Text post today as I’m still waiting on parts to arrive. I’m anxiously awaiting availability of the nRF51422 Dev Kit and debating going to the Global Tech tour in Boston, MA on the 24th of Sept.
I knew it wouldn’t be long before I’d see Rotor’s new slightly reduced size version of their power crank. Essentially the exact same idea I had and apparently it is licensed from the old MEP powermeter which decided to use Bluetooth instead of ANT+. What I didn’t know is that they were on track with a pedal stroke analysis algorithm which is something I’ve been working on using error measurement analysis theory I developed in my Engineering Masters degree. Apparently they are also working as an ANT+ Alliance member to update the power profile to include this. That’s good as it’ll be a direct implementation to V3 once I get the basic and crank profile code cleaned up. Though I often wonder why we haven’t seen an influx of Chinese companies since most designs are based on an expired patent and therefore not patentable.
It does look like Rotor is taking a dig at the Quarq and SRM decoupling Algorithm here by saying:
“Independent positive and negative power output. unsurpassed accuracy due to symmetrical crank deformation with 8 strain gauges (no need for extensive calculus or data ltering).”
What does this mean. Well, it’s just saying they (and by extension, my design) doesn’t require a more complicated decoupling algorithm. I suspect they mean “altering” and not “ltering”.
Pioneer has released a new attempt at a power meter which I have huge reservations about. This is not about the “how” it’s just about the glue — literally the glue. Strain gauging can be done in a few ways. “Lick-and-Stick” which means Cyanacrylate glue, a high quality crazy/superglue, or higher temp cured epoxy. Well the second won’t be happening as a LBS will need to bake the bike crank for up to eight hours. Improbable. There are almost room temperature cured epoxies that require specific clamping pressure but generally not recommended for transducer quality installations. Essentially you are trying to get a LBS to do a precision technician job. I think that this could lead to people with poor reading power meters if this gets to market.
Looks like someone was working on the same thing as I did but put in more than the two evenings of two to three hours of work. What I’m referring to is this and my own Arduino style version here. It’s Bluetooth smart, which I have no idea about. I suspect it’s the sports implementation of Bluetooth LE. It uses a similar algorithm as I previously indicated of
Power = regressed curve + Flywheel mass moment of inertia * angular acceleration * mean angular velocity over sample period
and in a coast down this becomes
regressed curve/angular velocity = Flywheel mass moment of inertia * angular acceleration
More updates to follow. If I can get the bike crank this coming week I’ll do some documentation on how to install strain gauges.