Comparison of SRM, PowerTap, and CompuTrainer Power Measurements

Many cyclists now utilize, or even depend upon, power measurements made during training. Power can be measured on the road using a device like the SRM (which measures power at the cranks) or the PowerTap (which measures power at the rear hub). CompuTrainers also provide a power reading; however, the number is calculated according to a load curve programmed into the trainer. The advantage of the PowerTap and SRM is that they may be used outside while training and racing. Unfortunately, expense and upkeep prevent many riders from installing power meters on personal bicycles (the devices range in price from $1000-$3500 and must be maintained as any bicycle component). A CompuTrainer offers the advantage of recording each session regardless of the components on the bicycle -- every cyclist training in our lab has the opportunity to train with the use of power measurements.

A common question in the lab: "How does the CompuTrainer power calculation compare with the PowerTap or SRM?" Although the engineering for each device is different, our experience is that all three power meters yield readings that are remarkably close. We frequently perform CompuTrainer training sessions in tandem with an SRM to verify that the numbers agree. This testing also ensures that each of our eight CompuTrainers is behaving like the others. For accurate comparisons, each meter must be calibrated and/or zero'd according to manufacturer's instructions.

To demonstrate the consistency of different power meters, a CompuTrainer step test was performed with both an SRM professional crank (Dura Ace) and a PowerTap Pro installed on the bicycle. The test protocol began at 100 watts and increased 25 watts every two minutes. Results from the SRM are shown in red, the PowerTap in blue, and the CompuTrainer in black (see graph and table below). Note that the SRM numbers are consistently highest. Because power is lost in the drive train, power delivered to the crank is more than power at the hub. Similarly, it is impossible for power measured at the hub to be less than power at the CompuTrainer, even though the PowerTap numbers are sometimes 1-2 watts below the CompuTrainer readings. Each CompuTrainer is dynamically calibrated on a dynamometer, however, and without knowing the exact calibration procedure, one can not be certain where the power will fall for CompuTrainer compared to the other power meters. Each power meter claims its own margin of error: 1.5% for the PowerTap Pro, 2.5% for the SRM Pro, and 2% for the CompuTrainer (Lab version). For our demonstration, all meters were within 2% of each other at 150 watts and within about 1% of each other at power outputs above 200 watts. Note: It is critical that each power meter user appropriately calibrate their meter to achieve consistent results. This particular test was performed at random and is indicative of results obtained from similar tests performed monthly at Incline Training.

In the final analysis, the numbers from each meter are identical within claimed margins of error and certainly close enough to make useful comparisons.