# Derivation of magnetic field of a Solenoid: Biot savart law

## Main Question or Discussion Point

Hello,

I have seen that biot savart's law works for infinitely narrow wires:
"The formulations given above work well when the current can be approximated as running through an infinitely-narrow wire."

When I wanted to derive the magnetic field of a solenoid, I had to do this substitution:
$n_o = N/L$

$k = n_o dx$
Where k is the number of turns per dx.. But shouldn't K be an integer? so I can substitute it in the formula for circular coils. That means I have infinite number of turns and turn density of something like $\frac{a}{dx}$ where a is an integer.

Is there is something wrong or that this is the idealization that we do to the solenoid? Wouldn't it be way off the correct value?

Page 8, Example 9.

Related Other Physics Topics News on Phys.org
BvU
Homework Helper
2019 Award
But shouldn't K be an integer?
No reason. If you have an 11 cm coil with 10 turns, you have 90.91 turns/m.

Note: never, never ever write something like "$\frac{a}{dx}$ with $a$ finite".
$dx$ is universally seen as an infinitesimal (something that goes to zero). So $\frac{a}{dx}$ does not exist.

No reason. If you have an 11 cm coil with 10 turns, you have 90.91 turns/m.
Don't we idealize a solenoid as a number of circular coils?

https://i.imgur.com/RCO3qcQ.png

If we take a dx piece of this solenoid and treat at as k number of coils,

The equation for a single coil is:
$B = \frac{ u_o i R^2}{2 ( R^2 + x^2)^{\frac{3}{2}}}$
If dx has 2 turns then we multiply by 2, If it has k turns then we multiply by k. But a dx piece cant have a 2.5 or a fraction of a coil (It doesn't make sense), Can it?
That is why I said $k = n_o dx$ has to be an integer.

BvU
Again: do NOT use infinitesimals this way -- it will get you into trouble. A infinitesimal $dz$ piece has zero turns and nevertheless contributes to B with an infinitesimal contribution $dB$ proportional to ${N\over L}$. It is only when you integrate $dB$ over a finite range in $z$ that you get a finite result. The derivation (it's not a proof) in the pdf is just fine.