Invented in 1936 by Nicola Barra the cantilever brake is still almost a century later considered one of the more mysterious parts on a bicycle to adjust. This is because unlike on most brake systems with a cantilever brake you can adjust the brakes mechanical advantage dramatically altering its feel and braking power. Too little mechanical advantage results in too little power and too much results in the brake lever hitting the handlebar before the brake pads hit the rim.
The brake is only one half of the braking system, the other being the brake lever. In this text I’m going to cover my own modest research in the differences in mechanical advantage between different types of brake levers. I’m also going to discuss the limits imposed to the mechanical advantage of cantilever brakes by clearance required by mudguards, wide tyres and possible racks. These two factors together set hard limits for the maximum braking power of the braking system and I haven’t seen this combined effect studied in the same text thoroughly.
The mechanical advantage of the brake lever
The equation for the mechanical advantage of the braking system can be simplified as the mechanical advantage of the brake lever multiplied by the mechanical advantage of the brake. I’ll cover the mechanical advantage of the cantilever brake later but the mechanical advantage of the brake lever is the distance from where the lever is pulled to the pivot point (AB) divided by the distance from the pivot point to the brake cable (BC).
Campagnolo Chorus 10 speed 2nd generation Ergopower lever pictured.
Since the difference in the AB distance between different brake levers meant for drop handlebars is little and we’d prefer to have our fingers fall on roughly the same place regardless of the brake lever, different brake levers can be grouped to different mechanical advantages by just the BC distance. This is also the quickest way to check if you’re unsure if you have a v brake or cantilever brake lever on hand.
These measurements were taken by me and might not be 100% correct but should be accurate enough for our purposes. This covers brake levers from roughly the 70s (the Vintage Mafac) to today (Tiagra 4700 is still in production) and we can see that there are clear differences. As we shall later see, just 2mm of difference is important when it comes to mechanical advantage. Here less distance means more mechanical advantage.
The common wisdom is that aero levers have more mechanical advantage than non-aero ones since you’re suddenly also braking from the hoods and not just the drops which reduces the AB distance discussed above and mechanical advantage on that side of the equation. That seems to be the case: the non-aero Shimano 600 has a BC distance of 24mm and the aero offerings have a BC distance of just 19mm. An interesting case is the vintage Mafac lever with a BC distance of just 22mm. I think the explanation is that the lever was actually designed to be used with cantilever brakes (and centerpulls) unlike any of the others.
Despite this common wisdom we see the distance increase and the mechanical advantage decrease in the Shimano levers. Shimano has named its different braking systems starting chronologically with SLR (Shimano linear response), then Super SLR, New Super SLR and finally SLR EV. Sadly, I didn’t have a SLR lever to measure but I think it would measure 19mm. These correspond with changes in design with caliper road brakes which have increased in mechanical advantage. The same can be seen with Campagnolo going from the 2nd generation of Ergopower to the 3rd. In my understanding SRAM currently also uses the 23mm pivot distance.
It’s unclear whether the Tiagra lever is New Super SLR or SLR EV. They’re supposed to have about the same mechanical advantage so for our purposes there’s no distinction. The lever is also very tricky to measure as the point of contact with the lever and the cable is hidden inside the lever body but I’ve come to the 25mm figure by estimating, calculating and comparing measurements online. Measured to where the cable head sits the distance would exceed that of the sole V-brake lever which would misrepresent how the lever actually works.
The point here is that with modern brifters you might have less mechanical advantage than old non-aero brake levers and with Shimano brifters you definitely will have.
The mechanical advantage of the cantilever brake
Benno Belhumeur’s article Cantilever Brake Geometry: Setup and Mechanical Advantage written in 2010 strips much of the mystery in the adjustment of cantilever brakes. You can find the article here along with an interactive tool. The mechanical advantage is mostly dependent on the yoke height. The lower the yoke, the higher the mechanical advantage. The problem with cantilever brakes is usually the lack of power and lowering the yoke is the simplest way to increase it. There’s more to Belhumeur’s article that I’m not going to cover here as I’m focusing on the brake levers.
The figure above from Belhumeur (2010) plots the mechanical advantage of the brake as a function of yoke height. I’m going to refer to the brakes Paul Neo Retro and Paul Touring by their type in my text. The Neo Retro is a wide profile cantilever brake closely resembling old Mafac models. The Touring is a low profile cantilever brake, a design which became popular in the 1990s on mountain bikes. Both types are commonly available today.
The unnamed medium profile brake would be something like the Shimano Deore XT BR-M730 from the late 1980s and this type is not produced anymore. That is a shame because in my view it’s the best design of the three. The ultra wide profile brake would be close to the original Barra design from the late 1930s.
As we can see of the two main types of cantilever brakes, low and wide profile, the low profile one always gives more mechanical advantage. As Belhumeur explains the down side is that it’s trickier to adjust than the wide profile brake. The medium profile brake would be as easy to adjust as the low profile one but it provides a lot more mechanical advantage at the same yoke height, which can be a limiting factor.
Interestingly Shimano moves to the higher mechanical advantage low profile brake at the same time when it moves to the lower mechanical advantage Super SRL brake lever. It was introduced to Dura-Ace in 1990 and to the mountain bike brake levers soon after that. The other reason for this is the increasing tyre size in mountain bikes at the time. With the low profile brake the yoke will be a lot higher than with the wide profile brake giving more tyre clearance. Increased distance between the tyre and the straddle cable could also be seen as a safety feature.
Picture of the front end of Pack Rat as it currently is set up.
This segues us to the problem low profile cantilever brakes have with the yoke height. I’ve a Surly Pack Rat which I’ve written about before on the blog in Finnish (link). It has 47mm tyres, mudguards on top of those with about 20mm of clearance and a Pelago Commuter front rack that attaches to the fork crown. I’ve modified the L bracket of the rack but the lowest height the yoke can go is still about 100mm. With low profile brakes this would give a mechanical advantage of about 1,9 according to Belhumeur’s graph. (This is ignoring the differences in design between different models of the same type).
The Surly presents a pretty typical scenario in modern bikes with cantilever brakes. 47mm 650B tyres with mudguards and a front rack is a common setup on bikes like it. Losing the rack would only lower the yoke about 1cm. With a 90s mountain bike with 26” wheels and even wider tyres you’d likely have even higher minimum yoke height. The wheel size doesn’t affect the yoke height but smaller wheels usually mean wider tyres.
I built my bike with the Tiagra 4700 levers and wide profile cantilever brakes (Tektro CR-720) and the braking was ok when you were familiar with them but I kept wanting more power. With a lot of work they could even be very powerful but that didn't last long with the brakes quickly going out of adjustment. After switching to low profile brakes (low end old Shimano Acera X) I’ve been much happier with the brakes. Next I’ll look into why that is.
Mechanical advantage of the brake system
Now we’ll look at the mechanical advantage of the whole system. The equation is simple: the MA of the brake lever is multiplied by the MA of the brake. On the following graphs I’ve calculated the MA of the brake as 1,9 for the wide profile cantilever and 2,5 for the low profile one. These are derived from Belhumeur’s graph with the brake types and yoke heights I’ve used on my Surly. I've cut a few corners and haven't calculated the exact mechanical advantage with my fork and brakes but the results are close enough.
For the first graph let’s look at the mechanical advantage of the system when braking from the drops. I’ve estimated the distance from the fingers to the pivot to be 73,5mm. The brake levers I measured curve outwards at the bottom at about 9 to 10cm from the pivot so this would be about where the lever would be pulled from when braking in the drops with two fingers on the lever.
Braking from the drops.
As we can see there are big differences between the different levers and the different brakes. In my experience though the Tiagra lever with the wide profile brake was still pretty good when braking in the drops. It’s possible that mechanical advantage above 8 might be too much and that’s something I haven’t (yet) tested. The V-brake Tektro lever is there as a control and should not be used with any other type of brake.
Modern drop bars are generally meant to be ridden from the hoods. When braking from the hoods you have a lot less leverage on the brake lever with the distance from your fingers to the pivot being a lot shorter. I’ve estimated the distance to be 4cm in the following table. In any case or with any lever braking from the hoods will not be as powerful as braking from the drops and the bike should be set up so that both methods are comfortable to perform.
Braking from the hoods.
The mechanical advantage is clearly reduced and this is usually where the problems in combining modern brifters with cantilever brakes will show. In my experience the Tiagra levers have been at best acceptable with the wide profile brakes braking from the hoods and good with the low profile ones. I wouldn’t recommend them with the wide profile brakes. By this calculation the older Campagnolo or the TRP levers should be as good from the hoods even with the wide profile brakes.
Shimano still makes aero brake levers without shifters and they still use the same Super SRL pivot distance. The design seems to have been unchanged in every way for decades now. Seeing how common these are I’d hazard a guess they probably would be ok from the hoods but have not tested them myself.
Conclusion
As we have seen the cantilever braking system has several factors to consider to get a well braking total. In my experience I recommend against using Shimano brifters with wide profile brakes and wide tyres. I’d also be wary about using modern Campagnolo or Sram brifters too. With low profile brakes or narrower tyres all of them should work.
This is a shame since the Shimano brifters are by far the most common and the low profile cantilever brake has a bad reputation as being difficult to set up. The reputation is mostly to blame on the Shimano Link Wire system which makes it impossible to adjust the yoke height which we have seen to be the most important adjustment on the brakes. Though as can be seen on Belhumeur’s graph getting the yoke height just right is more precise too than with other brake types.
What about the cantilever brake BR-CX50 that Shimano is still making and that is compatible with the New Super SRL brake levers? It probably is a very good brake but it has limited tyre clearance having been designed for 35mm tyres without mudguards. The straddle cable can be changed to allow a higher yoke and the brake has pretty much a low profile design. It also has a very Shimano-like adjustment process which is made irrelevant by changing the straddle cable in my eyes making it pointless to get them. If you already have a set and need more clearance, changing the straddle cable makes sense.
Another brake type that suffers from lesser mechanical advantage in modern brifters is the long reach caliper brake. As the brake pads move further away from the brake’s pivot point the mechanical advantage of the brake is reduced. Unlike cantilever brakes caliper brakes have no other way than moving the brake pads to alter their mechanical advantage.
As I wrote in the beginning you can also have too much mechanical advantage. Having more mechanical advantage in the brake lever, where it can’t be adjusted, frees a bigger potential range of adjustment in the brake where it can be easily altered. It also makes adjusting the brake easier as you’re not as likely to be struggling to find more mechanical advantage to make the brakes more powerful.
Petri Vaittinen