View Single Post
Old 07-03-2019, 12:57 PM
Mark McM Mark McM is offline
Senior Member
Join Date: Jun 2006
Posts: 6,364
Originally Posted by weiwentg View Post
Well, I don't make powermeters, and I'm not any sort of engineer, so this answer is off the cuff. The issue there is that they don't know how much the fudge factor is. The amount to fudge the reading probably varies in real-world conditions - GPLama himself said that the differences aren't as pronounced in mixed terrain, and they seemed to get worse with increased power.
If the "fudge factor" was a simple offset and/or linear multiplier, than it would be intrinsically included in the calibration. I suspect that it is more complicated than that. I suspect that it is not a matter of non-linearity (strain in metal is very linear below the yield point), but instead is a matter of sensitivity to off-axis loads. And off-axis loads can vary quite a bit with pedaling technique and terrain.

Originally Posted by weiwentg View Post
This is a much more detailed critique than what Wakeham posted on his blog (I understand that Wakeham may have a YT vid that goes into more detail, I haven't watched it). This does raise a question. I don't buy the title of Wakeham's post, that Shimano made a design blunder with their cranks. They are, after all, a bike component manufacturer. However, bikes are headed towards greater integration of parts. Lots of people are putting power meters on their bikes. Should Shimano design their crankarms more symmetrically on account of this? What will pressure them to do so?
Left cranks are well suited to torque measurement, because they are long, relatively uniform beams, which will have uniform strain fields in the middle of their spans (which is why you'll find most left crank power meter strain gauges mounted right in the middle of the arm). There will still be some off-axis load sensitivity here, but this can be cancelled with a few extra co-located strain gauges.

A right crank has spider arms sprouting from it, often some distance out the crankarm. Loads are transferred from the crankarm to the spider arms in different proportions, depending on crank position (and off-axis loads). The affect of the spider arms can be mitigated somewhat by moving the strain gauges further out on the crank arm. Unfortunately, you'll then run into space constraints, as there may not be enough clearance between the crankarm and the chainring and front derailleur cage.

Right crank power measurement can be accurate and repeatable, if the crank arm is designed properly. But that would take a great departure in shape from the current Shimano crank design.
Reply With Quote