Power meters on road bikes

The bike in the diagram is the same model as I ride. It is the best fit and feel for a road bike that I have ridden in the past several years. Started with an entry level Cannondale R700, then tried Italian with a Bianchi 1885, briefly used a Surly Cross-check for a New Zealand bike tour, rode a full carbon Bianchi 928 for a couple of days, and finally bought a used Lemond Tete de Course from eBay. The combination of titanium and carbon produces a very comfortable ride. Together with my Brooks B-17, never need to use bike shorts, and saddle sores are a thing of the past. All the previous beasts were sold to engineers on my team and we train together a few times a week as a team-building exercise.

I like testing out new bike gadgets--it appeals to my engineering mind and the lack of a car in HK together with extremely low personal taxes (15% is the maximum tax rate), I have enough funds to satisfy this expensive hobby. About a year ago, I tried the iBike unit. It mounts on the handlebars and measures the power by measuring the wind resistance and subtracting the bike speed. Although cheap, I found the unit to be very difficult to use and seemingly un-accurate. For example, wind speeds were displayed as almost double or triple the actual wind speed (there is no way I can pedal 30 kph in a 60 kph wind). The power readings were also erratic. In the end I stopped using it as it was quite large and not the most attractive thing on the handlebars. I then tried out a PowerTap wheelset on my Lemond Tete de Course. It worked wonderfully until I was caught in a heavy rainstorm and it has now ceased to function (I did remove the cover and replaced the internal batteries, but the unit still doesn't respond).

As for the Ergomo and the SRM, they are not only prohibitively expensive, but also require more advanced mechanical knowledge to install. Plus, I am perfectly happy with my current bottom bracket and crankset--both work and from my short experience as a cyclist, if your BB + cranks work, do not replace them or try to fix them. I have heard bad reports about the Polar unit regarding installation and accuracy (though it is rather innovative in using the frequency harmonics of the chain in order to determine torque and therefore power). Will most likely test out the Quarq unit next for two reasons: minor modification to my existing setup (already have Dura-Ace spider arm cranks) and it uses open source software. This means I can go in and make any modification I want to the system, customizing to my needs. The combined setup (computer + torque unit) is rather expensive, but it uses an open wireless standard so it can conceivably work with other bike computers, like the upcoming Garmin 705. I also used the Garmin 605 unit. It was pretty good and fairly robust; except I could never get it to recognize its own speed and cadence sensors (apparently a common problem).

As one can see from the bike diagram, there are several locations on the bike for measuring power. There are still several locations left for a set of enterprising individuals and I will explore each location in upcoming posts.

Web-links:
1. PowerTap: Can purchase from www.jensonusa.com
2. Polar: Amazon.com offers best prices
3. Ergomo: eBay has best prices
4. SRM: eBay has best prices
5. Quarq: Product will be released in December 2007
6. iBike: Can purchase from www.speedgoat.com

Lung Infection & Non Rectangular QAM

Seemed to have contracted a nasty lung infection on 29 September. Had to stop training and been off the bike ever since (with a 3 day exception in the middle of the month averaging 20-30 km). None of the doctors have been able to figure out which bacteria it is; except that it is not one of the common ones. Zithromax seemed to help a little bit last week, but started regressing again when it wore off.

While staying at home for some people may seem like a vacation, it starts to feel like jail for awhile. After starting and completing Command & Conquer (Tiberium Wars), gained enough mental facility to consider an interesting problem somebody posed to me at work. It has to do with constellations and QAM. For QAM, engineers use 4, 8, 16, 64, .... QAM for signal modulation. The friend was thinking of using hexagonal tiling instead, but some group published a paper last year. I suggested to use a Penrose Tiling. It uses quasi-five fold symmetry and initial results does show that it is more efficient than other symmetries (assuming the size of the n-fold polygons are the same with elements at the vertices). In terms of the inter-element spacing, n-fold polygons with smaller n are more efficient (completely logical). There is probably some horribly complex math that can be done to prove all of this, but Apple Keynote worked fine enough for me. What's interesting is how everything works in circles....The 5-fold symmetry is a problem I spent a year doing at UCSC when working on my thesis: 3D Visualizations of Quasi-Crystals (published a couple years later in a Physics journal).

For more information on Wikipedia:
http://en.wikipedia.org/wiki/Quadrature_amplitude_modulation
http://en.wikipedia.org/wiki/Constellation_diagram