Separation works only when the right-hand side factors into an x-part times a y-part. If it does, you can divide the y-part over to the dy side and you're ready to integrate. If it can't be factored that way — most often because it's a sum — you need a different method.
Spotting separability is its own skill — and a tested one. For each equation, decide whether the right side factors into (an x-part)(a y-part). Watch for sums that don't factor, and products hiding inside things you can pull apart.
The whole method is four moves: separate, integrate both sides, solve for y, then apply the initial condition. Build your own equation — pick an x-part, a y-part, and a starting value — and watch the method solve it, then thread your exact solution through its slope field. Switch to Test me to choose each step yourself.
The method needs the right side to factor — an x-part times a y-part. A sum almost never qualifies.
Divide the y-part across to the dy side, then integrate both sides. That single move is the heart of the whole technique.
Integrating both sides produces a constant on each — but combine them into a single +C on the x-side. Two constants is a red flag.
The general solution carries an unknown constant. The initial condition is the one point that fixes it, giving the particular solution.
Six checks across the whole method — recognizing the shape, separating and integrating, solving for y, and pinning the constant with a condition. Try each before revealing.
Six fast multiple-choice checks across the whole lesson. Pick an answer to see whether it's right and why — then work through all six.
The instant you integrate, a constant appears. Leave it off and the rest of the problem is wrong — and on a free-response it's an immediate point lost. Write +C the moment you integrate.
Both sides produce a constant, but they merge into one. Put a single +C on the x-side and move on — two separate constants signals a misunderstanding.
If the problem gives an initial condition, you're not done at y = A·e^(…). Apply the point to solve for the constant and report the particular solution.
When you exponentiate ln|y| = … + C, the e^C becomes a single new constant (call it A). Don't drag the C along inside the exponent.
dy/dx = x + y can't be separated — no division pulls the variables apart. Separation needs a product; a sum means a different method entirely.
Integrating dy/y gives ln|y|, not ln y. It usually washes out into ±A when you solve for y, but the careful step is ln|y|.