Gyroscope — steady precession derivation: Visual walkthrough

Step 1 — Draw the top and name its parts

WHAT. We have a heavy spinning wheel on a stick. One end of the stick sits in a pivot — a fixed point we call $O$ that the stick can swing around but never leaves. The other end carries all the mass; its balance point (the center of mass, the single point where the object's weight effectively acts) sits a distance $ℓ$ out along the stick.

WHY. Before any physics, we must fix a picture and a vocabulary. Every later arrow hangs off this diagram, so we earn each label now.

PICTURE. Look at the figure. The stick tilts away from straight-up by an angle $θ$ (the tilt angle, measured between the axis and the vertical). "Vertical" is the up-down direction, drawn as the black dashed line and labelled $\overset{z}{^}$ (the little hat means "an arrow of length 1 pointing that way" — a pure direction, no size). The red arrow is the wheel's axis, our star object.

Figure — Gyroscope — steady precession derivation

Step 2 — What "steady" means, drawn as a cone

WHAT. Steady precession is the special motion where the tilt never changes ( $θ$ fixed), the axis swings around the vertical at a constant speed we name $Ω \equiv \dot{ϕ}$ , and the wheel spins at a constant rate.

WHY. "Constant" is what lets us balance one steady push against one steady turn. If $θ$ were wobbling (that wobble is Nutation), nothing would cancel cleanly.

PICTURE. The red axis sweeps the surface of an invisible ice-cream cone whose point is at $O$ . The tip traces the red horizontal circle at the top. $Ω$ is how fast we run around that circle.

Step 3 — The spin makes a big angular momentum arrow

WHAT. A spinning object carries angular momentum — call it $L$ , "how much rotation it has stored, and about which axis." For a fast wheel, $L$ points straight along the spin axis and its length is $L_{s} = I_{3} ω_{s} .$

WHY. Angular momentum is the quantity that torque acts on (next step). We need its picture — an arrow — before we can push it. We use $I_{3}$ , not $I_{1}$ : $I_{3}$ is the moment of inertia about the symmetry axis, the very axis the wheel spins about. Using the sideways moment $I_{1}$ here would answer the wrong question.

PICTURE. The red arrow $L$ lies along the tilted stick, same direction as the spin (right-hand rule: curl fingers with the spin, thumb gives $L$ ). It is long because the wheel spins fast.

Step 4 — Gravity's torque points sideways

WHAT. Weight $m g$ pulls the center of mass straight down. Its turning-effect about $O$ is the torque $τ = r \times m g$ , where $r = ℓ \overset{e}{^}_{3}$ is the arrow from $O$ to the center of mass, and $\overset{e}{^}_{3}$ is the unit direction along the stick. The size is $∣ τ ∣ = m g ℓ sin θ .$

WHY. Torque is the only thing that can change $L$ . The pivot's push acts at $O$ , so it has zero lever arm and zero torque — gravity is the whole story. We use the cross product $\times$ because torque asks "how much does this force twist us?", and $r \times F$ is exactly the tool that returns a perpendicular twist-axis whose size is force times perpendicular lever arm.

WHY $sin θ$ ? The lever arm — the horizontal distance from the vertical line through $O$ out to the center of mass — is $ℓ sin θ$ . When the stick lies almost flat ( $θ \to 9 0^{\circ}$ ) that offset is largest; when it stands straight up ( $θ \to 0$ ) the offset is zero and there is no torque.

PICTURE. Gravity points down (black). The red torque arrow points horizontally, straight out of the vertical plane — perpendicular to both gravity and the stick. This sideways direction is the whole secret.

Step 5 — Torque nudges the arrow: $d L = τ d t$

WHAT. The master law of rotation is $τ = \frac{d L}{d t},$ read: "torque = how fast the angular-momentum arrow changes." Over a tiny time $d t$ , the arrow gains a tiny piece $d L = τ d t$ , added in the torque's direction.

WHY. This is Newton's second law for rotation. It is the single hinge of the whole derivation: torque doesn't move the top, it edits its $L$ -arrow.

PICTURE. The red $L$ -arrow's tip sits at the top of the cone. The black $d L$ tacked on is horizontal (Step 4), perpendicular to $L$ 's own horizontal shadow. Adding a perpendicular nudge doesn't lengthen the arrow — it swings its tip along the circle. That swing is precession.

Step 6 — How fast the tip circles: $\frac{d L}{d t} = Ω L_{s} sin θ$

WHAT. Only the horizontal shadow of $L$ actually goes around the circle. That shadow has length $L_{s} sin θ$ (project the tilted red arrow onto the flat ground). A horizontal arrow of length $L_{h}$ spun at rate $Ω$ sweeps its tip at speed $\frac{d L}{d t} = Ω L_{h} = Ω L_{s} sin θ .$

WHY. For anything rigidly rotating about $\overset{z}{^}$ at rate $Ω$ , calculus gives $\frac{d L}{d t} = Ω \times L$ , whose size is $Ω L sin (angle to \overset{z}{^}) = Ω (L sin θ)$ . The vertical part of $L$ sits on the axis of rotation, so it never moves — only the shadow does.

PICTURE. Split the red $L$ into an up part (grey, frozen) and a flat shadow (red, length $L_{s} sin θ$ ). The shadow is the clock hand; its tip races around at $Ω \times (its length)$ .

Step 7 — Equate the two, watch $sin θ$ vanish

WHAT. Step 4 says the tip is pushed at rate $m g ℓ sin θ$ . Step 6 says the tip actually moves at $Ω L_{s} sin θ = Ω I_{3} ω_{s} sin θ$ . These are the same physical rate, so set them equal: $m g ℓ sin θ = Ω I_{3} ω_{s} sin θ .$

WHY. $τ$ is $d L / d t$ — one arrow, two ways of measuring its rate. Balance = steady precession.

PICTURE. Both bars carry an identical $sin θ$ block (grey). Cancel the matching blocks and read off what remains: $Ω = \frac{m g ℓ}{I _{3} ω _{s}}$

Step 8 — Edge & degenerate cases (drawn)

WHAT & WHY. A good picture must survive every input. We test three limits.

(a) Straight up, $θ = 0$ . Lever arm $ℓ sin 0 = 0$ , so $τ = 0$ — no torque, no precession. This is the "sleeping top", perfectly upright. The formula's cancellation quietly assumed $sin θ \neq = 0$ ; at exactly $θ = 0$ there is simply nothing to precess.

(b) Spin dying, $ω_{s} \to$ small. $Ω = m g ℓ / (I_{3} ω_{s})$ blows up — the top tries to precess infinitely fast, which it cannot sustain. The fast-top picture breaks; real tops start nodding (Nutation) and fall. The honest limit lives in the exact quadratic below.

(c) Exact reality check. Keeping the transverse inertia $I_{1}$ (Lagrangian Mechanics does this for a Symmetric Top) turns the balance into a quadratic: $I_{1} Ω^{2} cos θ - I_{3} ω_{s} Ω + m g ℓ = 0,$ which has real $Ω$ only when $(I_{3} ω_{s})^{2} \geq 4 I_{1} m g ℓ cos θ$ — "spin it fast enough or no steady cone exists."

PICTURE. Three mini-frames: (a) upright, no red torque; (b) fast-spinning slow-circle vs slow-spinning wide-fast circle; (c) the discriminant crossing zero.

The one-picture summary

Everything collapses into one loop of logic: gravity twists sideways → the sideways twist edits $L$ → the edited $L$ drags the axis around the cone → the drag rate $Ω$ balances the twist, and $sin θ$ cancels out.

Recall Feynman retelling of the whole walkthrough

Picture a heavy spinning wheel on a stick, one end poked into a fixed corner. Gravity yanks the far end down — but here's the trick. "Twisting force" (torque) always points sideways, at right angles to gravity, because that's just what a cross product does. And torque doesn't shove the wheel; it slowly edits the wheel's big spin-arrow. Since the edit is sideways, the arrow doesn't drop — it swings around in a circle, like a clock hand. How fast it swings is set by a tug-of-war: gravity's twist on top, the spin's stubbornness on the bottom. When we write both sides down, the leaning factor $sin θ$ shows up on both and cancels — so the walk-around speed doesn't care how much the top leans. The finish: $Ω = m g ℓ / (I_{3} ω_{s})$ . Spin it harder and it circles lazier; let it die and it circles frantically, wobbles, and flops. Straight up, there's no sideways offset at all, so it just sleeps.

Recall

Which factor cancels between torque and the sweeping angular momentum? ::: $sin θ$ — it appears in both $m g ℓ sin θ$ and $Ω L_{s} sin θ$ . Why is the torque horizontal, not vertical? ::: It is $r \times m g$ , a cross product, hence perpendicular to the downward gravity. Which moment of inertia sets the spin $L$ ? ::: $I_{3}$ , about the symmetry axis — not the transverse $I_{1}$ . What happens to $Ω$ as $ω_{s}$ shrinks? ::: It grows (they are inversely related); the top precesses faster then falls.