The basic cadence sensor uses a magnet on the crank, it turns the motor ON when you start pedaling and turns it OFF when you stop pedaling. It works more or less like a switch.
Using the cadence sensor, you have to control the boost level and speed by adjusting the assist mode manually up and down. Most basic e-bikes have this.
The advantage of the cadence sensor is that it’s not expensive but the disadvantage is that the pedal assistance can feel response delayed.
The torque sensor is a totally different technology that uses a precision strain gauge. It measures your actual force on the pedal, sampling at 1,000 times per second over the entire pedal stroke.
The harder you pedal, the more power it gives to the motor. If you pedal lighter, less power goes to the motor. It makes this adjustment in real time so it is technically amplifying your every input. It feels like you are bionic.
Most high-end e-bikes use torque sensors . In some countries, cadence-only sensors are not allowed to be sold. The disadvantage is the cost is significantly higher to implement this technology as the precision component is relatively more expensive.
Torque sensors “feel” the pressure applied to the chain, irrespective of the bike’s condition (stationary or moving) and then allow the electricity to surge from battery to the motor and control the power output continuously.
The cadence sensor can be a bit tricky when you are in situation, let’s say, traffic signal somewhere in the middle of a hilly road and now you need to rotate the crankset 2-3 times before the motor starts assisting you, the torque sensors feel that the chain is getting tighter and immediately help you