Automatic Bicycle Transmission
Check it out!! We are on hackaday!! [[1]] Link to our Youtube Video made during our demo: http://www.youtube.com/watch?v=KFmDDD-pwwc [[2]]
Contents |
Team Members
Andrew Toth, Jenna MacCarley, Peter McHale, Nicolas Mellis
Project Description
Cars use an automatic transmission to most effectively change the gear ratio to accommodate driving conditions, without driver interaction. Motorcycles use a clutch and sequential gearbox to accomplish the same thing. Bicycles have no need for a clutch, but also require the rider to change gears themselves.
What if the bike changed gears for you, so you (the rider) can focus on keeping your cadence and go up, down, and around as quickly as possible?
Using magnets and hall-effect sensors, we intend to keep the rider's cadence at 80RPM (a number that is generally agreed upon in avid cyclist groups :: Source: Robert Abramson, CMU Cycling Club) and shifting gears when that cadence drops below 60RPM (and shifting to a lower gear to return closer to 80RPM) or if the cadence rises above 100RPM, shifting up to return to a preferred cadence. Servos are used to automate the gear shifts.
This has been attached to a 2012 Jamis Aurora bicycle, and could be a drop-in attachment to other bicycles with a few adjustments.
Other features that were scrapped in favor of demoing a working implementation within a week include:
- Using the rear wheel's RPM to confirm the bicycle actually shifted gears, to be able to compensate for the slack in the cable that occurs naturally during a bicycle's lifetime
- Switching between all 27 speeds the bicycle has to offer--our servos do not have enough torque, nor do we have time to appropriately experiment other solutions. Other solutions we suggested included: 1- lengthening the servo arm 2-Connecting multiple servo in parallel & securely fastening the cables to the servo 3-Using a worm driver 4-Using a gear system to increase torque
Worklog
- Sunday
*Begin machining the servo mounts
- Monday
*Continue machining the servo mounts *Test whether a servo has enough torque to pull the shift cable
- Tuesday
*Mount the magnets to the wheel and pedals/crankshaft *Mount magnets to the back wheel. *Test the Hall-effect sensors to ensure they work as expected *Machine additional mounts for more servos with intent to combine the servos mechanically in parallel to share torque load between multiple servos
- Wednesday
*Debug Hall-effect sensors *Test the combination of multiple sensors/actuators connected simulataneously *Making permanent connections between things on a protoboard
- Thursday
*More system debugging *System Integration *Definition of System constants
Media
Hackaday: http://hackaday.com/2013/01/18/building-an-automatic-bicycle-transmission-in-a-week/ Youtube of Demo: http://www.youtube.com/watch?v=KFmDDD-pwwc
Files
Fritzing & Arduino code: https://www.dropbox.com/sh/skt3uydbxyfz927/iu0U3UM_RM
Parts List
- Arduino MEGA
- i2c 16-channel 12-bit servo controller(http://adafruit.com/products/815)
- 4 x servos (http://www.pololu.com/catalog/product/1057)
- 2 x Pressure sensors (https://www.sparkfun.com/products/9376) [but let's order like 8, since they're cheap and I foresee them "accidentally" breaking a half-dozen times]
- 4 x Hall Effect sensors (https://www.sparkfun.com/products/9312)
- 8 x Spoke-mounted magnets (http://www.amazon.com/Cateye-Spoke-Magnet-Toolfree-Set/dp/B00180RTYG/ref=sr_1_1?ie=UTF8&qid=1354036565&sr=8-1&keywords=bicycle+spoke+magnet)
- 8 x Neodymium magnets (https://www.sparkfun.com/products/8682)
- Arduino Project box (https://www.sparkfun.com/products/10088)
- 12V Battery (amazon)
