I did a couple of test sessions over the past few days with a powertap and today realised that the torque offset was not zero. I’ve downloaded the files to cycling peaks and was wondering if there’s any way to adjust the data so I can salvage the tests. If anyone can save me from having to re-do the tests, I’d really appreciate it!
I did a couple of test sessions over the past few days with a powertap and today realised that the torque offset was not zero. I’ve downloaded the files to cycling peaks and was wondering if there’s any way to adjust the data so I can salvage the tests. If anyone can save me from having to re-do the tests, I’d really appreciate it!
Depends on whether the torque drifted over the course of the test or was constant.
Export the file to .csv then import to a spreadsheet. Examine the hub torque column for segments with zero cadence. If the minimum value you observe while coasting is constant (or nearly so) the file is salvageable. Otherwise, you’re SOL.
If it’s constant then you have a chance. First, calculate the ratio of watts/(hub torque*km/h). Verify that it’s a constant for all entries. Second, subtract the constant amount from all the hub torque values to re-zero them. Third, combine the zeroed hub torques with the speed column using the ratio in step one to recalculate the corrected power.
Next, repeat this all again with your eyes closed. You need to do this last step to be sure you’re blinded to the results.
Thanks for your help! It looks like the torque didn’t drift, so I was able to adjust the file.
I need to work on my blindfolded typing though.
How much variation is acceptable? I did the export like you suggested and analyzed my 3 hour ride from today:
8261 readings
141 @ zero cadence
30 of those had wattage other than zero - all positive from 9 to 366.
Is this file any good? i.e. are my avg watts etc accurate?