Why I Choose Centerpull Brakes

When I spec’d my new bike for this year’s Paris-Brest-Paris (and for our adventures in the Cascade Mountains), I opted for centerpull brakes. I didn’t choose centerpulls out of nostalgia. For the riding I do, they are the best choice.

Why not dics brakes?

It’s undeniable that the best hydraulic disc brakes offer amazing braking power. Isn’t more braking power always better? There is a limit to how much braking power you can use. Once your rear wheel lifts even with your weight all the way back, you can’t use any extra braking power. A good centerpull brake has just enough power to lift the rear wheel.

If the ultimate braking power is similar, then the choice of brakes comes down to other factors. For me, it’s about the fork rather than the brake itself: Disc brakes feed the braking forces into the left fork blade and flex it backward. If the fork isn’t stiff enough, the bike will self-steer when you brake hard. I once rode a test bike with a Wound Up disc fork that required a quick flick of the handlebars every time I braked hard, to compensate for the fork twist. It became second nature, but many riders might not like this.

Modern disc forks are much stiffer and no longer self-steer. The down side is that this stiffness transmits road shocks that are too big for the tires to absorb. Well-designed steel forks with thin blades flex up to 15 mm (0.6″), just enough to take the edge off these bumps. That’s not just more comfortable, but also faster.

There other reasons why I don’t like disc brakes, but they are relatively minor. One is weight – rotors, calipers and long brake lines all add up (although that can be mitigated somewhat if you use carbon rims.) Discs tend to bite quicker in the rain, but most discs I’ve ridden howled terribly when wet. Discs require more maintenance and care – the hydraulic houses are prone to kinking, and the pads often rub noisily on the rotors, requiring frequent adjustments of the calipers to recenter them. (Thru-axles have helped with that.)

Disc brakes have their place: They are an excellent choice for bikes with very wide tires. They don’t have to reach around the tire, so the brake’s weight and power are independent of tire width. That is why mountain bikes use them. With front suspension, the stiffness of the fork blades is a non-issue. On modern production bikes with wide tires, disc brakes make sense: They are what is available, and they work well. Simply choose the widest tires you can run, and you’ll get plenty of shock absorption.

For custom bikes with moderately wide tires, I think the main reason riders are tempted by discs is simply this: Most rim brakes for wide tires weren’t very good. But those problems can be overcome.

Why not cantilevers?

Our Rene Herse cantilevers are among the lightest brakes in the world. At 75 g per wheel, they weigh far less than most racing brakes. We used them on the J. P. Weigle for the Concours de Machines Technical Trials in France. They brake very well, too – as I could confirm when descending from the mountains in pouring rain during the Concours.

We’re very proud of our Rene Herse cantis, but I still prefer our centerpulls. The inherent drawback of all cantis is the location of the pivots on unsupported section of the (relatively thin) seatstays and fork blades. When you brake, the brake cable pulls upward, which tends to splay the brake posts outward. In addition, the pads are dragged along by the rim, which also tends to twist the brake. On the front, these two factors reinforce each other.

The fork blades twist, and this changes the angle at which the brake pads hit the rim. That is why you toe in the pads, which reduces the effect. But there is still a non-linearity as the pad surface increases as you brake harder.

For most rides, it’s not a huge deal, but when you brake deep into turns during twisty mountain descents, a brake that responds linearly to your input gives more confidence. And in the Cascade Mountains, we have plenty of twisty descents. When curve follows upon curve, when your instincts take over and your bike feels hardwired into your brain, then you want a brake that responds with linear force to your inputs. A brake where each increment of lever pull results in the same incremental increase of braking force.

That is where centerpulls come in. They eliminate the twisting problem by locating the pivots above the rim, where the stays (rear) and fork blades are well-braced. The result: There is no twist, the pad angle doesn’t change, and the brake action is linear and easy to modulate.

Modern racing brakes use the same pivot location – only the upper arms are more complex to eliminate the need for a straddle cable. Many of the best bikes now have direct-mount brakes, where the pivots are part of the frame, which further reduces flex (and weight). When we reintroduced direct-mount centerpull brakes, they were seen as oddballs. Today, they have become the norm.

Straddle cables have fallen from favor because they can cause lost motion. A thick straddle cable, as in the photo above, tends to curve over the straddle cable yoke. When you pull on the brake lever, the first part of the lever travel only pulls the straddle cable straight, without actually slowing you down.

Lever travel limits the power of all brakes: In theory, you could make the brake more powerful by increasing its mechanical advantage, but then the pads travel less for each increment of lever travel. And you can only pull the lever so far until it hits the handlebars. If you are wasting some of the lever travel to pull the straddle cable straight, you have less left over for the actual braking. You have to design your brake with less mechanical advantage – less braking power. And/or you need to set the pads closer to the rim, which increases the chance that they’ll rub if your wheel goes slightly out of true or if your brake goes slightly out of adjustment. (That is why discs tend to rub: They have a lot of mechanical advantage, so the gap between disc and rotor has to be tiny.)

There is a solution:  Use a thinner straddle cable that doesn’t bow. The straddle cable transmits less force than the brake cable, so a thinner cable works fine. (We use a derailleur cable, so replacements are easy to find.) The thinner cable bends smoothly around the straddle cable yoke (above). There is no lost motion when you apply the brake. Without the risk of bottoming out the brake lever, we had the freedom to design the Rene Herse brakes with more mechanical advantage. That way, we get as much brake power as a very good mechanical disc brake.

All the mechanical advantage in the world doesn’t do much if the brake flexes instead of squeezing the pads. Brake flex means less power for slowing the wheel. Most of the flex occurs between the pivots and the pads. This part of the brake twists when the pads are dragged along by the rim. The upper arms can be thin: They are stressed mostly in one plane (up/down). That is why centerpull brakes can be superlight: Their pads are much closer to the pivots than those of old-fashioned sidepull and dual-pivot brakes.

Not all centerpull brakes are created equal. The arms of our Rene Herse centerpulls have been optimized using Finite Element Analysis. We forge the brake arms for optimum strength, so we can make them thinner and lighter than CNC-machined arms. In fact, Rene Herse centerpulls are among the lightest brakes out there.

All our brakes are now available with titanium eyebolts for the pads. The centerpulls weigh just 137 g (per wheel, with pad holders, but without pads*). That is the same as a direct-mount Dura-Ace brake, even though the Rene Herse has room for 42 mm tires and fenders, while the Dura-Ace clears just 28 mm tires (without fenders).

For the titanium version of our brakes, we’re also using a titanium lower bolt for our Straddle Cable Yoke to save further weight. (The upper bolt is always made from super-strong CrMo steel, since it secures the brake cable.) The steel-bolt version of the brake remains available as a more affordable option.

The new custom-made titanium bolts are available separately, too. They are great for attaching bottle cages and fenders. (Please don’t use them on racks, where the full strength of steel bolts is needed!)

Light weight, excellent power, great modulation, low maintenance, and the ability to use flexible fork blades for comfort and speed – those are the reasons why I chose Rene Herse centerpull brakes for my new bike.

Further reading:

* Rather than get into a competition for the lightest (meaning: thinnest) brake pads, we weigh our brakes without pads. That way, we can use thick brake pads that last three times as long as the thin pads of most modern brakes. If you want ultralight pads, you can cut them down (or run well-used pads).

About Jan Heine

Spirited rides that zig-zag across mountain ranges. Bicycle Quarterly magazine and its sister company, Rene Herse Cycles, that turns our research into the high-performance components we need for our adventures.
This entry was posted in Brakes. Bookmark the permalink.

104 Responses to Why I Choose Centerpull Brakes

  1. Christian Bratina says:

    My DeSalvo Ti dirt road bike has a Paul Racer on small diameter, thin seat stays. I would love to compare it with the Rene Herse centerpull if you can provide threaded brake bosses for it.

    I chose to forego discs because they would require stiffer frame forks and i would not be able to swap out wheels readily with my other bikes. And I have found good cantilevers and linear pull brakes are more than adequate, even on our tandem.

    • Jan Heine says:

      The Paul Racer centerpull is also quite powerful. Way back, I worked with Paul on the design. He was the first to see the advantages of centerpulls, way back before.

      Compared to our brakes, Paul CNC machines his parts, so his Racers are bulkier to make up for the lower strength of the manufacturing technique. That makes them a bit heavier, and they don’t open quite wide enough to let through a wide tire without letting out some air. But those are minor points that you don’t notice when you ride with them!

      • Christian Bratina says:

        Are there threaded brake bosses that fit the Rene Herse so I can do a direct comparison?

      • Jan Heine says:

        The Rene Herse brakes fit on standard centerpull bosses, but I don’t know of any threaded ones. Ours are one-piece, to be brazed onto the frame.

  2. Ron says:

    The problem I’m having, is apparently you can’t (or shouldn’t) use modern brifters with centerpull brakes? Apparently when Shimano changed from having their brake lines from the brifters routed from out in front of the bars (in the way of a handlebar bag) to traveling under the bar tape, they also changed the amount of cable pull provided by the brifter. Basically making their brifters only compatible with their brakes. And was told the same situation with Campy.
    While everything that you espouse about bikes I agree with… not sure I’m going to enjoy reaching down to use down-tube mounted shifters. For the front derailleur it might not be so bad, but for the rear I tend to do more shifting as I’m going through rolling country.
    I have two bikes with Ultegra 11-speed and love them. But the custom randonneur I have being built will have down-tube shifters, and I hope I’m not going to regret the money I’m spending on it.

    • Jan Heine says:

      I haven’t run our centerpulls with the latest brifters. They work great with older generations. But they are quite forgiving – old Mafac levers have way more lever pull than aero levers, and yet both work great with the Rene Herse centerpulls. We’ll have to do some testing, but I suspect it’ll work fine.

      Downtube shift levers aren’t a big issue. On long rides, it’s good to move your hands off the bars frequently to prevent numbness, and doing this to shift is an easy way to do it. Downtube shifters ‘fall easily to hand’ (as they used to say). The only real advantage of brifters is that you can shift while riding out of the saddle – useful especially in the city.

      • Kevin says:

        Hmmm…Ron’s post might (or might not) explain the very poor braking on my bike. Compass centerpulls with Ultegra 6700 series brifters. Quite poor in the dry, frightening in the wet! I am studying the close up photos of Jan’s new bike and comparing set up – need to have a bit of a play with mine.

    • Steve Palincsar says:

      @Ron – You said, “Basically making their brifters only compatible with their brakes. And was told the same situation with Campy.” I can tell you this much from personal experience: I am using gutted 11-speed Campagnolo Chorus Ergopower levers (i.e., the shifter parts have been removed, after having failed) as brake levers on my JP Weigle, which is equipped with Rene Herse cantilevers. They work perfectly. Jan’s CdM Weigle has gutted 10-speed Ergo levers for brake levers, and they work perfectly too.

      • Jan Heine says:

        Looking at the pivot locations, the difference in mechanical advantage between the different lever generations cannot be very large.

      • Frank says:

        If using Ergopower levers in such a fashion it is better to keep the central shaft on which the shifting bits are mounted, as it also has a structural role.

    • tonebone says:

      I recently set up a pair of brazed paul racer centerpulls with Ultegra R8000 levers and they worked great

    • Bob C says:

      FWIW, I run Compass Centerpulls on several bikes using both Campy (Chorus 10 speed) and Shimano (Dura Ace 11 speed) with absolutely no incompatibilities. Plenty of braking power, great modulation, etc. I know several other people with similar setups and no issues with compatibility.

    • Mitchell Hull says:

      I run brazed-on MAFAC Raids (as I understand it, functionally identical to RH) using Campy 11 brifters on my single bike which has 584-42 BSPs and fenders. They work great.

  3. Willem says:

    My loaded tourer uses Magura HS66 hydraulic rim brakes because they can handle 52+ mm tyres, modulate very well, and are powerful enough to cope with a heavy load. However, the new ‘fast tourer’ that I am planning for day rides and credit card tours will have 42 mm tyres and that bike will indeed have centre pulls on braze ons for that precise feeling that the cantilever brakes on my current fast bike are missing.

  4. Had bolt on Zeus 2000 centre pull brakes on my first custom frame. Loved them; they seemed much more powerful than Campagnolo side pulls of that era. And they were beautiful. But I moved on to MAFAC tandem cantilevers for subsequent bikes cos I couldn’t get braze on posts for the Zeus brakes. The MAFCs were powerful if somewhat crude. Recently I’ve used Avid cantilevers (useless and over rated), Paul Mini Motos (strong but not enough clearance for mudguards); TRP mini cantilevers (poor quality fittings that failed); TRP HY/RD (powerful and well modulated but fiddly); and Shimano 805 hydraulic discs (powerful but fiddly, pads rub & wear quickly if organic; wear slowly if metallic, but are often impossible to fit around the disc when new; are bastards to work on when travelling far from home; and have vulnerable hydraulic hoses when transporting. They also locked me into brifters, which I found uncomfortable and less convenient than bar end shifters. My latest bike has RH c/pulls. These are simple, powerful from the drops with TRP levers, but require more hand strength than discs or Minimotos from the hoods; they’re light and classic looking, and they modulate extremely well. They’re not as ultimately powerful as the Shimano discs but they seem more usable. Even though they don’t look as good as the Zeus c/ps, they’re still pretty, they’re more powerful ( cos of braze on posts, I suspect), fit wider tyres, and I love them more.

  5. PStuart says:

    Since your fork is flexible, do you run into the problem of brake chatter/shudder that some experience with cantilever brakes (e.g., https://www.velonews.com/2010/09/cyclocross/technical-qa-with-lennard-zinn-return-to-cross_101807)?

    • Jan Heine says:

      Absolutely no brake judder – on dozens of bikes we’ve spec’d with these fork blades, both with centerpulls and cantis.

      The judder on cantis is caused by a number of factors. Mostly, it’s an issue of the fork blades actually being too stiff compared to the steerer tube. So the steerer flexes as you brake, which releases the brake tension, which in turn unflexes the steerer, which increases the brake tension and so on.

      Beyond that, it’s also due to the twisting of the brake. Going from the really flexy Mafac cantis to the much-stiffer Rene Herse has cured the brake judder on my cyclocross bike…

  6. PK says:

    A disc brake stops the hub. The spokes stop the bike and I have seen more cracked eyelets in my shop on disc bikes than on rim brake-equipped bikes by far. Replacing a cracked rim is no small feat especially when OEM rims are not always available as repair parts.
    I have found that a bike built for its intended use from even mid-range parts performs well and lasts nearly forever. A bike built with the highest quality parts and abused will fail. If I was going on a plane to Paris for a huge event I’d never even consider untested, super light, esoteric parts. Strictly proven, developed, ubiquitous parts for me (der wires, pads, free wheels, 27.2 post, etc.)

    • Johan Brox says:

      Cracked spokes: Recently I have come to discover that not all wheelbuilders consider the elasticity of the spokes together with the elasticity of the rim when picking components. Some use thick, inelastic spokes on lightweight, elastic rims, which can lead to spokes pulling loose from the rim as the rim flexes more than the spokes allow.

      High-profile rims are stiffer per kg, but have lead many manufacturers to stop using crack-preventing eyelets. I can only imagine that disc brakes exacerbate the problem, both for the reason you describe, and because rim producers can use thinner extrusions since the sides of the rim are no longer subject to brake wear.

  7. Dr J says:

    How about simple, old v-brakes?
    + cheap (starting at $20 per wheel)
    + light (~145g with pads)
    + easy to setup and adjust
    + fit any frame and fork with brake boses, unlike centerpulls
    + much more powerful than cantis
    + easily replaceable, with spare parts available in any bike shop
    + don’t require cable hangers
    – may flex some lighter frames or forks during braking

    • Jan Heine says:

      V-brakes are powerful. As you mention, fork twist results in relatively poor modulation. That isn’t a big deal in flat regions, where you brake hard only during emergency stops… but it’s less ideal for twisty mountain descents.

    • Conrad says:

      As I unfortunately do not own a custom bike as of yet, that is what I do. Cane Creek V brake levers paired with V brakes with salmon pads are really, really good. I have them on 3 or 4 bikes. Modulation is excellent and they are never wanting for power, even when it is wet. The only problem is that they interfere with a front rack that attaches at the usual hole in the fork crown. I use a Nitto big front rack that gets around that issue.

  8. Larry T. atCycleItalia says:

    Just another example of CHANGE not being the same as improvement. How many of us were seduced back-in-the-day by the sexy looks of the new sidepull brakes? “Centerpulls?” they said, “How quaint!” Same s–t, different day now with hydraulic disk brakes being “the thing”.

    • Jan Heine says:

      I think the main advantage of sidepulls was that they were dead-easy to set up.

      • Stuart Fogg says:

        WAY back (ca. 1975) Campy Record sidepulls felt better to me than anything else. Perhaps it was due to the pads or to generally high quality, dunno. I used the same set until about 2015.

      • Jan Heine says:

        I also raced on Campy sidepulls. It wasn’t a problem when I was riding at the front, but when I was in the middle of the pack, I almost ran into the riders in front of me who had dual-pivot brakes…

    • John C. Wilson says:

      The note about running into back of riders in front with dual pivots brings back memories. Much of that was due to the guys with the newer brakes simply grabbing too much brake on a system that had a very light pull. Nuovo Record brakes would lift the rear wheel as well as any other good brake, they did require hand strength to do so. Racers were assumed to have hand strength and who else would be using such brakes.

      Have noticed a lot of inconsistency over the years mixing Shimano levers with Mafac brakes. Until proven otherwise it would be a good idea to assume compatibility is a maybe. Modern shop practice is to run the brake pads very close to the rim. The pad>rim distance in the photos above is immense compared to what I see day to day on high-end bikes. My own bikes the gap is even greater. I began when the object was to be able to remove the Paris-Roubaix shod wheels w/o even thinking about cable release. Still works fine, not going to sell it to anyone young.

      • Jan Heine says:

        The beauty of a stiff brake with little flex or lost motion is that none of your lever travel is wasted. So you can set the brake pads quite far from the rim and not risk running out of lever travel. Even when I squeeze the levers with all my might while the bike is standing, I don’t come close to the bars. This means that even if my wheel goes out of true (unlikely with wide tires cushioning the blows to the rim), it won’t rub on the rim.

        If someone wants their pads closer to the rim, it’s easy to set them up that way!

  9. PS. I always get confused about what is meant by high or low mechanical advantage. Do high mech. adv. levers have more or less cable pull than low mech. adv. levers? I know you say RH c/ps can work well with most levers, but which are optimal—low or high m.a. levers? Are TRP levers sufficient?

    • Jan Heine says:

      High mechanical advantage means that a lot of travel pulls little cable. So there is a lot of power, but the pads have to be close to the rim. If you imagine a simple set of pliers, high mechanical advantage means long handles…

  10. gcziko says:

    Why does the straddle cable transmit less force than the brake cable? Is that because its force is divided between the two brake arms?

    • Jan Heine says:

      Yes. Unless you run the straddle cable very low (which doesn’t provide any advantages with centerpulls), you divide the brake force among two cables.

    • Frank Krygowski says:

      It’s a little more complicated. It also depends on the angles involved. It’s a vector equilibrium problem, an upside down version of the following example problem with the Clown Picture at https://www.physicsclassroom.com/class/vectors/Lesson-3/Equilibrium-and-Statics In our problem, the brake cable tension takes the place of the picture’s weight.

      The vertical component of each angled tension is the tension times the sine of its angle to horizontal. If that angle is 30 degrees, (whose sine is 0.5) the tension in the clown hanger wires equal the weight. (Summation of all vertical forces must equal zero.) In our brake cable case, for 30 degrees, the straddle cable tension equals the main cable tension. For angles to horizontal smaller than 30 degrees, the straddle cable actually has more tension than the main cable.

      I don’t think it’s a problem, though. In my experience, cables break from fatigue in bending, not from straight tension.

      • Jan Heine says:

        You are right. We did that calculation when we developed our brakes. The thinner straddle cables are more than sufficient.

        Also correct that straddle cables fray from fatigue during bending, right where they are clamped to the brake arm. Since the angle between brake and straddle cable changes as the brake is applied, the cable flexes a lot right in that one spot. The solution is to make the straddle cable anchor so it swivels and never gets flexed. Then the straddle cable lasts for decades.

  11. Dann Zinke says:

    “That is why discs tend to rub: They have a lot of mechanical advantage, so the gap between disc and rotor has to be tiny.”

    Do you think this advantage could be reduced (and hence the gap increased for the better) by using short-pull levers with long-pull disc calipers?

    • Jan Heine says:

      It’s the overall mechanical advantage that matters here – it doesn’t matter whether it’s at the lever or at the caliper. If you made the rotor larger, you wouldn’t need as much clamping power, and thus you could get more clearance. However, a larger rotor is more likely to develop a wobble, so probably you wouldn’t gain much.

  12. Ted Durant says:

    What’s the maximum practical tire width for Compass Centerpull brakes?

  13. Derek says:

    I don’t dispute that fork flex can be beneficial, or that rim brakes allow more flexible blades, but I’m surprised that you claim 30mm of movement. How did you measure that?

    • Jan Heine says:

      I went back and checked our notes, and we documented only 15 mm flex. We didn’t take the setup to the max – simulating a tire that bottoms out on the rims – but I changed the blog text to make sure we don’t claim something we haven’t tested. We’ll do more testing and hopefully get a better measure for this.

    • Francisco says:

      Using a simple FEA model for a Columbus SL-tubed frameset I easily reach elastic deflections in that range. The steerer tube, top tube and down tube contribute significantly to the total deflection.

  14. Matej says:

    Could RH centerpulls be installed on a Cinelli Tutto frame that has front and rear removable V-Brake pivot studs? I find v-brakes aesthetically not pleasing, even mini V-brakes. The frame does allow for 35 mm tyres (Bon Jon Pass it will be). The bike will be fixed gear with brakes for safety reasons and modulation on downhills. This years PBP has really opened my eyes toward fixed gear freedom. Thank you in advance.

  15. Jacob Musha says:

    Is there a plan to make a longer straddle cable for the Rene Herse cantilever brake? I would love to try them, but my bikes use >52mm tires and fenders, and the specs say there is only clearance for 42mm tires and fenders.

  16. steve parkes says:

    “Discs require more maintenance and care – the hydraulic houses are prone to kinking, and the pads often rub noisily on the rotors, requiring frequent adjustments of the calipers to recenter them.”
    In my shop and personal experience, disc brakes need less adjustment and the pads last longer than rim brakes given average rider case use. Add to that the availability of good performing cable actuated brakes for short and long pull levers at low cost, and I’m not sure the above statement is entirely honest or accurate. There’s enough fear of disc brakes out there without fanning the flames.

    • Jan Heine says:

      All Bicycle Quarterly test bikes with disc brakes needed adjusting the calipers at least once during our testing. I can’t think of a bike with rim brakes that needed adjusting in just 200-500 miles.

    • Conrad says:

      I have seen riders go entirely through a set of disc pads in one race. Then they have no brakes. I can usually get two seasons out of my cantilever pads.

      • samuli says:

        I used hydraulic discs with sintered pads in a 7400 km race with mountains, rain and gravel and then continued to use the same pads for about a year, through the winter. Didn’t adjust them either.

  17. Bob C says:

    I use Compass (oops, Rene Herse!) centerpulls on two bikes. I like them enormously EXCEPT for the constant messing with them required to prevent the hellish squealing they can emit. Even with the indented washers you sell to provide a bit of toe in, the brakes seldom stay silent for long. I’ve taken to profiling the pads the with a wood rasp (no kidding) to provide enough of an angle to the pads to prevent the squealing.

    It’s a profound aesthetic problem, it’s a just a horrible noise and every time I hear it I curse the brakes. It takes some of the joy from bike riding as you wince every stop But it is also something of a safety problem. I live in San Francisco with a LOT of pedestrians around and I’ve had pedestrians jump away and fall down because they heard my brakes shrieking as I came to a stop at a crosswalk behind them. They had no idea what the banshee was that was heading toward them.

    I realize Jan says this doesn’t happen or has minimized the problem in the past. But everyone I know who has these brakes has struggled with it and intensely dislikes the ear piercing squealing. This is 100% an engineering issue that you can solve if you just provide users with an effective means of adjusting the toe in. The washers don’t cut it. Please remedy this. Your loyal customers are begging, and I mean BEGGING, for some relief on this score… It’s a major flaw – the only one, frankly — and it can absolutely be fixed if you choose to.

    • Jan Heine says:

      Hi Bob,
      I am sorry to hear about this issue. If even adding toe-in by reprofiling the pads doesn’t help, then either a) toe-in isn’t the problem or b) your brake posts are so much out of alignment that your pads are still toed out even after increasing the toe-in to the max. Please send me some photos, so we can try to trouble-shoot this issue. As you mention, it’s extremely unusual for this to happen.

      • Bob C says:

        Thanks for the response. The reprofiling of the pads with the rasp generally works for about, hmmm, maybe 600-700 miles or so. Depends a bit on how much rain I encounter over those miles. FWIW, both bikes are MAPs and as such I would be surprised if the posts are out of alignment. I’ll send some photos. Also I got a message from a friend who saw this and he tells me that moving to Koolstop Black pads completely stopped the squealing on his RH cantis — at a tiny cost of braking power. Might try that as well.

      • Jan Heine says:

        Sometimes, there is an incompatibility between pad and rim materials. This was my next suggestion – use a different pad like the Kool-Stop black ones.

    • I have RH c/p brakes on my latest bike, custom built by Ewen Gellie here in Australia. Extremely accurately assembled and installed brake posts. Kool Stop salmon pads. So far no squealing whatsoever from the get go, wet or dry, on the flat or on fast descents.

      As for discs: powerful, yes. Squealing? Not much with metal pads. A pain to maintain, esp. when travelling? Yes. YRMV.

  18. davidmcneary says:

    re: straddle cable bowing on that XTR canti – why not coax the straddle cable around the yoke with a convincing set of pliers? that was how I was taught to set up cantilevers properly:
    https://blackmtncycles.com/get-the-most-out-of-your-canti-brake/

    • Jan Heine says:

      That is what I did on the green bike in the photo after taking the photo. When dealing with an existing brake, this ‘fix’ is fine, but when designing a brand-new brake, why not do it properly in the first place?

  19. Rick Thompson says:

    I’m sure that I do not ride as hard as you guys, so may not appreciate the performance at the limits, but my RH Rinko centerpulls stop as well as any other brake I use, including discs.
    I take the Fitz apart pretty much every week to pack in a car, these brakes make it easy by just popping off the straddles and coiling up the cables. I can Rinko pack for the car (not strapping the bundle or bagging) in about 8 minutes now, and with no tools needed.
    The same Rinko features make it easy to swap handlebars. I have drop and flat bar cockpits complete with full brake cables that swap in a few minutes.
    One minor issue is the straddle yoke. On first set up, I can run a long brake cable past the pulley, adjust length on the bike, mark the cable, then pull through and cut the cable to length. Not too bad. To adjust later for stretch and wear, I now have to remove the pulley, pull out some cable, cut a little off, re-install everything, then check if the length is good. If not, repeat. For brake levers with adjusters this tweaking can be coarse, but for levers with no adjusters it’s hard to get the cable length just right. Is there any way to put some easy adjustment into the yoke design?

    • Jan Heine says:

      You should never have to touch the straddle cable yoke once you’ve set it up. The adjustment for pad wear is part of the brake pad design: When new, set the pads so they are close to the brake arms. As they wear, simply loosen the 10 mm nut and slide them closer to the rim. That way, the geometry of the brake doesn’t change as the pads wear. Since the toe-in is set only once and doesn’t need to be adjusted every time the pads are loosened, sliding the pads closer to the rim takes only a few seconds.

  20. Just wondering if you could think Antelope Hill tyres in without fenders?
    I’m looking at a build that would take Antelope Hill with tiny clearances but have enough space for fenders and smaller tyres as a sort of Monster Road thing

  21. Brendan says:

    What is your analysis on the old Shimano U brakes? Do they perform just as well, other than being chunkier?

    • Jan Heine says:

      I haven’t used them, but C. S. Hirose, the great Japanese builder, uses them as centerpulls on his tandems. There is no reason why they shouldn’t work well.

    • Conrad says:

      I have an XT U brake on the back of an old mountain bike. It is a good brake. Pad setup can be challenging with narrow rims. The old Araya RM 20 worked well because the rim sidewalls were not quite parallel .

  22. Matthew J says:

    Are there mechanical reasons no one is making wider longer reach direct mount caliper options?

    • Jan Heine says:

      Like a modern racing brake, but with longer reach? The only disadvantage over a centerpull is the weight. With longer lower arms, you also need longer upper arms, and those will flex a lot if they aren’t made very bulky. The straddle cable is a very elegant solution, as it replaces the parts that are in tension with a thin and light wire.

  23. Matt B. says:

    Does anyone have experience using modern, high-tech, synthetic cordage for the centerpull straddle cable? Something like 1.2mm or 1.5mm diameter dyneema SK90/SK99 kite-surfing control lines or tent support guy lines. Also, various braided fishing lines are also made from dyneema or spectra. Should be “stronger than steel,” stiff (in tension) enough, and **probably** more flexible (in bending) than steel bike cables. The pre-stretched, heat treated SK99 dyneema cord might be ideal, if it can be sourced and someone can figure out how to terminate the ends to eliminate the slack in the straddle cable. Maybe just a simple Brummell eye splice replaces the barrel on one end.

  24. Frank Krygowski says:

    Regarding lost motion in cantilever straddle cables (7th photo, green mountain bike): I recognized that problem long ago and solved it to my satisfaction by pre-bending the cable around the saddle, so it runs in straight lines from the saddle to each brake arm.

    I’ve since used that trick with every canti equipped bike I own, but it’s especially valuable on our tandem. I’ve set up its cantilevers with a shorter straddle cable for more mechanical advantage. As you say, that means less travel of the brake shoes, so any wasted motion is a larger problem.

  25. Tommo says:

    A bit off-topic, but still on the subject of your new PBP bike.
    I was wondering if Mark used the Rene Herse bottom bracket shell and chainstay combination for construction, or are these customized?

  26. Drew says:

    One important factor left out in the discussion about rim brakes here is the wear of the rim sidewalls, which is something rim brake users need to keep in mind. Having a worn rim crack could end the ride, and/or cause a crash from the tire blowing off. The rim brake slowly wears down the structure that keeps the inflated tire where it belongs.

    On my road bike with rim brakes I occasionally hold a straight edge against the rim to see how much of a gap there is where the pad makes contact. Anything more than a half millimeter may be too much, and I replace it; especially if it is a front rim.
    Another technique is (after putting on earmuffs) is I inflate the tire to twice the pressure I normally use. If the rim bows out or cracks anywhere, it’s obviously worn out.
    If you ever feel an of bump-bump-bump when braking, check your rim. It may have just cracked.

    • Jan Heine says:

      Excellent point. There is a way to measure the remaining rim wall thickness with calipers: Place a thin metal rod (like a piece from a spoke) on the sidewall and another inside the rim, so it sticks out beyond the bead hook. Measure the total thickness with calipers, deduct the diameters of rods, and you have the remaining wall thickness. You need at least 0.8 mm, better 1.0 mm.

      Usually, front rims last a long time, while rears wear quickly if you brake a lot on the rear. (During hard braking, the rear wheel barely touches the ground, so proper technique is to use the front brake only unless it’s so slippery that the front wheel might skids and lose traction.)

      With disc brakes, you don’t have that problem, but you do have pads that can wear very quickly. I wore out a set of OEM pads on my TRP Spyres during a single, long mountain descent in Mexico, where traffic didn’t allow me to let the bike roll…

  27. Tim says:

    I have successfully rebuilt my pair of c1972 mafac “Racers” with your various online catalogs materials. “Thanks” doesn’t cover it, I really appreciate your work. They are better than original, perhaps in part because the new pad material? Anyways, happy trails.

    • Jan Heine says:

      Really glad you could rebuild your brakes and make them work better than new. You are right, the new pads are better, and so are the bushings. The original pad holders also could develop play. (Ours are one-piece to prevent the post from loosening.)

      The original Mafacs were brilliant engineering, and the first ones in the 1950s were quite well-made, since they were marketed as the best (and most expensive) brakes in the pro peloton. As Mafac as a brand moved downmarket over time, the quality did, too. But the forgings for the arms were always excellent, and the remaining parts can be replaced using our hardware.

  28. Peter Alspach says:

    How is the performance of centerpull brakes that don’t have specific brake bosses but instead use a typical center mount attachment? Curious about performance for retro fit to frames without the expense of adding braze ons.

    • Jan Heine says:

      Bolt-on centerpulls can work very well, but the difference is noticeable. I’d recommend adding braze-ons to the fork and using a bolt-on rear. Then you only have to repaint the fork. Or get a new fork with more flexible blades, a front rack, and all the other things that most older bikes lack.

      A bigger issue is that some centerpulls work better than others. Especially the wide-tire Dia-Compes don’t seem to brake very well. The mechanical advantage is quite different, and the arms seem to have more flex. If you go the bolt-on route, try to find some old Mafacs Raids, or get our bolt-on Rene Herse brakes. If you don’t need as much clearance, the Mafac Racers are plentiful and work very well. You’ll probably want to replace the bushing and pads (and maybe the rest of the hardware) – we offer kits that make this easy.

  29. alderbanks says:

    Those are beautiful center-pulls. My question refers to the distance between the pivot point on the cantilever arms versus the center-pulls. The center-pull pivot looks like it is twice as far away from the rim. If the arms are of similar thickness, or stiffness, one would think that the twisting phenomena would be more of an issue with the center-pulls.

    • Jan Heine says:

      It’s the fork blades that twist, not the brake itself. The brakes, if well-designed, are quite stiff. You are right, the lower arms on a centerpull are longer, so the upper arms are much longer, too, to get the same mechanical advantage.

  30. Ronan B says:

    How did you manage to kink a hydraulic hose?

  31. Raoul Morley says:

    Would your Centrepull brakes clear an Antelope Hill tyre without mudguards?

  32. John Clay says:

    Cantilever brakes make a good case for Continental oval for blades. They are more resistant to flecture from the loading profile placed on them by cantis than Imperial blades. A Little flexibility is likely lost but with the better, butted blades I’d expect the loss to be acceptable. One could always have them drawn with a lower leg profile similar to those found on your Imperial oval blades.

    • Jan Heine says:

      Absolutely. I’ve even been thinking of using round fork blades on a bike with cantilever brakes. Making flexible round fork blades would be easy – all fork blades start our round before they are squeezed oval. Making a fork crown that fits the correct diameter would be the only issue, but a twin-plate crown with round holes is not that difficult to make…

  33. Bill Lindsay says:

    I’m skeptical that a flexy fork really gives 30mm of travel. I’d believe 10-15mm, but 30mm sounds like a lot. My Lauf Grit suspension fork offers 30mm of travel. I’ve never bottomed it out but can claim that it flexes a lot more than my Norther/Lyon flexy rando fork. Did you measure this in your static load setup?

    • Jan Heine says:

      See above – we’ve documented 15 mm, and I changed the post to correct this.

    • Francisco says:

      [[I’m skeptical that a flexy fork really gives 30mm of travel]]

      There is a difference between how much a structure can safely flex and how much it flexes in normal use. Jan demonstrated 15mm of deflection in a light steel fork, and it will flex at least another 15mm before reaching its elastic limit if my FEA model can be trusted. So lets say it has more or less the same total “travel” as your Lauf fork. The difference is in the spring rates: the Lauf will bottom out at a fraction of the force needed to fully flex the steel fork; conversely, only a fraction of the potential deflection of the steel fork is accessed in normal use. But it is still there.

  34. Scott Turnbull says:

    Great run down. For my commuter I still prefer disc brakes for the initial bite in the wet and the fact I wear a rotor rather than the rim. I ride by a number of schools and the number of times I need to emergency brake due to parents in a hurry is shameful. I also ride in sloppy, muddy trails as part of my daily commute so disc brakes excel here as well. I do however agree that the howling can be annoying and loud especially with sintered pads. That said, for me at least it tends to be intermittent when conditions change (e.g., dry to wet), but goes away under constant conditions (e.g., contested wet). I believe it has something to do with the material transferred from the pad to the rotor. I think it could be an interesting article researching how to deal with this issue.

  35. James says:

    For the R. Herse brake to function properly, do the pads need to be toed in or can they run parallel to the rim?

    • Jan Heine says:

      On centerpulls, there isn’t much fork blade twist, so the pads can be parallel to the rim. If the brake makes a lot of noise even after bedding the pads, a little toe-in will quieten it.

  36. Derek says:

    I would like to have a bike with RH brakes on it but I like Rat Trap Pass tires so much. Neither of your brake designs fit over 54 mm tires and fenders. I don’t want to go back to tires that are “only” 42 mm, or give up fenders. Will you please offer a longer straddle wire for the RH canti brakes at some point?

  37. Tim Quinlan says:

    Hi Jan, I have a slightly off-topic question. What headset are you using on your new bike? Looks like the top is a King and the bottom is something else.

  38. Willem says:

    A few people have mentioned V brakes. These can indeed be excellent for heavily loaded bikes with wide tyres, but I want to stress that they cannot be combined with normal road brake levers because the cable pull is different. You will need the special Tektro drop bar levers for V brakes (which rules out sti).

    • Derek says:

      A Problem Solvers travel agent will allow sti and v brakes to work together. I changed one of my tandems to this setup 5000 miles ago. I wouldn’t spec a new bike this way but it gets the job done.

    • CHRISTIAN Bratina says:

      You can use V brakes with 9, 10, and 11 speed Shimano/Campy brifters by adding a Problem Solvers Travel Agent. They are commonly used on tandems, including ours. You should use a short arm V brake with them, such as Paul’s Minimoto on a single bike to avoid too much leverage.

    • Martin says:

      Nice article! Practical Question: what is the recommended torque Range for the screws that fix the compass centerpulls to the pivot axles?

      • Jan Heine says:

        We don’t spec torque settings because consumer torque wrenches have too much variation. The bolts are dimensioned so that a standard wrench (box or open-ended) will provide the correct torque with a moderately high amount of hand force. (That is why a 10 mm wrench is longer than an 8 mm, etc. It automatically gives you the right torque once you’ve learned how much hand force to apply.)

        The bolts are made from ultra-strong CroMoly steel, so it’s hard to overtorque them. Lubricate the threads (grease, beeswax, Loctite) to make sure that there isn’t excess friction that stops the bolt before it is tight. Also, if the bolt doesn’t go in smoothly, chase the threads with a tap. Sometimes, the bosses get distorted a bit during brazing…

  39. I have always wanted to know what’s the difference between centerpull and Sidepull brakes. The classic Campagnolo/Weinmann/Shimano sidepull design doesn’t use a straddle cable, but aside from that, is there any advantage?

    • Jan Heine says:

      A sidepull has the pivot above the wheel, so there is a long brake arm that can flex. Also, the pivot is a thin 6 mm bolt – again adding flex. The centerpull has the pivots next to the tire, close to the rim. It’s basically an upside-down cantilever. Even the bolt-on versions can be very stiff, since the arched piece with the pivots doesn’t move, so it can be bolted firmly to the fork crown. Direct mount always is better, of course, as it is even stiffer (and lighter).

  40. Mike Morrison says:

    I prefer cantilevers (old-school, not v-brakes) for their greater tire and fender clearance. Unfortunately, my bike came stock with cantilevers on only the front wheel. The rear wheel has U-brakes, a.k.a. centerpulls. I have no complaints about their stopping power, but locating the bulk of the brake mech close to the tire really limits the tire and fender clearance. I could clear 60 mm tires in the rear easily without the u-brakes, but with them, I can just squeeze in Rat Traps and the Honjo fenders. Thus I’d like to (someday!) pay for cantilever braze-ons for the rear wheel.

Comments are closed.