A Closer Look at Haven Mercer's Mechanical Automatic Transmission Concept

The bicycle drivetrain has seen a multitude of performance-enhancing changes over the last couple of decades, culminating in very reliable, easy to use options available over an astonishingly vast range of price points. We've seen the widespread adoption of narrow-wide chain rings that improve chain retention, clutch derailleurs that do the same (as well help keep things quiet on the trail), and the consolidation of the 1x system with wide range cassettes and climbing gears that, as my dad tells me, "you could ride up the side of a house with."

Then, there was the advent of wireless, electronic shifting, and the more recently introduced auto-shifting and coast-shifting features for eMTBs. More recently still, we've seen gearbox technology, which has been around for years, combined with an eBike motor in a single, sealed unit that does it all.

While all of these things are impressive, they are, arguably, somewhat incremental. It's rare that we see something completely left of field when it comes to the drivetrain. Hence, my interest having piqued in the last few days with the publication of Haven Mercer's patent on an "Automatic Transmission System for a Bicycle".

What we have here is a continuously variable transmission, meaning that the gear ratio varies between an upper and lower limit with no set number of gears to speak of; the number is effectively infinite. And, not only is the system fully automatic, it is also entirely mechanical. We got the details from the inventor himself.

---

I was delighted to hear that Mercer's continuously variable transmission is no mere concept. Indeed, he has created a working, rideable prototype.


And, he's taken it mountain biking. Indeed, his goal is to develop the system such that it works well enough for trail riding, reasoning that if it can cope with the demands of mountain biking, it can probably cope easily with the demands of road cycling, and so on. Fair enough.

How does it work?
The system is centered upon the ability of the chainring (for want of a better word) and rear sprocket to automatically expand and contract in response to varying torque inputs. Both consist of a number of pulley wheels distributed about the center of rotation, each supported on a linkage, the position of which is determined by a spring and the amount of torque applied in any given moment.

At the chainring, the mechanism is configured to bias the the pulley wheels to an extended position, such that they are each at their furthest point away from the center of rotation when there is no load on the pedals. Here, the chainring is in its largest effective size.

Before I continue with this explanation, it is essential to note that each pulley wheel sits on a one-way bearing. On the chainring, the pulleys can rotate clockwise, but not anti-clockwise.


So, when torque is applied to the pedals and the chain is pulled through the system, the last pulley to engage locks into the chain, effectively becoming a torque-sensing element. The more torque applied, the more the spring is forced to compress and, concomitantly, the smaller the effective chainring size becomes.

Indeed, as the torque-sensing pulley compresses, all of the others do so in unison. The fact that the pulley wheels are able to rotate in a clockwise direction, allows them to feed the chain through such that, as the distance between neighboring pulleys diminishes, the chain doesn't simply get ejected. The same is true for when torque is reduced, and the effective chainring size gets larger.


The mechanism is reversed at the rear sprocket. Here, the spring and linkage system is configured to bias the sprocket towards its smallest possible size, with the pulley wheels closest to one another.

So, in an unloaded state, the transmission sits in its hardest gear. Stamping on the pedals forces the rear sprocket's effective size to increase, and the chainring's effective size to decrease, moving the system towards a relatively easier gear. As momentum increases, and the rider's torque input backs off, the rear sprocket will shrink, and the chainring size grow. The transmission automatically adjusts the gearing to the demand of the rider.


Doesn't it have a built-in chain tensioner?
It occurred to me that it could be, theoretically, possible to run this system without the use of any chain tensioner. It's not so difficult to see that as the chain length requirement at the chainring increases, the chain length requirement at the rear sprocket decreases. While that is the case, the two are not equivalent.

Haven Mercer, inventor of this continuously variable transmission, explained that the system does in fact need not one, but two chain tensioners. That's because the rate at which the two ends expand and contract is different, and that can introduce undesirable chain slack along the upper and lower chain lines. The upper tensioner also doubles up as a sort of chain guide, which appears to be rather necessary given how violently the chain jumps up and down as it is fed into the chainring.


Who is this alternative bicycle transmission for?
As I mentioned, inventor Haven Mercer intends to develop this transmission to the point where it's good enough for mountain biking, technical climbing and all. There are issues with the current prototype, the main one being that the springs are too light, and are thus too easy to compress under pedaling loads. Because of that, he says it rides a "bit squishy" at the moment.

It's easy enough to imagine that the perfect spring rate may not exist, or at least it won't be the same for every rider or every kind of terrain. That said, Haven plans to engineer it such that the springs are easily swapped out by the end user.

Another concern is the durability of the one-way bearings, not to mention the numerous durability issues that could arise from having so many moving parts.

Without riding it, it's hard to gauge how well it would work under any circumstance. However, the fact remains that this invention does eliminate the need for shifting, allowing the rider to pedal along without having to think about what gear they're in, or what gear they want to be in. The transmission would simply adapt to their input.

That could genuinely present an improvement in safety, particularly in the context of riding through heavy traffic. And, particularly for the individual who is a very, very occasional cyclist who is not tremendously well-versed in bike handling. The less that person has to think about, the more attention they can give to the road and other road users around them.

Posted In:
Reviews and Tech First Looks