News: Blowing our own trumpet
Thursday, February 10th, 2011, by Barry
If we don’t blow our own trumpet then it ain’t gonna get blown!
The pod on one shoe combines the values and sends them to the display.
Part of our realtime data display system.
A sample of what the raw data from our motion sensors looks like.
The Brim Brothers power meter measures not just power, but also cadence and a big chunk of other details about what’s going on with your feet and your pedals. It measures forces, pedal angles, crank angles and speed of rotation 100 times per second and uses this to calculate torque, power and cadence. The pods on the left and right shoes make these measurements and calculations independently and then communicate directly with each other over the ANT wireless system. One of the pods acts as the master by combining the values and transmitting them to an ANT+ compatible display. It’s all just plug-and-play, with nothing to be configured. The pods automatically decide which of them is the master, so there’s nothing to be configured for that.
If there’s only one pod then it will automatically double its own power measurement (as an estimate of the power on the other pedal) and send that to the ANT+ display. That means that if you use just one pod instead of a pair you will still have a usable power meter. By the way, the pods on each side are identical so it makes no difference which way round you use them. Each pod automatically works out whether it’s on the left or the right shoe. A major design goal in developing the power meter was that it should be easy to own and use, with absolute minimum configuration and calibration (and I’ll talk about calibration in another post soon) so we’ve come up with clever ways for the pods to work out things for themselves.
While the main purpose of the power meter is to send torque, power and cadence data to an ANT+ display, the detailed data stream is extremely interesting to look at. This gives us extra ways to see what’s going on as you pedal. The simplest is to look at how you apply force as the cranks rotate. Theoretically a rider should apply the maximum push-down force as the crank is half way through the down stroke, but we could see in our realtime graphs that our test rider was applying force very late and then continuing in an attempt to stretch the cranks at the bottom. Seems like he might benefit from adjusting his pedaling style a little, particularly by moving the application of force to slightly earlier in the pedal stroke.
Watching these graphs in realtime while cycling on an indoor trainer is quite intriguing. You can see immediately the effect of trying to, for example, apply force earlier in the down stroke. During these tests we had a discussion about the mental drivers and triggers that set the way you apply pedal force, and we went on to discuss how a rider uses the beat of music when training indoors and whether this could be used to improve pedaling efficiency. From observation it was apparent that our test rider pedals such that the music beat occurs just after the crank passes the horizontal position, so we did a short unscientific experiment where he consciously changed so that the beat occurred just before the crank passed the horizontal position. The effect was immediately obvious on the realtime graphs – the point of maximum force moved back in the crank rotation. We didn’t do any experiments to see if this resulted in any measurable improvement in efficiency (how would we quantify that anyway?) but I’m sure there are lots of knowledgeable people out there who will do these experiments as soon as we deliver the technology to them.
There is lots more stuff I want to share with you, but I’ll leave that for another day (hopefully within a week or two). Two things to finish off today. First, we have a 45 second video introducing ourselves. It’s kind of a bland corporate thing so probably only of interest to you if you’ve only just found us (and I swear that’s the last time anyone will get me in front of a video camera). Second, you may notice a blog post here called “The problem with the Kionix KXSD9” that’s password protected. The article is a very technical explanation of a serious problem we had with a component, and is not open for public reading at the moment. If you have a particular interest in the KXSD9 contact me.