

Converting to exponential form:

x1⋅x2=23=8x_1 \cdot x_2 = 2^3 = 8


Step 5: Verification by finding individual roots.

Factorizing the quadratic: (y−1)(y−2)=0  ⟹  y1=1,y2=2(y - 1)(y - 2) = 0 \implies y_1 = 1, y_2 = 2.

  • For y1=1y_1 = 1: log⁡2x1=1  ⟹  x1=21=2\log_2 x_1 = 1 \implies x_1 = 2^1 = 2

  • For y2=2y_2 = 2: log⁡2x2=2  ⟹  x2=22=4\log_2 x_2 = 2 \implies x_2 = 2^2 = 4

Product of roots: x1⋅x2=2×4=8x_1 \cdot x_2 = 2 \times 4 = 8 (Verified!)

The product of all real roots is 88.

Takeaway: For quadratic equations in y = log b x, the sum of roots y 1 + y 2 = -B/A immediately delivers the product of original roots x 1 x 2 = b^-B/A in one line.

Example 5.3: Logarithmic Inequality with Base in (0, 1)

Solve for all real values of xx:

log⁡1/3(x2−3x+2)≥−1\log_{1/3}(x^2 - 3x + 2) \ge -1

Show solution

Step 1: Establish the Domain Constraint FIRST.

The argument of the logarithm must be strictly positive:

x2−3x+2>0x^2 - 3x + 2 > 0

Factorizing the quadratic:

(x−1)(x−2)>0(x - 1)(x - 2) > 0

By the wavy-curve method, the product is positive outside the roots:

x∈(−∞,1)∪(2,∞)— (Domain D)x \in (-\infty, 1) \cup (2, \infty) \quad \text{--- (Domain } \mathcal{D}\text{)}


Step 2: Solve the inequality with the Base Direction Flip.

Notice that the base is b=13b = \frac{1}{3}.

Because 0<13<10 < \frac{1}{3} < 1, the logarithmic function is strictly decreasing, so the inequality sign flips from ≥\ge to ≤\le:

x2−3x+2≤(13)−1x^2 - 3x + 2 \le \left(\frac{1}{3}\right)^{-1}

Compute the right-hand side:

(13)−1=3\left(\frac{1}{3}\right)^{-1} = 3

Therefore:

x2−3x+2≤3x^2 - 3x + 2 \le 3

x2−3x−1≤0x^2 - 3x - 1 \le 0


Step 3: Find the roots of x2−3x−1=0x^2 - 3x - 1 = 0.

Using the quadratic formula:

x=−(−3)±(−3)2−4(1)(−1)2(1)=3±9+42=3±132x = \frac{-(-3) \pm \sqrt{(-3)^2 - 4(1)(-1)}}{2(1)} = \frac{3 \pm \sqrt{9 + 4}}{2} = \frac{3 \pm \sqrt{13}}{2}

Thus, the quadratic is non-positive between the roots:

3−132≤x≤3+132\frac{3 - \sqrt{13}}{2} \le x \le \frac{3 + \sqrt{13}}{2}

Note the numerical approximations (since 13≈3.606\sqrt{13} \approx 3.606):

  • Lower bound: 3−3.6062≈−0.303\frac{3 - 3.606}{2} \approx -0.303

  • Upper bound: 3+3.6062≈3.303\frac{3 + 3.606}{2} \approx 3.303


Step 4: Intersect with Domain D\mathcal{D}.

We must intersect [3−132,3+132]\left[\frac{3 - \sqrt{13}}{2}, \frac{3 + \sqrt{13}}{2}\right] with (−∞,1)∪(2,∞)(-\infty, 1) \cup (2, \infty):

  • Left segment: [3−132,1)\left[\frac{3 - \sqrt{13}}{2}, 1\right)

  • Right segment: (2,3+132]\left(2, \frac{3 + \sqrt{13}}{2}\right]

The complete solution set is x∈[3−132,1)∪(2,3+132]x \in \left[\frac{3 - \sqrt{13}}{2}, 1\right) \cup \left(2, \frac{3 + \sqrt{13}}{2}\right].

Takeaway: When base is in (0, 1), flip the inequality sign immediately, then strictly intersect the resulting interval with the argument positivity domain.

Example 5.4: Logarithmic Inequality with Variable Base Bifurcation

Find the complete set of real values of xx satisfying the inequality:

log⁡x(2x+3)<2\log_x(2x + 3) < 2

Show solution

Step 1: Domain Declaration.

  • Base constraints: x>0x > 0 and x≠1x \ne 1

  • Argument constraint: 2x+3>0  ⟹  x>−1.52x + 3 > 0 \implies x > -1.5

Intersection gives the global domain:

D:x∈(0,1)∪(1,∞)\mathcal{D}: x \in (0, 1) \cup (1, \infty)

Because the base xx can lie either above 11 or between 00 and 11, we must split into two separate cases.


Case 1: Base is strictly greater than 11 (x>1x > 1).

Here, the base is an expander (>1> 1), so the inequality direction is preserved:

2x+3<x22x + 3 < x^2

x2−2x−3>0x^2 - 2x - 3 > 0

Factorizing the quadratic:

(x−3)(x+1)>0  ⟹  x∈(−∞,−1)∪(3,∞)(x - 3)(x + 1) > 0 \implies x \in (-\infty, -1) \cup (3, \infty)

Intersecting with the case condition x>1x > 1:

x∈(3,∞)— (Case 1 Solution)x \in (3, \infty) \quad \text{--- (Case 1 Solution)}


Case 2: Base is strictly between 00 and 11 (0<x<10 < x < 1).

Here, the base is a contractor (0<x<10 < x < 1), so the inequality direction flips from << to >>:

2x+3>x22x + 3 > x^2

x2−2x−3<0x^2 - 2x - 3 < 0

Factorizing the quadratic:

(x−3)(x+1)<0  ⟹  −1<x<3(x - 3)(x + 1) < 0 \implies -1 < x < 3

Intersecting with the case condition 0<x<10 < x < 1:

x∈(0,1)— (Case 2 Solution)x \in (0, 1) \quad \text{--- (Case 2 Solution)}


Step 5: Combine solutions from both cases.

S=(0,1)∪(3,∞)\mathcal{S} = (0, 1) \cup (3, \infty)

The complete solution set is x∈(0,1)∪(3,∞)x \in (0, 1) \cup (3, \infty).

Takeaway: When the base contains a variable, bifurcate into two mutually exclusive cases: x 1 (keep inequality direction) and 0 < x < 1 (flip inequality direction).

Example 5.5: Telescoping Logarithmic Series with Difference of Squares

Evaluate the exact sum of the series:

S=∑k=299log⁡10(1−1k2)S = \sum_{k=2}^{99} \log_{10}\left(1 - \frac{1}{k^2}\right)

and calculate the exact value of 10S10^S.

Show solution

Step 1: Factorize the argument inside the logarithm.

Notice that 1−1k21 - \frac{1}{k^2} is a difference of squares:

1−1k2=k2−1k2=(k−1)(k+1)k⋅k=(k−1k)⋅(k+1k)1 - \frac{1}{k^2} = \frac{k^2 - 1}{k^2} = \frac{(k - 1)(k + 1)}{k \cdot k} = \left(\frac{k - 1}{k}\right) \cdot \left(\frac{k + 1}{k}\right)


Step 2: Convert the sum of logarithms into the logarithm of a product.

S=∑k=299log⁡10[(k−1k)⋅(k+1k)]=log⁡10[∏k=299(k−1k)⋅∏k=299(k+1k)]S = \sum_{k=2}^{99} \log_{10}\left[\left(\frac{k - 1}{k}\right) \cdot \left(\frac{k + 1}{k}\right)\right] = \log_{10}\left[\prod_{k=2}^{99} \left(\frac{k - 1}{k}\right) \cdot \prod_{k=2}^{99} \left(\frac{k + 1}{k}\right)\right]


Step 3: Evaluate the first product P1=∏k=299(k−1k)P_1 = \prod_{k=2}^{99} \left(\frac{k - 1}{k}\right).

Write out consecutive factors:

P1=12×23×34×⋯×9899P_1 = \frac{1}{2} \times \frac{2}{3} \times \frac{3}{4} \times \dots \times \frac{98}{99}

Every numerator cancels with the preceding denominator, leaving only:

P1=199P_1 = \frac{1}{99}


Step 4: Evaluate the second product P2=∏k=299(k+1k)P_2 = \prod_{k=2}^{99} \left(\frac{k + 1}{k}\right).

Write out consecutive factors:

P2=32×43×54×⋯×10099P_2 = \frac{3}{2} \times \frac{4}{3} \times \frac{5}{4} \times \dots \times \frac{100}{99}

Every denominator cancels with the preceding numerator, leaving only:

P2=1002=50P_2 = \frac{100}{2} = 50


Step 5: Multiply the two telescoping products.

Total Product P=P1×P2=199×50=5099\text{Total Product } P = P_1 \times P_2 = \frac{1}{99} \times 50 = \frac{50}{99}

Therefore:

S=log⁡10(5099)S = \log_{10}\left(\frac{50}{99}\right)


Step 6: Compute 10S10^S.

10S=10log⁡10(50/99)=509910^S = 10^{\log_{10}(50/99)} = \frac{50}{99}

The exact sum is S=log⁡10(5099)S = \log_{10}\left(\frac{50}{99}\right), and 10S=509910^S = \frac{50}{99}.

Takeaway: Factor 1 - 1/k^2 as ((k-1)/k) ((k+1)/k) to split the logarithmic product into two independent telescoping cascades, yielding (1/n) ((n+1)/2).

Example 5.6: Infinite Geometric Series Inside a Logarithmic Argument

Solve for all real values of xx:

log⁡3(x+x24+x316+x464+… )=1\log_3\left(x + \frac{x^2}{4} + \frac{x^3}{16} + \frac{x^4}{64} + \dots\right) = 1

Show solution

Step 1: Analyze the infinite series inside the argument.

The series is:

S∞=x+x24+x316+x464+…S_\infty = x + \frac{x^2}{4} + \frac{x^3}{16} + \frac{x^4}{64} + \dots

This is an infinite Geometric Progression (GP) with:

  • First term: a=xa = x

  • Common ratio: r=x2/4x=x4r = \frac{x^2 / 4}{x} = \frac{x}{4}


Step 2: Enforce the convergence and domain criteria.

  1. Convergence Criterion: An infinite GP converges if and only if ∣r∣<1|r| < 1:

    ∣x4∣<1  ⟹  ∣x∣<4  ⟹  −4<x<4\left|\frac{x}{4}\right| < 1 \implies |x| < 4 \implies -4 < x < 4

  2. Logarithmic Argument Positivity: The sum S∞S_\infty must be strictly positive: S∞>0S_\infty > 0.


Step 3: Compute the sum to infinity S∞S_\infty.

Using the sum formula S∞=a1−rS_\infty = \frac{a}{1 - r}:

S∞=x1−x4=x4−x4=4x4−xS_\infty = \frac{x}{1 - \frac{x}{4}} = \frac{x}{\frac{4 - x}{4}} = \frac{4x}{4 - x}


Step 4: Solve the logarithmic equation.

log⁡3(4x4−x)=1\log_3\left(\frac{4x}{4 - x}\right) = 1

Convert to exponential form:

4x4−x=31=3\frac{4x}{4 - x} = 3^1 = 3

Multiply both sides by (4−x)(4 - x):

4x=3(4−x)4x = 3(4 - x)

4x=12−3x4x = 12 - 3x

7x=12  ⟹  x=1277x = 12 \implies x = \frac{12}{7}


Step 5: Verify the convergence and positivity constraints.

  • Convergence check: ∣x∣=∣127∣≈1.714<4|x| = \left|\frac{12}{7}\right| \approx 1.714 < 4. (Converges!)

  • Positivity check: 4(12/7)4−12/7=48/716/7=3>0\frac{4(12/7)}{4 - 12/7} = \frac{48/7}{16/7} = 3 > 0. (Positive!)

  • Logarithm check: log⁡3(3)=1\log_3(3) = 1. (Verified!)

The unique real solution is x=127x = \frac{12}{7}.

Takeaway: When an infinite series appears inside a logarithm, evaluate the sum using S infty = a / (1 - r) and strictly verify the convergence criterion r < 1.

Practice questions

Question 1 · CAT Level

Find the product of all real solutions to the equation:

xlog⁡3x−4=7x^{\log_3 x - 4} = 7

  1. A

    9

  2. B

    27

  3. C

    81

  4. D

    243

Show answer and explanation

Answer: (C) 81

Taking the base-33 logarithm of both sides gives log⁡3(xlog⁡3x−4)=log⁡37\log_3\left(x^{\log_3 x - 4}\right) = \log_3 7.

Bringing down the exponent yields (log⁡3x−4)log⁡3x=log⁡37(\log_3 x - 4)\log_3 x = \log_3 7.

Setting u=log⁡3xu = \log_3 x produces the quadratic equation u2−4u−log⁡37=0u^2 - 4u - \log_3 7 = 0.

Since the discriminant D=16+4log⁡37>0D = 16 + 4\log_3 7 > 0, there are two real roots u1,u2u_1, u_2 whose sum is u1+u2=4u_1 + u_2 = 4 by Vieta's formulas.

Since u1+u2=log⁡3x1+log⁡3x2=log⁡3(x1x2)=4u_1 + u_2 = \log_3 x_1 + \log_3 x_2 = \log_3(x_1 x_2) = 4, the product of solutions is x1x2=34=81x_1 x_2 = 3^4 = 81.

Question 2 · CAT Level

If x,y,zx, y, z are positive real numbers in geometric progression with x<y<zx < y < z, such that:

log⁡3x+log⁡3y+log⁡3z=12andlog⁡3x⋅log⁡3y⋅log⁡3z=48\log_3 x + \log_3 y + \log_3 z = 12 \quad \text{and} \quad \log_3 x \cdot \log_3 y \cdot \log_3 z = 48

find the value of zz.

  1. A

    243

  2. B

    729

  3. C

    2187

  4. D

    6561

Show answer and explanation

Answer: (B) 729

Since x,y,zx, y, z form a GP, y2=xz  ⟹  2log⁡3y=log⁡3x+log⁡3zy^2 = xz \implies 2\log_3 y = \log_3 x + \log_3 z, so log⁡3x,log⁡3y,log⁡3z\log_3 x, \log_3 y, \log_3 z form an AP.

Represent the AP terms as v−d,v,v+dv - d, v, v + d. Their sum is (v−d)+v+(v+d)=3v=12  ⟹  v=4(v - d) + v + (v + d) = 3v = 12 \implies v = 4.

Their product is (4−d)⋅4⋅(4+d)=48  ⟹  16−d2=12  ⟹  d2=4(4 - d) \cdot 4 \cdot (4 + d) = 48 \implies 16 - d^2 = 12 \implies d^2 = 4.

Since x<y<zx < y < z, the progression is strictly increasing, so common difference d=2d = 2.

The largest term is log⁡3z=v+d=4+2=6\log_3 z = v + d = 4 + 2 = 6, which gives z=36=729z = 3^6 = 729.

Question 3 · CAT Level

Find the value of the infinite series summation:

S=∑n=2∞log⁡2(1+1n2−1)S = \sum_{n=2}^{\infty} \log_2\left(1 + \frac{1}{n^2 - 1}\right)

  1. A

    0

  2. B

    1

  3. C

    2

  4. D

    ∞\infty

Show answer and explanation

Answer: (B) 1

Rewrite each term inside the logarithm: 1+1n2−1=n2n2−1=nn−1⋅nn+11 + \frac{1}{n^2 - 1} = \frac{n^2}{n^2 - 1} = \frac{n}{n-1} \cdot \frac{n}{n+1}.

The finite partial sum is SN=∑n=2Nlog⁡2(n2(n−1)(n+1))=log⁡2(PN)S_N = \sum_{n=2}^N \log_2\left(\frac{n^2}{(n-1)(n+1)}\right) = \log_2(P_N), where PNP_N is the partial product.

Expanding PN=(2⋅21⋅3)(3⋅32⋅4)…(N⋅N(N−1)(N+1))P_N = \left(\frac{2 \cdot 2}{1 \cdot 3}\right) \left(\frac{3 \cdot 3}{2 \cdot 4}\right) \dots \left(\frac{N \cdot N}{(N-1)(N+1)}\right).

All intermediate factors telescope and cancel, leaving PN=21⋅NN+1=2NN+1P_N = \frac{2}{1} \cdot \frac{N}{N+1} = \frac{2N}{N+1}.

Taking the limit as N→∞N \to \infty, lim⁡N→∞PN=2\lim_{N \to \infty} P_N = 2, so S=log⁡2(2)=1S = \log_2(2) = 1.

Question 4 · CAT Level

Find the number of integer values of xx satisfying the inequality:

log⁡1/3(x2−3x−1)≥−1\log_{1/3}(x^2 - 3x - 1) \ge -1

  1. A

    2

  2. B

    4

  3. C

    6

  4. D

    Infinitely many

Show answer and explanation

Answer: (A) 2

First, the base 13∈(0,1)\frac{1}{3} \in (0, 1) reverses the inequality sign: x2−3x−1≤(13)−1=3x^2 - 3x - 1 \le \left(\frac{1}{3}\right)^{-1} = 3.

This simplifies to x2−3x−4≤0  ⟹  (x−4)(x+1)≤0x^2 - 3x - 4 \le 0 \implies (x - 4)(x + 1) \le 0, yielding candidate interval x∈[−1,4]x \in [-1, 4].

Second, argument positivity requires x2−3x−1>0x^2 - 3x - 1 > 0, whose roots are 3±132≈−0.30\frac{3 \pm \sqrt{13}}{2} \approx -0.30 and 3.303.30.

Thus, valid values must lie in [−1,3−132)∪(3+132,4]\left[-1, \frac{3 - \sqrt{13}}{2}\right) \cup \left(\frac{3 + \sqrt{13}}{2}, 4\right].

Testing integers in [−1,4][-1, 4]: x=−1x = -1 and x=4x = 4 satisfy the condition, while x∈{0,1,2,3}x \in \{0, 1, 2, 3\} produce negative arguments, leaving exactly 22 integer solutions.

Question 5 · CAT Level

Find the complete set of real values of xx satisfying the inequality:

log⁡0.5(log⁡4(x2−5))>0\log_{0.5}\left(\log_4(x^2 - 5)\right) > 0

  1. A

    x∈(−3,3)x \in (-3, 3)

  2. B

    x∈(6,3)x \in (\sqrt{6}, 3)

  3. C

    x∈(−3,−6)∪(6,3)x \in (-3, -\sqrt{6}) \cup (\sqrt{6}, 3)

  4. D

    x∈(−3,−5)∪(5,3)x \in (-3, -\sqrt{5}) \cup (\sqrt{5}, 3)

Show answer and explanation

Answer: (C) x (-3, - 6) ( 6, 3)

The argument of the outer logarithm must be strictly positive: log⁡4(x2−5)>0  ⟹  x2−5>40=1  ⟹  x2>6\log_4(x^2 - 5) > 0 \implies x^2 - 5 > 4^0 = 1 \implies x^2 > 6.

Since the outer base is 0.5<10.5 < 1, the inequality flips: log⁡0.5(log⁡4(x2−5))>0=log⁡0.5(1)  ⟹  log⁡4(x2−5)<1\log_{0.5}(\log_4(x^2 - 5)) > 0 = \log_{0.5}(1) \implies \log_4(x^2 - 5) < 1.

Since the inner base is 4>14 > 1, inequality direction is preserved: x2−5<41=4  ⟹  x2<9x^2 - 5 < 4^1 = 4 \implies x^2 < 9.

Combining both constraints gives 6<x2<96 < x^2 < 9.

Taking square roots yields the symmetric open intervals x∈(−3,−6)∪(6,3)x \in (-3, -\sqrt{6}) \cup (\sqrt{6}, 3).

Try real CAT questions

Study this concept with interactive practice in PrepSoc.

Open the concept library