Why should the motor cable be run twisted?
The magnetic fields of the outgoing and returning conductors cancel; twisting repeats that cancellation at every pitch, cutting radiated and received noise.
The motor cable is run twisted because the current flows in opposite directions in the outgoing and returning conductors and the magnetic fields they produce cancel each other. Twisting the two wires repeats that cancellation at every twist pitch; the radiated noise falls considerably. For the same reason a twisted pair also stands up better to magnetic interference arriving from outside.
Where the noise comes from
In a brushed DC motor the brush arc produces broadband noise, while a switching driver creates fast current changes. Two cables run straight and parallel behave like an efficient antenna, because the field between them stays one-directional. The cable itself is not the source but it is the radiator; that is why the answer is sought in cable geometry. The typical symptom is the remote control cutting out the moment the motor runs, the analogue reference wandering, or a counter or sensor nearby giving a false reading.
What twisting solves
- Radiated noise: At every twist pitch the field direction reverses and cancels against the neighbouring pitch.
- Received interference: An external magnetic field induces voltages of opposite sign in successive pitches, and the sum converges to zero.
How it is applied in practice
Rather than giving an exact number of turns, look at this criterion: the outgoing and returning wires should be as close to each other as possible everywhere along the cable, and the gap between them should stay constant along its length. Keeping an even pitch by hand is difficult; fixing the cable at one end and turning it from the other gives a pitch that looks regular to the eye. Nor is very tight twisting needed: a twist that stretches the insulation and tires the copper creates a risk of breakage over time.
- Twist the two motor leads as a single bundle with a regular pitch; twist the supply pair the same way. A pitch that is irregular and opens up here and there does not give the cancellation you expect.
- Apply the twist over the whole cable. Wires that open out over the last hand's breadth into the terminal give back a good part of the benefit gained.
- Do not route the outgoing and returning conductors along separate paths; as the area between the two grows, the twisting loses its point.
- Keep motor cables at least 50 mm away from analogue and RF signal cables; where that is unavoidable, cross them at a 90° angle.
Next steps if twisting is not enough
Twisting reduces magnetic coupling, not capacitive coupling. If the problem persists, use shielded cable and ground the shield at one end only, preferably on the driver side; grounding at both ends creates a ground loop and 50/60 Hz interference. Suppressing the source is possible too: a capacitor or a ferrite bead fitted at the motor terminals directly reduces the noise produced by the brush arc.
What twisting does not solve
- A poor chassis connection. Noise riding on the common return resistance has nothing to do with cable geometry; the chassis and the grounding are put right first.
- The antenna in the wrong place. Even a twisted cable raises the noise floor if it is resting against the receiver antenna. Move the antenna away from the motor cable.
- External sources. Against strong sources such as an inverter or a welding machine, twisting alone is not enough.
In installations with an RF receiver (the ST42 and PT500 can be used with a remote control) these details affect the range directly. When you have finished the wiring, run the motor under load and check that the remote control responds from the furthest working point; if the range shortens while the motor turns, the problem is in the cable route.
The full topic: DC Motor Driver Selection Guide