Originally Posted by
spdntrxi
so it's within the 2% out the box and your not happy?
it was just outside the 2% originally. even when its 'within 2%' i also don't want the L and R to be at different ends of the spectrum (2% high on right and 2% low on left = ~right value for total but then the LR-balance is inaccurate which would bug me). i am admittedly insane.
really what I'm looking for is it to report the same values as when i had it installed on my old bike, even if it was wrong on the other bike.
Originally Posted by
Duke of Kent
Buy one of the many used SRMs on eBay and be done with it. Send it in once a year for re-calibration.
i just bought a new bike and when i was picking out a crankset i 100% should have just gotten the SRM at that point. next time...
Originally Posted by
Enthalpic
Even calibrated scales have error, especially when working near the limits of their calibration ranges. If you did one measurement on one scale you have no idea of what the true value is. You need the mean and std dev of both values to compare properly.
2% bias is so far from "crap values" it's funny. We pay for "certified" products where the 95% confidence interval is huge.
i have more confidence in the scale's accuracy than any of the other variables involved. but you're right, theres tons of variables involved and plenty of room for measurement error all over the place.
as for my unhappiness with 2% - i agree, 2% would be great, and thats all the error I saw when doing the torque test, but its hard to square with the meter reporting on the order of 10% lower (than both my previous bike and any equation would indicate) in my field tests which is what sparked this whole endeavor. i think my main frustration is that i really trusted the values i was previously getting but now that i've switch bikes i'll constantly be doubting it.