Monday, August 26, 2013

Accuity V4.1 Working Well; Safety First!

I went on my third ride last night (fit file here) with zero dropouts, but there is one spike for some reason I’ve yet to determine. The Garmin connect link is here for those who are interested. To do is check the calibration. To get to a minimum viable product I need to finish the enclosure and auto shutdown when not pedalling for a few minutes. 150 hours of life (estimated) doesn’t seem that long when it spends 22 hours a day not in use.

I made a video tonight explaining how pedal based power meters such as Garmin Vector and Look Power Pedal work. It also explains how Resonant String Gauges for the Swedish Adrenaline indiegogo project likely work (here). The video is still uploading, and it’s fairly basic whiteboard video. It’ll be here later tonight. Direct link / post tomorrow.

I want to apologize for taking video in an unsafe manner for Accuity V4.1’s first ride. Please don’t do the same. It was stupid and reckless endangering myself unnecessarily. If you must take video buy a cheap camera mount or a GoPro. I ordered a mount today to take some video. It’s cheap from China but I suspect it’ll do the job.


  1. Looking good. There's a handful of not-ridulously-high-but-still-wrong power spikes in there too, which all appear to happen when the left leg is [apparently] producing 100% of the power:

    Note the faint dots (individual records) at the top of that chart (left leg 100%), whereas there aren't any on bottom after 200W.

  2. I think this is the same anomaly I was seeing while riding. I'm still trying to determine if it's from the first partial or the first full rotation, but always after I stopped pedalling for a little big. I noticed it always skewed left and then gave a buggy result. I'm working on detecting when rotation has stopped and resetting the reading of the ADC until it detects rotation again.


  3. Thanks for the video regarding the pedal power meter strain measuring designs.
    From the video I can see the need to measure the moment and the offset. But I'm not sure I understand how the second set of gauges measures that moment/couple (Maybe I need to watch the video again). Are the second set of gauges located at the very end of the pedal spindle and therefore more sensitive to the couple and less sensitive to spindle bending moment? Or is it more to do with how the gauges are orientated on the spindle surface and gauge orientation is optimised to detect twisting rather then bending?

    1. I haven't done an official post about the video yet. There is some accompanying diagrams that I'll include in that.

      It's more about how they are oriented. When you bend a beam the strain on the top and bottom is related to the bending moment. If you apply a moment at the end it's constant throughout. So if you use a differential setup it doesn't matter the location of the force or if there is a moment applied, it rejects both of it.

      Constant moment = constant strain, so they are both offset.

      If the force is applied after the two gauges, they both increase proportional to one another. So with 1 Newton of force at 1cm from the second gauge you might read 40 and 20 microstrain. Move it to 2cm and it goes up to 80 and 60 microstrain. Electrically they can be setup to subtract from one another giving the slope. So in each instance you read only 20 microstrain. Applying a bending moment is similar, it's rejected as well.

      I'll try and do a better job explaining it when I write up the accompanying blog post.