I have been trying to solve this equation for over a week now: $$\tan5x-2\tan3x=\tan3x\tan5x$$

I found one solution $x=k\pi$ but I cannot prove that this is the only solution. It is equivalent to: $$\sin 5x \cos 3x - 2 \sin 3x \cos 5x = \sin 3x \sin 5x$$ $$\sin 2x =\sqrt 2 \sin 3x \sin(5x+ \frac{\pi}{4})$$ $$2\sin x \cos x =2\sqrt 2 \sin x \cos (2x +1) \sin(5x+ \frac{\pi}{4})$$ So one solution is $x=k\pi$. But how to proceed with the remaining: $$\sqrt 2 \cos x = 2\cos (2x +1) \sin(5x+ \frac{\pi}{4}) $$ It seems impossible to use derivatives to prove that it has no solution because the argument of the third cosine is shifted by $\frac{\pi}{4}$

  • $\begingroup$ What does tg5x mean ? $\endgroup$ – Amr Nov 24 '12 at 15:28
  • $\begingroup$ is tg= trigonometric? $\endgroup$ – clark Nov 24 '12 at 15:29
  • $\begingroup$ Corrected. It means tan5x $\endgroup$ – Adam Nov 24 '12 at 15:34
  • 5
    $\begingroup$ Wolfram Alpha thinks there are more solutions (none of which --apart from the multiplies of $\pi$-- are very attractive): wolframalpha.com/input/… $\endgroup$ – Blue Nov 24 '12 at 15:51
  • $\begingroup$ @Blue: do you mind making your comment as an answer? $\endgroup$ – Willie Wong Jan 29 '13 at 13:13

Multiply $$ \tan(5x)-2\tan(3x)=\tan(3x)\tan(5x) $$ by $\cos(3x)\cos(5x)$ and cancel to get $$ \begin{align} 0 &=\sin(5x)\cos(3x)-2\sin(3x)\cos(5x)-\sin(3x)\sin(5x)\\ &=\tfrac12(\sin(8x)+\sin(2x))-2\cdot\tfrac12(\sin(8x)-\sin(2x))-\tfrac12(\cos(2z)-\cos(8x))\\ &=\tfrac12(\cos(8x)-\sin(8x)+3\sin(2x)-\cos(2x))\\ &=\tfrac12(\sqrt2\sin(8x+\tfrac34\pi)+\sqrt{10}\sin(2x-\arctan(\tfrac13)))\\ \end{align} $$ There doesn't appear to be any simple way to solve this other than numerically.

The function has a period of $\pi$. Plotting, it appears that this function has $4$ roots in $[0,\pi)$.

$\hspace{8mm}$enter image description here

Numerically, the roots in $[0,\pi)$ are $$ x=\left\{\begin{array}{} 0.0000000000000000000\\ 0.3598476319540602998\\ 1.6915164344292107407\\ 3.0829964597422711815 \end{array}\right. $$


Making my comment into an answer, as requested:

Wolfram Alpha thinks there are more solutions (none of which ---apart from the multiples of $\pi$--- are very attractive). See here.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.