So, I know that $$\lim_{n\to\infty}\frac{F_{n+1}}{F_n}=\phi$$
where $F_n$ stands for the n'th Fibonacci number
I was interested in measuring the error of the convergence of the above limit and was drawn to the conjecture that:
$$\sum_{n=1}^\infty \left(\phi-\frac{F_{n+1}}{F_{n}}\right)=\frac{1}{\pi}$$
How might we go about proving this result?
Edit:
The solution is actually not $\frac{1}{\pi}$. I had thought that it was due to how close the sum and $1/\pi$ actually are. I am curious (if one exists) if there is a closed form for the sum.
So, we get that:
$$\sum_{n=1}^{\infty}\left[\phi-\frac{\text{F}_{n+1}}{\text{F}_n}\right]=\sum_{n=1}^{\infty}\left[\phi-\frac{\frac{\phi^{n+1}-\frac{1}{\left(-\phi\right)^{n+1}}}{\sqrt{5}}}{\frac{\phi^n-\frac{1}{\left(-\phi\right)^n}}{\sqrt{5}}}\right]=\sum_{n=1}^{\infty}\left[\phi-\frac{1+(-\phi)^n\phi^{2+n}}{\phi\left(\left(-\phi^2\right)^n-1\right)}\right]=$$ $$\sum_{n=1}^{\infty}\frac{1+\phi^2}{\phi-\phi\left(-\phi^2\right)^n}=\frac{\sqrt{5}\left(5\pi-i\left(\ln(5)+2\psi_{\frac{1}{2}\left(-3-\sqrt{5}\right)}^{(0)}\left[1-\frac{2\pi}{\pi-i\text{arccosh}\left(\frac{3}{2}\right)}\right]\right)\right)}{2\left(\pi-i\text{arccosh}\left(\frac{3}{2}\right)\right)}$$