The comparison is usually settled with two lines: a planetary gearbox is more efficient, a worm gearbox holds position without a brake. Both lines are true, and neither decides anything, because they describe the two families as if they occupied the same design space at the same ratio. They do not. A worm gearbox reaches a ratio in one stage and pays for it in heat. A planetary gearbox reaches the same ratio in two or three stages and pays for it in length. Once that is clear, most of the remaining comparison follows, including the part that the efficiency figures leave out.

The Same Ratio Is Not the Same Gearbox

A single worm stage covers roughly 5:1 to 100:1. A single planetary stage covers about 3:1 to 10:1, so a 50:1 planetary unit is three stages stacked, and a 50:1 worm unit is one. That difference is not cosmetic. Three stages add axial length, add cost at each stage, and multiply their losses rather than adding them. A frame that reaches that range, such as ZGX42RM, gets there by stacking.

It also changes what the quoted efficiency means. The figure of 90 to 97 percent often published for planetary gearboxes describes a stage, or at most two. Each stage keeps roughly 95 to 98 percent, and a three-stage unit therefore arrives somewhere around 85 to 93 percent. At 50:1 the planetary is still ahead of the worm, but by less than the headline numbers suggest, and the comparison has to be made at the ratio the machine actually needs.

The Worm’s Efficiency Is a Function of Its Ratio

Worm efficiency is not one number either. It depends on the lead angle, and the lead angle falls as the ratio rises. A low-ratio worm with a multi-start thread can reach the high seventies. A single-start worm at 50:1 or more commonly lands between 40 and 70 percent, and the lower end of that range is where self-locking begins. So the efficiency gap between the two families widens with ratio rather than staying constant.

This matters because a comparison table built at one ratio cannot be applied at another. Two machines needing 8:1 and 60:1 output ratios, both driven by a permanent magnet DC motor of the same frame, are not facing the same trade at all. At 8:1 the worm’s penalty may be tolerable and its right-angle output may settle the layout. At 60:1 the penalty is large enough that the planetary’s extra length and cost are usually worth paying, unless holding position without power is a requirement.

The Motor Is Part of the Comparison

This is the part that comparison tables leave out, and it changes the numbers. A gearbox does not deliver a fixed torque; it delivers whatever the motor puts in, times the ratio, times the efficiency. So the same output torque requires more motor torque behind a worm than behind a planetary, and a permanent magnet DC motor produces more torque only by running slower and drawing more current.

Take a 24 V motor with a no-load speed of 4000 rpm and a stall torque of 60 mN·m, a ratio of 50:1, and an output of 0.5 N·m. Behind a three-stage planetary at 90 percent, the motor must produce about 11.1 mN·m and turns at roughly 3260 rpm, giving 65 rpm at the output. Behind a worm at 55 percent, the motor must produce about 18.2 mN·m, turns at roughly 2790 rpm, and the output lands near 56 rpm.

Two consequences follow, and neither appears in an efficiency table. The worm system runs about 15 percent slower at the same nominal ratio and the same load, because the motor has been pushed further along its speed-torque line. And the motor’s current rises with its torque, by about 64 percent, so the copper loss in the motor rises by a factor of roughly 2.7. The gearbox loss is paid a second time, as heat in the motor, which is the part that is hardest to cool on a small frame.

That is why a battery-powered machine notices a worm gearbox more than the gearbox efficiency figure implies, and why swapping gearbox types late, with the motor already chosen, rarely lands where the calculation said it would.

What the Envelope Decides

Before efficiency is worth discussing, the two families have to fit. A worm output shaft is at right angles to the motor, which is sometimes the only way to get a drive into a shallow housing, and sometimes the reason the machine cannot be built at all. On a unit like ZWL-47FP, driven by an RF370 motor, the output leaves the assembly at ninety degrees to the motor axis. A planetary unit is coaxial, so the motor and the output shaft share one line, which is what a wheel hub, a robot joint or an inline actuator needs.

The length follows from the ratio. At high ratios the worm is one short stage on a right-angle axis, while the planetary is three stages in line and grows with each one added. A machine designed around one of these will not accept the other without changing the bracket, the shaft position, or both, and that decision is usually made before anyone opens an efficiency table.

Backlash and Holding Position

Backlash separates them more cleanly than efficiency does in precision work. A worm drive typically shows backlash in the range of half a degree to a few degrees, because the tooth form and the clearance between worm and wheel require it. A standard planetary unit sits around a tenth of a degree to one degree, and precision grades go below one arcminute. Where the axis reverses direction under load, that difference is visible in the product.

Holding position without power is the worm’s clear advantage, and it is worth being precise about what it costs. Self-locking generally requires the efficiency to fall below about half, because the same friction that prevents the load from driving backwards is the friction that absorbs the input. So the ratio at which a worm begins to hold is also the ratio at which it wastes the most, and the two features cannot be bought separately. The property also belongs to the lead angle rather than to the family name, so it has to be confirmed for the specific unit rather than assumed from the ratio. A worm that holds at 25 degrees Celsius may not hold at 60, because the friction doing the locking falls with temperature, and a change of grease can undo an arrangement that was marginal to begin with. Where the position matters for safety, the answer is to verify it on the actual machine at the actual working temperature, not to rely on the catalogue.

What to Send

Four items decide this comparison, and the first is the one most often left out. Give the output torque and the output speed required at the ratio the machine needs, not at a ratio drawn from a catalogue, since the gap between the two families changes with ratio. State the duty, because the worm’s losses become heat that the gearbox has to shed. Give the envelope available, including whether the output has to be right-angled or coaxial, since that often decides before efficiency does. And say whether the load has to be held with the power removed, and if so how much torque has to be held, because that is the only number that makes a worm worth its efficiency penalty.