The Browning Automatic Bicycle Transmission

by Chester R. Kyle, Ph.D.
Reproduced from: Cycling Science Winter '95


Bicycle transmissions have continuously evolved for over a century and although they are deceptively simple, they are extremely difficult to improve [1]. One might suppose that we are nearing the end of the development process. Not so. The last decade has seen the rapid acceptance of index shifting, the grip shift, 8 speed cassettes, the Mavic electric rear derailleur and other innovations. Amazingly, there is still one important avenue that has been left almost entirely unexplored - the automatic bicycle transmission. Browning Research, near Seattle, Washington, is probably the first company to successfully solve the formidable problems associated with fully automatic shifting.

The basic mechanical system behind the Browning transmission was invented in 1974 by Bruce W. Browning and developed by Bruce and his sons Marc, David, Paul and Chris. This unique shifter uses a hinged sprocket sector which can swing either out or in to guide the chain to the next sprocket. See Figure 1 and Figure 2.
Figure 1

Figure 1. The elements of the Browning three-speed transmission are shown on the left. On the right is a Browning four speed rear cluster.
Figure 2
Figure 2. The left hand view shows a front Browning three-speed chain ring shifting down from the large sprocket to the middle sprocket. The right view shows the chain shifting up from the middle sprocket to the large sprocket. Note in both views the chain is in driving engagement with both sprockets at the same time.

The critical feature that differentiates the Browning from standard derailleur systems is the continuous gear/chain engagement during shifting, there is no skipping or grinding even under the heaviest loads. The transmission wil1 shift under any combination of speed, cadence and pedaling force. Shifts are smooth, fast and nearly silent. The chain stays fully in contact with the gear teeth during either up or down shifts. In contrast, even with the best standard derailleurs, on upshifts, the chain loses contact with the smaller sprocket before it is fully engaged on the large sprocket which often leads to grinding or skipping under heavy loads.

Another unique feature of the Browning transmission is that it is ideady suited for complete electric operation including computer controlled automatic shifting. This eliminates derailleur cables and cable adjustment. Although the Browning automatic transmission is not yet in mass production, it has been extensively tested in the field and the laboratory and it is now a fully functional design. Versions of the transmission have been produced in the past and orders can presently be placed for twospeed, three-speed, four-speed and twelve-speed transmissions.Another unique feature of the Browning transmission is that it is ideady suited for complete electric operation including computer controDed automatic shifting. This eliminates deraiDeur cables and cable adjustment. Although the Browning automatic transmission is not yet in mass production, it has been extensively tested in the field and the laboratory and it is now a fuDy functional design. Versions of the transmission have been produced in the past and orders can presently be placed for twospeed, three-speed, four-speed and twelve-speed transmissions.


Bruce's Grandfather was the brilliant gun designer, John M. Browning, inventor of the Winchester 30/30, the pump shotgun, the Colt 45 automatic, the U.S. Army WW I and WW II machine guns, the Browning automatic shotgun and more than one hundred other guns, many of which are still manufactured today. John's father was also a gun maker. A hallmark of Browning's gun designs is their rugged reliability, their simplicity and flawless performance. They work without failure under the worst conditions. Bruce cut his teeth designing elegant gun mechanisms for the Browning Arms Company. He left Browning in 1969 to work on his own inventions and designs.

In the early 1970's, Bruce and his sons formed a family research company to invent, research, prototype and sell ideas. Among these ideas was the concept of an automatic/electric bicycle transmission. For a short time in the early 1970's, Browning Arms marketed a bicycle as part of their sporting goods business. Under an agreement with Browning Arms, Bruce and his sons worked on an automatic bicycle transmission as part of this project. In 1977, Browning Arms closed their bicycle business, so Bruce and his sons continued to develop the transmission on their own. Since then it has undergone continuous change and improvement at the Browning Research. The company buildings are nestled in the woods, a half-hour ferry boat ride from Seattle, but about a century away in their idyllic surroundings.

During the 1980's, a two speed mechanical version of the Browning transmission was adapted for BMX racing. The essential elements of the present device were developed for this BMX transmission. In the late 1980's, a three speed ATB chainwheel was introduced which was later licensed exclusively to Suntour. The transmission was produced under the name "The Beast" Unfortunately, like many components companies which were under intense competitive pressure from Shimano, Suntour fell on hard times and the production of the ATB transmission was halted.

Research on a fully automatic, computer operated version of the Browning transmission continued. By 1990, two 12 speed prototypes were finished, one that was fully automatic, and one that used four electric push buttons (up and down for both front and back gears).

In May 1993, the Browning "Smart Bike" was completed. The development, which is still continuing, has been a cooperative effort among a team of 17 persons, including project management and mechanical design by Marc Browning, software development and mechanical design by David Browning, testing and product reliability by Paul Browning, general administration and legal affairs by Gloria Browning and mechanical design and guidance by Bruce Browning (see the photo). This makes the fifth generation of Brownings who have been involved in creating basic inventions. Their two decades of effort have again produced a truy superior system.


The Browning automatic transmission consists of five basic elements:, the onboard computer, the chain length compensator, the front and rear sprocket clusters, the gear selectors, and two communications buttons on the handlebar. The whole system is powered by a common 9 volt portable radio battery which will last between 200 and 300 miles, depending upon the average speed of the bicycle. After redesign for manufacture the power requirements will be lower and the battery life will be extended to over 1000 miles.

The waterproof computer container shown with its cap removed and the battery being inserted. The computer shown is an early version. The size will be much diminished in later models.
The computer (see the photo) utilizes a small eight bit microprocessor to electronically control the shifts. It is housed in a small aluminum canister, completely sealled against the elements with O-rings. The whole bike can be dipped in a river or hosed off for cleaning without damaging the electronics. Wires connect the computer with the shift mechanism and the controls. There are no cables to maintain or install.

When starting from rest, one crank revolution will wake up the computer from its power-conserving sleep mode. Two front sensors use magnets and reed switches to feed pulses to the computer to determine the direction and rate of crank rotation. Two rear sensors are used to determine the wheel speed and the sprocket speed. By comparing the number of sensor pulses of the front crank with the freewheel sensor pulses, the computer can determine the current gear. After each shift, the computer updates the current gear. Using the wheel speed, the computer constantly searches for a "desired gear" that will allow the rider to maintain a preferred cadence. If the current gear is different from the desired gear, when the rider is pedaling forward, the computer will automatically shift toward the desired gear with no attention from the rider.

The rider can adjust the preferred cadence by pressing either the up or down button to shift to another gear. The computer remembers this selection and will choose gears, within the gears available, to maintain this new cadence.

I rode the Smart Bike for about 40 miles through a hilly region of Northern California. With steady riding, it shifted between 8 and 18 times per mile depending upon the terrain. This was much more than I usually shift, but the shifts were generally welcome. After the novelty had worn off, the shifting became essentially invisible, I forgot about it during the last part of the ride. At the beginning, when going up a long grade, I overrode the computer a few times to select a more comfortable gear, but later I seldom needed to do this.

David Browning, who developed the computer software, has solved some very difficult but fascinating problems. Today's common 24 speed road or mountain bikes, really only have 13 or 14 effective gears because of the close overlap between certain ratios. In contrast, the Browning Cross Bike that I rode had 12 gears, but they were in a neat progression, so all of them can be programmed into the computer software. The bike I rode had some of the shifts blocked out, leaving some fairly steep changes, but this can be modified in later versions of the software. The computer program allows the cadence to vary within a band width, so that the shifting will not "dither" (shift back and forth) when close to a shift point.

Figure 3. A side view of the Chainlength Compensator. Except for the length of the S shaped arms, this device is generic to all varieties of the Browning rear transmission. It does not derail the chain. It passively follows it.
Unlike a derailleur, the chainlength compensator requires no force or power, it is completely passive. See Figure 3. It acts as a chain guide, it maintains chain tension, and it keeps about 180 degrees of chain wrap around the sprocket at all times.

The three front chainwheels on the bike I rode were 48, 38 and 30, and the four speed rear cluster was 12, 17, 23, and 32 which gave a gear range from 26 to 110 inches in 12 steps, see Table I. Swinging sectors are included on all but the smallest gears. On the 12 tooth sprocket and the 30 tooth front chainwheel, no swinging sectors are necessary since some of the teeth are shortened, and the adjacent sprocket sector is able to swing over into the chain line and pick up the chain when shifting up.

When a sector moves either out or in, it positively guides the chain to the next sprocket like a ramp on a freeway, or a railroad switch guiding a train from one rail to another. The sector provides a helical path for the chain to follow, and maintains positive contact during the switch. The shifts are therefore smooth and flawless. Shifts are timed so the swinging sectors can be moved without chain interference and they require very little force. At present a limited number of gear sizes have been made for the prototypes, but this selection can be expanded upon manufacture.

To assure the reliability of the shifting system, the Brownings have built testing machines that have subjected the transmission system components to thousands of hours of service under various loads while undergoing millions of cycles. Any failures have been analyzed with high speed videos and corrected, until the system is now superbly reliable.

Figure 4. The front and rear gear selectors in neutral, up-shift, and down shift positions
Front and Rear Cams in Neutral Position
Cams in Switch Up Position, From Smaller to a Larger Sprocket
Cams in Switch Down Position, From Larger to a Smaller Sprocket

To change gears, a pawl is displaced by a selector cam, and all of the sectors are hinged either outward or inward depending upon which cam groove is followed. See Figure 4. Once the chain is derailed, the pawl is released and a small spring returns the sectors to the neutral position. The rider provides all of the force necessary to displace the cam and complete the shift, so the gear selector requires very little power. Without the computer power drain, one 9 volt battery will complete from 80,000 to over 100,000 shifts. A small reversible electric motor is contained within the selector. A 50 millisecond pulse accelerates the motor and an inertial hammer strikes a trip mechanism. A latch and key system then locks the cam in either of its two switching positions. Once the shift has started, the trip mechanism, then releases the latch and holds the cam in its neutral position.

Figure 5. A conceptual drawing of the Browning handlebar mounted control assembly which will house the computer, a speedometer and an odometer with ergonomically placed control buttons.

A variety of button and control assemblies are possible with the Browning transmission. Shown in Figure 5 is a two-button control for a twelve-speed automatic. The housing includes the computer which controls the shifting and a read-out for speed and distance.

A small control box is placed on the handle bar for easy finger tip operation. A toggle switch allows the rider to select either manual operation or automatic. In manual, two buttons allow the selection of either a higher or lower gear. The transmission will remain in this gear until the rider commands another shift. In automatic, pressing the buttons will cause a shift, but it will also cause the computer to readjust the preferred cadence for future automatic shifting.

A four button manual version of the Browning electric shifting transmission has been developed for mountain bikes. Two finger tip control buttons for each of the front and back shifters are placed in two small control boxes on the handlebars. All of the other transmission elements are the same as on the Smart Bike.

The Browning transmission could be applied to bicycle racing in general. In the racing version, if the computer were housed under the seat, the aerodynamic profile of the bike would actually be less than a standard racing bike because of the elimination of cables and shifters. Also in a racing version, the weight of the Browning bike would be about the same as a high end conventional racing bike because of the elimination of shifters, mounts, cables, and other elements of a standard gruppo. Probably the manual/electric shift version of the Browning transmission would be preferred for road or mountain bike racing or in road time trials because of the unpredictable nature of conditions. Here, the finger tip shifting of the Browning and the non-slip performance would be very useful features.

There are other racing events where the automatic Browning transmission could prove invaluable. The UCI regulations permit gear shifts on the track for individual time trial races under article 49, Section III: "For track competitions, other than where the rider takes the track alone, (500m, kilometer TT Individual Pursuit and individual record attempts), derailleur gears and brakes shall be prohibited."

In track racing, in the individual time trial events such as the 1000 meter or the 4000 meter individual pursuit, cyclists must accelerate from a standing start and maintain speeds of from 33 to 38 mph for several laps of the track. Track racers currently use a fixed gear of about a 93 inches for the entire race. In these events, gear shifts might be advantageous because they could improve acceleration from a standing start.

Gear changers have been used in the past in international track racing. The kilo champion of Canada, Jocelyn Lovell used a two speed rear hub in the World Championships in the early 1980's. Lovell's system of drive chains on both sides of the rear axle was used by Gene Samuel of Trinidad in the 1992 Olympics to place 8th. Also in the 1992 Olympics, Adler Capelli used a derailleur gear shift in the 1000 meter time trial, and placed 5th. The Browning automatic transmission is technologically advanced compared to these earlier systems and can produce the same result in a cleaner, more efficient manner shifting by using a computer model along with available research data. In 1976, I used a 200 pound unbraked flywheel to measure the rate of acceleration of a sprinter from a standing start [2]. The flywheel was driven from a stationary racing bicycle fixed to a stand. The flywheel speed was recorded using a light emitting diode and a fast strip chart recorder. The energy input to the flywheel could then be determined by calculating the kinetic energy of the flywheel. From this, I found the average crank torque as a function of time and other variables. Figure 6 shows the result. The rider, Mario Palombo, was a good sprinter but not nationally ranked.

Figure 6. An unbraked 200 pound flywheel dynamometer was accelerated from a stop by a cycle sprinter pedaling a stationary racing bicycle. The curve shows the crank torque, rider power, and crank RPM versus time. The power reaches a peak between about 120 and 140 RPM and declines as the crank speed becomes very high.
The crank torque in Figure 6 can be expressed accurately by a simple equation:

( 1 ) T = 270 - 35.23(RPM)0 3576

where T iS the torque in ft-lb, and RPM is the crank speed in revolutions per minute. Since world class track riders are more powerful than our test subject, the coefficients in the above equation were revised upward by about 7% to more accurately match current track times. This gives:

(2) T = 289 - 37.69(RPM)0 3576

Knowing the speed of the bike, and the gear ratio, the crank RPM can be calculated. From this, the wheel thrust T in pounds due to the crank torque can be expressed as a function of the bike speed V in ft/sec, the wheel radius rw in ft, the number of teeth on the chainwheel Nchw, and the number of teeth on the rear sprocket Ns as follows:

(3) T = (NsNchw)(1/w(289 -37.69 [9.55V Ns / (rwNchw)]0.3576)

Newton's equation in differential form will give the time for a cyclist to reach a certain maximum velocity:

(4) dt = MdV/[T - (W(A0 - A1V) + A2V2)]

Where t is the time in seconds, M is the total mass of the rider and bicycle in slugs, W is the total weight in pounds, A0 is the static rolling resistance coefficient, A1 is a coefficient which expresses the variation of rolling resistance with speed, and A2 is a coefficient expressing the wind resistance of the rider and bicycle. Let us analyze a track race on a 250 meter velodrome and calculate the acceleration time to 125 meters.

Typically a kilo racer who can do 1000 meters in 1 minute 7 seconds can turn the first half lap in about 12 seconds, and his top speed is about 37 mph. From wind resistance data, and rolling resistance data from previous tests, a realistic estimate of the above coefficients for track racing with modern equipment would be A0 = .0022, A1 = .000029, and A2 = .00255.

Another differential equation gives the distance covered in a certain time:

(5) dx = Vdt

I solved Equations 3,4 and 5 for various gear combinations using numerical integration. The results compared favorably with actual split times taken during velodrome races.

Suppose we try a 14/19/24 tooth rear cluster with a single 48 tooth front chain ring and a 700C rear wheel. This would give gears of 91, 67 and 53 inches. With no shifts, using the 48/14 gear, the computed half lap time to 125 meters would be 12.08 seconds. If we try shifting when the crank RPM exceeds 100, using only one shift from 19 to 14, the computed time would be 11.53 seconds. Using one shift from 24 to 14, the time would be 11.47 seconds. Using two shifts from 24 to 19 to 14, the time would be 11.31 seconds.

So, the computation shows that the time for the first half lap might be improved by as much as 0.8 seconds by using an automatic shifter. Since the chain loads on the rear sprockets are enormous during the explosive starts in track racing, the Browning transmission would be ideal for this service. Even though the first half lap split time appears to favor shifting, it is still not certain whether the total race time would be lower or higher. However, the technique looks extremely promising and it merits extensive field testing.


The Browning Automatic Bicycle Transmission is probably the first successful computer controlled shifting system, and it opens an entirely new avenue for the bicycle industry. The successful introduction of a fully automatic electric bicycle transmission could lead to the rapid expansion of a new market. The transmission has application to all types of cycling and might prove especially valuable in racing.

  1. Kyle, C.R. Alternative Bicycle Transmissions. Cycling Science, v3, nos. 3 & 4, (1991), pp 33-38.
  2. Kyle, C.R. & J. Mastropaolo. Predicting racing bicyclist performance using the unbraked flywheel method of bicycle ergometry. Presented to the International Congress of Physical Activity Sciences, Quebec City, Canada, July 11-16,1976. Biomechanics of Sport and Kinanthropometry. 6:211-220, (1976). F. Landry and W. Orban (Eds.), Miami Symposia Specialists.

Chester R. Kyle, Ph.D. is Research Director of Sera Sports, Sports Equipment Research Associates.