Consider the set
$X:=\{(x,y,z)\in\mathbb{R}^3 | x^2+3y^2=1+z^2\}$
I have to show that $X$ is a submanifold of $\mathbb{R}^3$ (and this is trivial); then, using on $T\mathbb{R}^3$ the standard coordinates $x,y,z,\partial_x,\partial_y,\partial_z$ I have to find the equations that describe $TX\subseteq T\mathbb{R}^3$ and $NX\subseteq T\mathbb{R}^3$.
Any suggestion? Thanks...
The tangent bundle $TX$ is the set of $6$-tuples $(x,y,z;\partial_ x,\partial_ y,\partial_ z)\in T\mathbb R^3=\mathbb R^3\times \mathbb R^3$ satisfying the equations $x^2+3y^2-z^2-1=0$ and $2x\partial_ x+6y\partial_y-2z\partial_z=0$.
The normal bundle $NX$ is the set of $6$-tuples $(x,y,z;2xr,6yr,-2zr)\in T\mathbb R^3=\mathbb R^3\times \mathbb R^3$ with $x^2+3y^2-z^2-1=0$ and $r\in \mathbb R$ .
Edit
At Frankenstein's request I'll add a few words of explanation.
The equation $x^2+3y^2-z^2-1=0$ just says that we are studying a point $P\in X$: this concerns the first three coordinates of our $6$-tuple $(x,y,z;\partial_ x,\partial_ y,\partial_ z)\in T\mathbb R^3=\mathbb R^3\times \mathbb R^3$.
At $P$ the gradient of $f(x,y,z)=x^2+3y^2-z^2-1$ is the vector $\nabla f(P)=(2x,6y,-2z)$.
A vector $v=(\partial_ x,\partial_ y,\partial_ z)\in T_P\mathbb R^3$ is tangent to $X$ if $v$ is orthogonal to $\nabla f(P)$, which translates into $\nabla f(P)\cdot (\partial_ x,\partial_ y,\partial_ z)=2x\partial_ x+6y\partial_y-2z\partial_z=0$.
A vector $v=(\partial_ x,\partial_ y,\partial_ z)\in T_P\mathbb R^3$ is normal to $X$ if $v$ is proportional to $\nabla f(P)$, which translates into $ (\partial_ x,\partial_ y,\partial_ z)=r \nabla f(P)=r(2x, 6y,-2z)=(2xr,6yr,-2zr)$.
A remark
Notice that the description of $TX$ is purely by equations whereas that of $NX$ is a mixture of an equation and a parametrization.
This is a linear algebra phenomenon:
$\bullet$ A plane in $\mathbb R^3$ is best described by the equation $l(x)=0$ where $l$ is a linear form, unique up to a non-zero constant.
$\bullet \bullet$ However a line must be described by two linear forms and there is absolutely no canonical choice for these.
So it is better to describe a line by choosing a non zero vector $v$ on it , which will be unique up to a constant, and take multiples $rv$ of that vector.
In other words, a line is better described parametrically than equationally.