On 06/07/18 01:48, AMuzi wrote:
On 7/4/2018 5:28 PM, James wrote:
On 05/07/18 02:48, jbeattie wrote:

You can buy Stages two-sided, if you're into accuracy,
which may be an issue if you're switching between bikes or
really concerned about accurate calorie burn numbers-- or
are in some sort of competition with other riders (who may
or may not be posting accurate power data).Â* Otherwise, I
think consistency is more important than accuracy. Spikes
can be filtered with data ceilings, etc., but the spike my
son had -- which was actually 1,732 watts is within the
one second power output of Chris Hoy.Â* It was recorded by
having the right crank forward on a bumpy descent and not
by massive quads, but the meter is just recording numbers
accurately. It's not going to filter that out.Â* Maybe he
is Chris Hoy! You can download the data and apply filters
later.


Is the crank position (cadence) sensor on the left crank and
left chain stay?

I wonder if the unit "thought" the cranks were rotating
because the rotation sensor recorded some jitter at the same
time the load cell did?

To eliminate this by design, a better crank position sensor
would help. An encoder ring on the crank and sensor on the
BB, for example.Â* Then you don't need to try to "fix" the
data afterwards, which I always see as a giant fudge for
poor engineering design.



Those are usually strain gauges inside or on the arm, not a
magnet/sensor interface to a chainstay sensor like a cadence counter.


The strain gauges are for the torque measurement, but to calculate power
the device needs to know the angular velocity of the crank as well.

The stages unit has an accelerometer for that purpose, as Jay noted, and
obviously that solution isn't fool proof because the data has erroneous
values.

--
JS