Second-Order Linear ODE with Constant Coefficients
Problem
Solve y'' - 5y' + 6y = 0. Find the characteristic roots r = 2, 3 and write the general solution y = C1 e^(2x) + C2 e^(3x).
Explanation
The general form
A homogeneous second-order linear ODE with constant coefficients has the form
"Homogeneous" here means the right side is (the #178 sense doesn't apply — different usage). "Constant coefficients" means don't depend on .
The whole method to solve these is one slick idea: guess . Because plugging that in gives and , the equation collapses to the algebraic characteristic equation
Solve this quadratic for — three possible outcomes, three kinds of solutions. (Repeated roots are in #185; complex roots are in #184.) This problem focuses on the clean case of two real distinct roots.
The given equation
Coefficients , , .
Step-by-step solution
Step 1 — Write the characteristic equation.
Step 2 — Solve it.
Factor: .
(Or use the quadratic formula: .)
Step 3 — Write the two basic solutions.
Step 4 — Combine into the general solution.
Because the ODE is linear and homogeneous, any linear combination of solutions is also a solution (see #200 on superposition). The general solution is
for arbitrary constants .
Verification
Differentiate :
Plug into the ODE :
The clean cancellation comes from the fact that each satisfies .
Why two arbitrary constants?
The ODE is order 2, so its general solution has a 2-parameter family. Geometrically, the solution space is a 2-dimensional vector space (over the reals), spanned by the basis .
Two pieces of data (typically an IVP , , or a BVP with boundary values at two endpoints) pin down and .
Initial value problem
Say and . Then
From the first, . Substitute: , .
Check: ✓; , ✓.
The three cases of the characteristic equation
Let the discriminant be . Then:
- : two real distinct roots . General solution . (This problem. Also #183.)
- : repeated real root . General solution . (See #185.)
- : complex conjugate roots . General solution . (See #184.)
Pattern: you always get exactly linearly independent solutions, but their form depends on which case the discriminant falls into.
Stability at a glance
Just by looking at the roots you can predict long-term behaviour:
- Both : solutions decay to — stable.
- Both : solutions grow unboundedly — unstable.
- One , one : saddle — generic initial data blows up, but a 1-parameter family of initial data decays.
For our problem — both positive, so every non-trivial solution grows exponentially. No equilibrium other than .
Common mistakes
- Forgetting to divide by . If , the characteristic equation is , not . (Or keep and solve directly — either way, don't drop .)
- Using only one constant. A second-order linear ODE needs two arbitrary constants in its general solution.
- Plugging in before solving for . Apply initial conditions after you have the general solution with both constants.
- Treating non-distinct roots the same way. If , then and are the same solution — you need to manufacture a second linearly independent one (it's , see #185).
Try it in the visualization
Slide the coefficients and and watch the discriminant cross zero. See the solution morph from two exponentials (Δ>0) into a repeated-root form (Δ=0) into oscillating exponentials (Δ<0). Each regime has a distinctly different shape.
Interactive Visualization
Parameters
Got your own math or physics problem?
Turn any problem into an interactive visualization like this one — powered by AI, generated in seconds. Free to try, no credit card required.