How accurate is the sensor?

The primary difference between the PowerCurve Sensor and a torque based power meter (like a PowerTap, cranked based power meters) is that the PowerCurve Sensor determines the power for the speed to power relationship of a specific trainer.

There are a few considerations to be aware with the speed to power relationship (the “power curve”) of a trainer:

  1. The calibration precision
  2. The “drift” as heat builds up and dissipates
  3. The manufacturing differences between trainers of the same model
  4. Sudden acceleration and deceleration

The calibration is done using the coast down time of the trainer. There is a small tolerance for the coast down time, and trainers with a larger freewheel have a longer coast down time, so there will be less variability in the calibration.  Another variability is the weight of your rear wheel compared to the wheel that was used to measure the power curve.

The “drift” as heat builds up can change the power curve itself. First, it is important before calibration to warmup the trainer and tire for 10-15 minutes. For wind trainers, resistance is created from the ambient air, so there should be very little heat drift. Magnetic trainers probably also don’t change their resistance too much with changes in temperature.

Fluid trainers could be affected by drift. First, the oil in the resistance unit will heat up and could make it less dense. Some manufacturers like Kinetic say that their oil is very “stable” and does not drift. This is probably true as the power curve measured for their fluid trainer was extremely close to their published power curve.

Fluid trainers tend to have “fins” to dissipate heat so the temperature should average out during the workout. So secondly, the type of workout, for example, steady state time trial or short sprints could also affect the oil temperature. Your average power too will also affect how much heat is generated.

Tire heat could also affect the resistance with the trainer roller. This stresses the importance again to warmup the tire and trainer before calibration. Different tire compounds will have different characteristics.

Trainers of the same model may have different physical characteristics. For example, who knows if a manufacturer has used the same oil in every production run? There might be slight changes in dimensions, magnet placement, bearings, etc…

Fast changes in speed will affect watts with a PowerCurve sensor. For example, if you do a 10 second sprint from a standing start, the power will be much lower on a PowerCurve. That is because the watts are determined from the speed, not the torque. On a 10 second standing start sprint, the maximum speed is not reached at maximum torque. It’s like having a lag, where the speed lags the torque. On the other hand, if you do an all out sprint from a rolling start at 30Km/H, the watts will be accurate because the maximum watts will occur very close to the maximum speed. During deceleration, the reverse happens, the watts will be higher and lag the speed. The average watts will remain accurate because for each acceleration, there will be a deceleration. What goes up must come down.

Even with all these considerations, during tests when comparing with a PowerTap, the discrepancy was usually very small, around 5-10 watts. In summary, don’t forget to warmup the trainer and tire. If the resistance unit generates heat, placing a small fan on the unit could keep things more consistent.