Force sensors measure linear push or pull (straight line). Think of a spring scale pulling a fish – that’s force. Torque sensors measure twisting (rotation). Think of a wrench turning a bolt – that’s torque.

In everyday language: force = pushing; torque = twisting. They use similar strain‑gauge technology but different mechanical designs. A force sensor cannot measure torque, and a torque sensor cannot measure linear force. However, some multi‑axis sensors measure both forces and torques (e.g., 6‑axis force/torque sensors).

Yes, but only indirectly and with lower accuracy. Common indirect methods:

• Motor current measurement: For electric motors, torque is roughly proportional to current. However, friction, temperature, and motor efficiency vary, so accuracy is typically ±5–10% – fine for simple monitoring, not for precision control.
• Strain gauges on a shaft: You could bond strain gauges to a shaft and build your own bridge circuit. This is technically possible but requires calibration and slip rings or telemetry – a complex DIY project.

Torque sensor accuracy is usually expressed as a percentage of full scale (FS). Typical ranges:

• Laboratory / calibration grade: ±0.05% to ±0.1% FS – very expensive, used in metrology labs.
• High‑performance industrial: ±0.2% to ±0.5% FS – for engine testing, aerospace, precision assembly.
• General industrial: ±0.5% to ±1% FS – suitable for most factory automation, process control, and torque auditing.
• Basic / hobbyist: ±1% to ±2% FS – affordable, okay for DIY or rough indication.

Remember: accuracy is affected by temperature, misalignment, and overloading. A ±0.5% sensor installed poorly can easily produce ±5% errors.

Most manufacturers recommend annual calibration for critical applications. However, follow these guidelines:

• Annually – for quality control, legal testing, or any application where readings affect safety or payment.
• Every 2–3 years – for non‑critical monitoring where trends (not absolute values) matter.
• After any overload – if the sensor has been subjected to a torque above its safe limit (typically 150% of rated capacity), recalibrate or replace it.
• After physical damage – dents, corrosion, or heavy vibration can shift calibration.

Yes, most torque sensors are bidirectional. They measure torque in both directions, with the output sign indicating direction (positive for clockwise, negative for counter‑clockwise).

Check the datasheet: some low‑cost sensors may be unidirectional (only positive torque). Also note that accuracy may be slightly different in each direction, though good sensors specify bidirectional accuracy.

Overload protection is a mechanical or electronic feature that prevents the sensor from being damaged by excessive torque. Some sensors have integrated mechanical stops that physically limit deflection once a certain torque is reached. Others rely on you – the user – to stay within the rated range.

Even sensors with overload protection have limits. The common specification is:

• Safe overload: typically 150% of rated capacity – you can go up to this without permanent damage.
• Ultimate overload: typically 300% – the sensor will break at this point.

Do you need it? If your application has occasional torque spikes (e.g., startup surges), choose a sensor with a high safe overload or add a mechanical torque limiter to your system.

Yes, but they are less common than contact (strain‑gauge) sensors. Non‑contact methods include:

• Magnetostrictive: Measures changes in magnetic properties of a shaft under torque. No strain gauges glued to the shaft – the shaft itself is the sensor.
• Optical (laser or encoder): Measures angular twist between two ends of a shaft using optical encoders. Very high accuracy, but expensive.
• SAW (Surface Acoustic Wave): Wireless, passive sensors attached to a shaft that are interrogated by an RF reader.

For most industrial applications, strain‑gauge rotary torque sensors with slip rings or telemetry are still the standard. Non‑contact types are used in specialty cases (e.g., MRI‑compatible, extreme speeds).

Prices vary greatly by capacity, accuracy, output type, and brand. Entry‑level hobbyist modules (e.g., strain‑gauge amplifier + DIY shaft) can be under $100, but they are not calibrated or rugged. For reliable industrial use, expect to invest accordingly.