This week I found a really cool planetary gearbox in a bin in my lab. I’ve always wanted to work through the derivation of the transmission ratios for planetary systems properly, so this seemed like the chance.

One of the genuinely interesting things about planetary gear sets is that you get several different transmission ratios out of the same hardware depending on which member you hold fixed and which you drive. The typical configuration fixes the ring and drives the sun, with a shaft on the carrier as the output.
Why the usual formula never satisfied me
For an ordinary gear transmission the ratio comes straight from the sizes of the gears, and the same relations hold between torque and rotational velocity.

That applies to planetary boxes too — but with one member stationary the overall ratio gets harder to see, and I was never satisfied just using the formulas some textbooks and websites hand you. So I derived them.

Setting up the geometry

| Ring teeth | 103 |
| Planet teeth | 39 |
| Sun teeth | 25 |
The constraint N_R = 2·N_P + N_S has to hold for the set to assemble, and it does: 2(39) + 25 = 103.
The key physical fact everything rests on is that linear velocities must match at the contact point of meshing gears — V = ωR. From there each configuration is just bookkeeping.

Three ratios from one gearbox

| Held fixed | Input → Output | Ratio |
|---|---|---|
| Ring | Sun → Carrier | 5.12 |
| Carrier | Sun → Ring | 4.12 |
| Sun | Carrier → Ring | 0.80 |
Same three gears, three quite different machines.
What actually limits the torque?

Since the sun, planet and ring share the same material and diametral pitch, the Lewis form factor and the face width are what decide which one gives first.
| Gear | Allowable tangential load |
|---|---|
| Ring | 1400 N |
| Planet | 966 N |
| Sun | 955.5 N |
The sun gear teeth are the limiting factor — maximum input torque of about 11 N·m. But thanks to the ratios derived above, that becomes roughly five times as much torque at the carrier output with the ring held stationary.
Checking the other failure paths

| Key in shear | 115 N·m |
| Shaft in torsion | 94 N·m |
| Sun gear teeth | 11 N·m |
So the ordering is clear: the teeth give out roughly an order of magnitude before either the key or the shaft. Between those two, the shaft would yield before the key, assuming the same material.
Reference
- Epicyclic gears — Roymech, which has a helpful figure showing how instantaneous velocities apply to planetary gears.