Old 11-22-10, 11:41 PM
  #13  
pjb
don't even
 
pjb's Avatar
 
Join Date: Sep 2010
Location: Chicago
Posts: 123

Bikes: Peugot Conversion, Jamis w.i.p.

Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Likes: 0
Liked 0 Times in 0 Posts
Not really that helpful right now, but...

My school has a class that is working on design of a power meter, I'd take it but no architects allowed

http://ipro.iit.edu/project-listings...Spring2011_324

Description:
The IPRO team is working to develop a system that measures the torque that a bicyclist applies at the crankset of the bicycle and transmits the data to a commercially available bicycle computer. Our goal is to have a design that is suitable for retrofitting existing cranksets for high-end racing bicycles. In contrast to existing solutions we want to be able to retrofit our system to existing cranksets, obviating the need to abandon parts that the bicyclist already owns. According to tests we have done during previous incarnations of this IPRO, this can be done, in principle, using sets of quite inexpensive strain gauges. However, being able to get accurate torque measurements will require some advanced processing of the signals from the strain gauges. These signals can then be transmitted wirelessly to a bicycle computer like the Edge~705 sold by Garmin. There is a defined wireless protocol (ANT+Sport) that has been developed specifically for the purpose of transmitting exercise data (including cadence, power output, heart rate, etc.) to small computers. Chipsets and development kits for this protocol are also available commercially. Our task will therefore be to find an optimal configuration of strain gauges that will be attached to the crankset, to develop an algorithm to process the strain gauge data in order to isolate a signal that is proportional to the applied torque, and finally to demonstrate that the system we are envisioning can provide sufficiently accurate power data to a commercial bicycle computer.
pjb is offline