Getting derivatives of trig functions right the first time
The chain rule eats people alive when you're just learning differentiation for trigonometric functions because you're juggling too many derivative formulas at once. You forget whether the derivative of cosine has a negative sign, you miss a chain rule multiplier, or you apply the product rule to something that doesn't need it. I've seen the same three mistakes repeat in every calculus class I've ever graded. The core set is small enough that memorizing it should take about an hour if you work through the proofs once. d/dx[sin x] = cos x. d/dx[cos x] = -sin x. d/dx[tan x] = sec²x. d/dx[cot x] = -csc²x. d/dx[sec x] = sec x tan x. d/dx[csc x] = -csc x cot x. That's it. Six derivatives. The inverse trig functions add six more, but you typically don't need all of them unless you're working with arc lengths or control systems. What people miss is that the quotient rule shows up constantly even though textbooks rarely warn you about it. When you see something like tan x / x, the quotient rule is your only path forward, and that's where most students lose points. You have to be comfortable switching between the product rule and quotient rule on the fly rather than converting everything to a single form first.
Here's a concrete example of how this plays out in practice. Let's say you need the derivative of f(x) = 3x² sin(2x). You apply the product rule immediately. The first term differentiates to 6x, and that multiplies the untouched sin(2x). The second part requires the chain rule on the sine: cos(2x) times the derivative of 2x, which is 2. Multiply through and you get 6x sin(2x) + 6x² cos(2x). No tricks. Just following the rules in order. A harder case that comes up constantly is f(x) = (tan(x³)). This looks intimidating until you break it into layers. Outer function is the square root, which gives you 1/(2tan(x³)). Then the inner tan function gives sec²(x³). Then the x³ gives 3x². Multiply all three pieces together and you get 3x² sec²(x³) / (2tan(x³))). Write each layer on paper rather than trying to hold it in your head. I've watched people lose 20 minutes re-deriving the same answer because they skipped the intermediate steps.
The edge case nobody talks about
Last year I was reviewing a student's work on a problem involving implicit differentiation of sin(xy) = x + y. The function wraps around itself in a way that makes the chain rule look completely different from what anyone expects. The derivative of sin(xy) with respect to x gives cos(xy) times the derivative of xy, and that derivative requires the product rule because y is a function of x. So you get cos(xy)·(y + x·y'). Then you differentiate the right side to get 1 + y'. Now you have y' on both sides and you need to collect terms. The final result is y' = (1 - y·cos(xy)) / (x·cos(xy) - 1). This took me about four minutes to write out on paper but a first-timer spent twenty minutes second-guessing themselves. The issue isn't complexity. It's that implicit differentiation combined with trig functions breaks the pattern recognition students rely on. My workaround is simple: always isolate y' at the earliest possible step. Don't solve for the entire left side first, then the right, then try to combine. Move every term containing y' to one side immediately after the initial differentiation. It cuts the algebra time roughly in half and eliminates sign errors, which are by far the most common mistake in these problems.
Get the Full Details
Where this method falls apart
Differentiation for trigonometric functions works cleanly when you're dealing with standard algebraic combinations of trig terms. It becomes fragile when you introduce piecewise-defined trig functions or when the argument itself is discontinuous. Take f(x) = sin(1/x) near x = 0. The derivative exists everywhere except at zero, but the limit as x approaches zero doesn't exist because the cosine term oscillates infinitely. If you're doing numerical work near that point, standard automatic differentiation tools will give you garbage results. You need to handle that singularity separately or switch to symbolic computation. Another limitation: higher-order derivatives of trig functions grow exponentially in complexity when nested. The second derivative of sin(x²) is straightforward. The fifth derivative? You're looking at a polynomial multiplied by sin or cos with dozens of terms. At that point, manual differentiation is essentially useless. I use a computer algebra system like SymPy for anything beyond the third derivative. The output is cleaner and faster, and you spend less time making arithmetic errors. The other practical bottleneck is numerical differentiation. If you're trying to compute derivatives of trig functions using finite differences on a computer, you'll hit precision issues very quickly. The derivative of sin(x) at x = /2 is exactly 1, but a naive central difference formula with a step size of 10 gives you something like 0.999999987 instead. That might seem fine until you're stacking these errors across thousands of iterations in a simulation. Use analytic derivatives whenever possible. The extra setup time pays for itself almost immediately.
If you want a reference sheet, most university math departments post one online. MIT OpenCourseWare has a clean two-page summary of all the standard derivatives with worked examples. The Engineering Mathematics textbook by K.A. Stroud also covers this material thoroughly with practice problems. Neither requires payment. Search for "MIT 18.01 derivative table pdf" or "Stroud engineering mathematics trig derivatives" and you'll find what you need within a few minutes.
Common pitfalls to avoid
The most frequent error is assuming the derivative of sin²x is cos²x. It's not. sin²x means (sin x)², so you need the chain rule. The derivative is 2 sin x · cos x, which simplifies to sin(2x). Students who skip the chain rule step keep writing the wrong answer and then can't figure out why their integration check doesn't work. Another trap is confusing the derivative of arctan x with the derivative of tan x. One is 1/(1+x²). The other is sec²x. They look somewhat similar if you're not paying attention, and mixing them up in an exam setting costs you easy points. Write the function name out fully before you start differentiating. It takes two extra seconds and prevents this particular mistake entirely. The quotient rule application on cot x is another one. Since cot x = cos x / sin x, applying the quotient rule gives you (-sin x · sin x - cos x · cos x) / sin²x, which simplifies to -(sin²x + cos²x) / sin²x. The numerator is 1, so you get -1/sin²x, which is -csc²x. If you don't remember this derivation, you'll second-guess the formula every time you see it. Deriving it once and writing it on your reference sheet takes about three minutes and removes that uncertainty permanently.
When you're working through practice problems, start with direct applications of the basic six derivatives. Then move to single chain rule problems. Then try product and quotient rule combinations. Save implicit differentiation and higher-order derivatives for last. The progression matters because each step builds on the previous one, and skipping ahead only creates gaps in your understanding that show up later. The material isn't difficult. It's just methodical. Treat it like a checklist rather than a collection of isolated facts, and you'll get through it without much trouble.
