Joining VeloClub not only supports the work we do, there are some fantastic benefits:
by Alan Coté
June 17, 2020
Is bicycle shifting about to get some entirely fresh technology? A recent patent application shows that SRAM has been hard at work designing a wireless front derailleur system unlike anything seen before – with three separate gear-changing mechanisms that rotate with the crankset. It’s made possible by motorized derailleurs and wireless communication with the shift lever, something that SRAM knows well.
I’ll give an overview of how this complex system works – it’s helpful to use the patent drawings as a guide. Patent drawings must be black and white, but I’ve added color for easy reference. Want a deeper dive? Check the US patent application which includes 78 drawings and 32,000 words of description covering 89 pages. Even for a patent that’s quite lengthy, underscoring how much there is to explain with this invention.
All of the design’s front shifting mechanisms are built into the crankset. There’s nothing – zero – mounted to the frame. The system includes a wireless receiver, battery, and one motor, all of which enable three separate derailleurs. It’s all tied together with linkages and cams and pivots and springs and sensors in a symphony of electro-mechanical action. I’ll skip over most of the internal fiddly bits, and focus on how shifting actually works.
To shift from big ring to small ring, there are two downshifting elements (212a, 212b), highlighted in green in the patent drawings. They are positioned outboard of the big ring, and each can swing back and forth on a pivot. Either one of those can knock the chain down off the big ring to the small – more on that in a minute. Press a shifter button like those used on present SRAM wireless systems, and the crank-mounted receiver activates the motor, and sets the downshifting elements in motion.
But those downshifting elements can’t upshift the chain from small ring to big. So there’s an entirely separate system for upshifts, which uses a set of protruding shift elements, or pegs (288, in orange). So call it a triple derailleur delight.
The big ring includes five holes, through which the five pegs can move inward when the rider wants to upshift. The pegs lift the chain upward to the big ring as the cranks rotate, with the pegs in a staggered orientation so each peg lifts the chain a bit higher than the previous one. The pegs retract out of the way when it’s time to downshift. The rearmost peg (605 in Fig 72 below) is special – it’s more of a hook, and actually takes the place of one tooth on the big ring. More on that later too.
If you’re starting to think of reasons why this design can’t work, you’re on the right track. SRAM starts the patent application touting problems with a conventional front derailleur: the frame is constrained to accommodate a front derailleur mount, and positioning and adjusting it is fiddly, with chain derailment too often a consequence. Just ask Andy Schleck.
But the problem SRAM creates by having the derailleurs rotate with the crank is that a chain can only be moved from ring to ring when a front derailleur is positioned at about 11 o’clock (with the chainring as a clockface). That’s because where a chain is engaged with the teeth of the chainrings, it won’t derail from the ring if pushed sideways. Conventional front derailleurs are located just before the spot where the chain engages with the chainring teeth.
SRAM engineers found a way around this, but it adds significant complexity to the system. Sensors know the rotational position of the cranks, so the downshifting elements only move into shift position when they’re between about 6:30 and 11 o’clock. When one downshift element approaches 11 o’clock, it encounters the side of the chain, and pushes the chain inward and down to the small ring. That’s why there are two downshifting elements: press a shift button, and the system waits until one of them is in the right spot. That means two shift periods per pedal stroke.
It’s the same with upshifting. Those upshift pegs only jump out from the inside of the big ring when they’re in that 6:30 to 11:30 zone. Then there’s the special upshift peg, the fifth and final one. It has a hook shape that helps to lift the chain high enough to hop on to the teeth of the big ring. The hooked peg is then positioned inline with the other teeth on the big ring, and fills in a “missing tooth” gap.
This means that an upshift is initiated only once per pedal stroke, as opposed to twice per pedal stroke with the dual downshift elements. The patent application does say that multiple upshift systems (e.g. a second set of five pegs) could be used – but it’s not pictured, and the whole assembly already looks stuffed, so that option seems more a matter of not limiting the patent disclosure.
The rider doesn’t need to be concerned with shifting at the wrong moment, as sensors delay everything until all is in the right spot. Sensors also detect if a rider is turning the cranks backwards, and prevents shifting under that condition.
When we might see samples in the field, let alone something for sale? That’s anyone’s guess. SRAM’s Michael Zellmann offered “no comment” about the patent filing and technology when asked, as is common practice. It is perhaps a surprising development from SRAM, who have become the mavens of single-ring 1x drivetrains in recent years.
Are there drawbacks to this new system? The design is certainly much more complicated – and therefore expensive to manufacture – than a conventional front derailleur, even a wireless electronic one. And the once-per-rev shifting isn’t ideal.
But the design could make for foolproof installation, and perhaps it achieves a new level of shifting smoothness and reliability. Also SRAM has already shovelled some money at it: a patent application of this length cost tens of thousands of dollars to prepare and file. Dwarfing that is the cost to develop the design – even if prototypes haven’t been built, a substantial amount of R&D time went into just what’s shown in the patent application.
Then there’s the cautionary tale of products like the Browning Automatic Transmission – a vaguely similar concept in which a pie-shaped section of the front chainring moves to shift the chain between rings. It was built in a number of different formats but never took off. Such can be the fate of new technology that’s either ahead of its time, or overly complicated for the resulting benefits.
Is this an engineering exercise from SRAM, or is it destined to become a product? While we don’t know, we will certainly be keeping our eyes on SRAM cranksets at pro races and elsewhere.
Registered Patent Agent Alan Coté is a mechanical engineer and principal of Green Mountain Innovations LLC. His writing credits include Outside, Men’s Journal, Bicycling and many others. He also serves an an expert witness in bicycle related legal cases.