Archive for the ‘Uncategorized’ Category

This week, I have been working on the mechanism within VECTr that will multiply the relatively short distance (4 mm) that a Shimano index shifter pulls to the greater distance (11 mm) of the VECTr settings. This is a critical part of the VECTr gear that will make it compatible with existing index shifters, such as a Shimano 7-setting shifter.

Index Mechanism 9.30.15

Work in progress

The latest version of my expanding chain ring idea is called VECTr, short for Variably Expanding Chain Transmission. How does it work? Find out here:
The key to the design is the simple locking mechanism which performs two functions: (1) holding the gear segments in the desired radial position while they are engaging the chain, and (2) unlocking from that position and moving to a new one when the rider moves a gear shifter.  The current design has a simple pin going through the gear segment and into a locking notch on the base of the device.
Sept 25 2015
Simply by being rotated as the rider pedals, the pin hits a curved control plate, gets unlocked, and slides along the contour of the control plate moving the gear segment to a new position.  When free of the control plate, a spring pushes the pin back to lock the gear segment in place so it can transfer power once again.
As you can see from previous posts, simplifying the design was not a very simple process.  I have found that the first, or most obvious (to oneself), way of solving a problem is seldom the simplest.  I tried to envision what I wanted my device to do, imagine a means of doing it, and then try to eliminate as many parts as I could while allowing the device to function.  The result of this long, meandering process is VECTr.
VECTr is a simple design that really works. The video on my website shows a very basic working model to illustrate proof of concept. The next step in this process will be to develop an actual prototype that can be safety tested. But, there is a little tweaking here and there that remains to be done. VECTr is patent pending and has attracted over 9,500 views on its You Tube channel, and has received some very positive feedback so far.
Now you know the long and winding path that led to VECTr. From here on out, this blog will feature updates on this work in progress. Share your thoughts on VECTr, and join the ride!

After realizing that the Continually Adjustable Bike Transmission (CABT) model was just too cumbersome to be appealing, I could not let the idea go. I still thought the problem of an expanding chainring system should be solvable, but it would take a new approach.  As my wife will gladly tell you, I am often guilty of over thinking things and making them more complicated than they need to be. So, I ditched the CABT design and started over from scratch.

I tried to think of the simplest way to get the gear segments both to be adjusted (i.e., change radial position) and  to lock into that new position. First, I decided that having gear segments slide along straight arms radiating from the center of a circle was the simplest way to for them to change radial position. (This is far simpler than my use of spiral-shaped gear segments being moved by a pin sliding within them, which thereby changed their radial position).

The other problem was how to allow them to lock into position while engaging the chain, and unlock and change position when the bicycle’s rider wishes to change gears. I got rid of the threaded screws that changed the position of the gear segments in the CABT model. I went through several designs for the locking mechanisms and tried building and testing them, using simple materials and working in my garage.

Initially, I thought the gear segments could be locked by means of a toggle which fit into a notch on the edge of the radial arms. I tried simple pivoting brackets, and an elaborate scissoring mechanism:

sketch 1

Toggle locking mechanism

The another design involved having pins pinch the radial arms when depressed by the force of the chain:

sketch 6

Pin locking mechanism

Neither of these seemed like they would work sufficiently well, and they were too complicated. So, I thought it would be most feasible to have a pin perpendicular to the plane of the base plate serve as a locking mechanism.  I thus had to figure out a way for this pin to move in and out as directed by the rider:

sketch 2

Perpendicular pin locking mechanism

Next, I tried a design inspired by clothespins:

sketch 3

Clothespin-inspired locking mechanism.

After that, I tried having the toggle swivel when making contact with the control plates, which would raise the locking pin from its notch:

sketch 4

Swiveling toggle locking mechanism

As many designs as I came up with and tried to build, they all seemed too complicated and/or heavy to be commercially feasible:

button-pin photo

Not it either!

It wasn’t until I thought of putting the locking notches within the groove along which the gear segments slide that I made the step which led to VECTr in its current design:

sketch 5

The penultimate design

After realizing that the gear segments had to be able to change radial position during the quarter of each revolution when they are not engaged in the chain, I also saw that I still had to work out the mechanics of how the gear segments would move. And this required framing the problem in a new way yet again.  Now the problem was: how can the gear segments change radial position when they are free from the chain, yet maintain their new position (and not slide inward) when they do engage the chain?  The problem was how to make the gear segments move when the bicycle’s rider wanted them to, yet remain fixed so as to serve as an effective power transmission.


CABT drawing

My next bright idea was to use long bolts radiating out from the center of a base plate, and affixing on them specialized nuts which would slide in a groove of the gear segments.  As the bolts turned, the nuts would be driven inward or outward, and the gear segments would expand or contract.  The turning of the bolts would occur as wheels on their outer end rubbed against control plates suspended next to the bolts as they pass the plates on each rotation of the crank.  Thus, depending on which side of the bolt rubbed against the control plate, the bolt would turn clockwise or counterclockwise, and the gear segment would pivot outward or inward.

Original CABT working model

Combining these ideas, I designed a device called the Continually Adjustable Bike Transmission (CABT) and set about trying to build and test the design.  After even more effort and many trips to hardware stores, I built a working model of that design.  I was quite pleased and excited that the CABT worked.  I even filed a provisional patent application and tried to interest various bicycle component manufacturers in licensing it. Unfortunately, I found no takers.  An engineer directed me toward Frank Berto’s book, The Dancing Chain, and becoming familiar with the history of the bicycle transmission, I realized that the CABT was too heavy and complicated to be marketable.

Back to the drawing board again. All of us have had to start over at some point, right?

In 2008, I started working in earnest again on the bike transmission. My wife and I had bought a home in the suburbs by now and were both busy with our careers. I ended up working at Sam Houston State University in campus ministry and teaching Philosophy, though the mechanical problem-solving part of me was still largely unfulfilled.  I channeled that desire into the bicycle transmission.

For my part, I kept trying to solve the problem of making a chainring expand by designing a device that had the sprocket broken up into gear segments hinged at one end and simultaneously pivoting them to increase the radius at which they would engage the chain.  As I looked into prior patents on expanding chainrings (and earlier “expanding pulleys”), I found that people have been attempting to solve this problem for almost as long as bicycles have existed.  Frank Berto, on p. 53 of his magisterial book The Dancing Chain, cites the Protean device from 1894 as the earliest commercial attempt to manufacture an expanding chainwheel system. (In another post, I will outline other earlier attempts.)


Examples of historical patents

For over 100 years, many previous inventors had attempted various ingenious ways of expanding a gear’s radius, yet almost all of them involved making all the chain-engaging elements move at the same time. Apparently none of them had been commercially successful, however, since none of them were available today.  And my own limited tests with the crude models I could construct seemed to indicate that there was a problem in having gear segments move or expand simultaneously.  While distance between the teeth of the gear segments remained fixed relative to each other, the distance between the gear segments (and so the teeth of adjacent gear segments) would change while the pitch of the chain would not.  (To address this issue with VECTr I recently made a particular refinement to my present design.

This new insight into the limitation of the idea I was pursuing allowed me to frame in greater specificity the expanding chainring problem: how could the radii at which the gear segments engage the chain grow or contract while still fitting into a chain whose links did not vary their pitch?  In one moment, while driving down a back road in Central Texas, the idea hit me that the gear segments could change their radial position (in fact they would have to do so) only when they were not engaging the chain.  I realized that on every turn of the crank, there was a quarter of the chain-and-sprocketsrevolution when the gear was free of the chain.  This is the space between where the chain comes to the chainring from the rear sprocket and where it leaves the chainring to return to the rear sprocket.  It had to be in this quarter turn of the crank that the gear segments could change position without hindrance from the chain, but they could still engage the chain with properly pitched teeth.  This flash of insight sent my design of expanding chainrings into a new direction, one that would ultimately culminate in VECTr.

This week, I was able to address a persistent constrain of expanding chain ring systems like VECTr – namely, ensuring the design will fit the pitch of a bike chain.  Standard chains have a half inch pitch, which means each roller of the chain is half an inch from its neighbor.  Thus the teeth on chainrings are half an inch from their neighbors as well, so they can fit with each roller of the chain.  The first constraint I encountered was that the system would not be able to spread the teeth out as the gear radius expanded.  My initial solution was to use gear segments (initially conceived as spiraling out from one hinged end – see my previous post).  But the space between gear segments still has to accommodate the pitch of the chain.  Each gear segment must be a whole integer number of links from the next gear segment (1, 2, 3, etc.; not 1½ , 2.33 or some fraction of a chain link) so that when the next segment engages the chain, there will be a tooth ready to fit over a roller on the chain, and subsequent teeth can fit into the pitch of the chain smoothly.

As I have been refining the VECTr design to move towards an actual prototype, I still have to design the notches in the radial arms of the base plate at the precise position where the teeth of the gear segments will match the chain pitch. That means they have to be a whole integer number of chain rollers from the next gear segment.  So, when VECTr has the size of a 50 tooth chainring, the adjacent gear segments are six chain rollers apart from each other, or 3 inches (since the pitch of a bicycle chain is ½ inch). When VECTr has the size of 45 tooth chainring, the gear segments must be five rollers apart or 2½ inches apart, and so on.  The measurements must be whole integer multiples of ½ inch (not fractional increments like 2.75 or 1.33 inches apart) if the chain is to run smoothly over the device, and not skip or slip.  (Slipping was a problem with working model, since tooth pitch was off.)   Getting the right distance between gear segments has turned out to be harder that I originally thought!

To solve this problem, I had to reach back into my memory and dust off my limited knowledge of high school trigonometry. Thanks to the beauty, symmetry and order of geometry in general, right triangles in particular, and that old mnemonic SOH CAH TOA (sine=opposite/hypotenuse, cosine=adjacent/hypotenuse, and tangent=opposite/adjacent) I was able to identify the proper locking positions of the gear segments on radial arms as 72 degree angles from each other (since there are five in the 360 degree circle).  The illustration below shows the calculations I made after some struggles to remember basic algebra and trig.

VECTr angle study

Click image to enlarge.

This exercise reminded me of what I both liked and hated about engineering: the problem solving through math, science and reason in general; the frustration in getting there.

One more problem solved…. doubtless more to come!

The idea of the continuous bicycle transmission kept nagging at me from the back of my mind. Someone should be able to solve this problem! Why not me? I loved bicycles, loved problem solving, and had built a rack and pinion steering system for my Lego car for goodness’ sake! I accepted the challenge.


In 1995, I was attending a Philosophy conference at the University of Notre Dame. While other conference attendees were debating the merits of various philosophical systems or gleefully buying Fighting Irish memorabilia, I was thinking about the bicycle challenge… and I had a breakthrough. The idea started forming in my mind of gear segments that radiated outward from a central point, which would be moved by cables. As soon as I got back to Houston, I started working on a rough design. Our tiny apartment did not have much room for a budding inventor’s work. I was pretty much relegated to an old desk  in a corner next to our dining table. Our PC at the time was ancient and slow. I still used graph paper. But, the idea was taking shape and I was obsessed with it.


The first sketch of what would eventually become VECTr.

Throughout graduate school, I kept working on the bike transmission design whenever I could… but it kept getting pushed to the back burner. I was soon writing my doctoral dissertation and starting to apply for teaching jobs across the country. I was also working part time to pay the bills, and had little time left over for the luxury of inventing. Over the years I often biked to work, and I noticed that the gear shifting was not always precise and the chain would fall off from time to time. This made me want to start again on the bicycle transmission, which I had largely abandoned in the general busyness of life.

The challenge wouldn’t let go. I had to find a way to solve it.