A Photo Says More than 1000 Words


This photo shows our Compass centerpull brake with the straddle cable released. (It also shows a Rinko fender, and a Babyshoe Pass Extralight tire after about 5000 kilometers…) More than a thousand words, this photo explains the design of our brakes:

  • The arms fit around 42 mm-wide tires and fenders with generous clearances.
  • You can install the wheel with its 42 mm tire inflated. (Most other brakes for wide tires require deflating the tire every time you remove the wheel.)
  • The brake is dimensioned so the pads are about 2/3 of the way down – not at the limit. This gives you some room for adjustments.
  • When the pads wear, it’s easy to slide them a bit further inward, without having to completely re-adjust the brake.


With the straddle cable closed, you can see how nicely the arms are profiled to fit over the fender. It’s a small detail, but it makes the bike so much more beautiful. (It’s also one of the reasons why the brake opens so wide.)

The slender arms don’t just look nice, but they also make this one of the lightest brakes on the market – lighter than Dura-Ace racing brakes. We used Finite Element Analysis to optimize the brakes’ stiffness, so they offer excellent braking power and superb modulation. And the arms are forged, not CNC-machined, so they are plenty strong despite their superlight weight.

Fender clearances are an important topic for “real-world” bicycles. The photos show 20 mm clearance between frame and tire. That is more than you find on most “fender-ready” bikes today. But there is a good reason for this: With that much clearance, the fender won’t rub even if a fender stay gets bent. More importantly, small debris will clear the fender rather than risk getting stuck and cause the fender to collapse and jam against the frame.


Of course, you don’t want so much fender clearance that the bike has that “Motocross” look, but my “Mule”, the bike you see in the photos here, is far from that. A bike with well-judged fender clearances looks graceful, yet it’s supremely functional.


Our brakes come with detailed instructions on how to get all these clearances right. The Compass brakes are part of a system: They work perfectly with our superlight CP-1 rack. Our centerpull braze-ons are pre-mitered to fit Kaisei “Toei Special” fork blades. Of course, you can also use other components, but this system makes it easier to build a bike that is both beautiful and functional.


Beauty, light weight, performance and superior function in every way – that is what we strive for with every Compass part we design.

Click here for more information about our brakes.

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.

44 Responses to A Photo Says More than 1000 Words

  1. Greg says:

    Might you ever release a version of these to retrofit onto older bikes, with traditional brake center bolts? Seems a shame to limit them to custom-frame installations/ frame modifications only….. 650b conversions come to mind, since I am thinking of doing one soon, on a bike that just won’t work, as-is, with its insanely-tight clearances around 21 mm 700c tires. If I add bosses, I lose the valuable original finish, and will need to spend a whole lot more money….

    • Making a yoke that connects the brake arms, as on the old Mafacs, would require another forging die. (We wouldn’t CNC-machine it, as it would have to be very bulky to be strong enough.) Instead, I think I’d rather make a bigger brake for Enduro Allroad bikes… In the mean time, it should be easy to find a set of Mafac Raid yokes. Compass brakes will fit on those perfectly.

  2. thebvo says:

    Nice work.
    How do the Switchback Hill tires fit into the equation? Do they squeeze into the brakes fully inflated? Do they fit onto the Mule or the Herse with these same fender/brake/fork crown clearances?

  3. Steve says:

    Hi – they are beautiful brakes. One question – are they compatible with modern ergopower levers, in terms of cable pull?

    • Yes, they are totally compatible with modern levers. Most of us use them that way. Since the brake is so stiff, you don’t get much flex, so even with Ergopower levers (which pull less cable), you can set the pads far enough from the rim to clear a slightly out-of-true wheel.

  4. Niels Lillevang Hansen says:

    How “powerful” are these brakes compared to say the v-brakes or discs of a modern mtb, when using the same effort at the brake levers? IE are they effective with even a light touch?

  5. cbratina says:

    How can I install them on the cantilever posts used by my Paul Racer Centerpulls? Is there a threaded post I can swap out? How about a brake comparison test?

    • Paul Racer centerpulls use cantilever posts instead of centerpull posts, so you are limited to Paul’s brakes with that setup. If you want to swap brakes, there isn’t anything you can do except braze on new posts. If your posts are designed for centerpulls – especially old Mafacs – then it’s often an easy swap.

      We’ve been thinking about a brake comparison test for Bicycle Quarterly. We did one years ago, and it was quite interesting – we found that pad material makes a huge difference, and that during repeated very hard braking, Aheadsets and wheel quick releases will come loose.

      • Owen says:

        …kind of scary given how many of us run Aheadsets and and QR wheels. Could you summarize your findings? Did this occur often or only in extreme situations?

      • Only in extreme situations. But basically, after a dozen full-on stops from 30 mph on a 15% hill – this means smoking brake pads! – the Aheadset had play (stem had moved) and the wheel QR was very loose. Both had been installed correctly. Again, in normal braking, these things don’t happen.

  6. Bob says:

    I’m curious about the possibilities for even wider centerpull brakes. Could it be done? I ask because I’m planning/saving for custom bike and appreciate the excellent braking of centerpulls. I also appreciate 559×54 mm tires, though. I can—and currently do—get by just fine with cantilevers on two bikes, but I recognize their shortcomings compared to my one bike with centerpulls. Thanks.

    • If you don’t use fenders, you can fit 54 mm tires under these brakes without problems. But fenders are very much part of the Compass philosophy, so we don’t really advocate this.

      In theory, bigger centerpulls could be made. The weight of the brake would increase significantly – as makers of dual-pivot brakes have found, you can’t just scale up existing models – but since the Compass brakes are still extremely light, the end result would be quite acceptable weight-wise.

      The big question is whether there is a market. On gravel, cantilevers work fine, because you don’t have much traction, so you can’t brake hard enough to twist the fork blades significantly, eliminating the major shortcoming of cantilevers that leads to poor modulation. So these brakes would be useful mostly for riders who use tires wider than 44 mm (with fenders) on the road. And since for most road riding, 44 mm still seems to be the sweet spot, the market may be limited. Still, it’s a tantalizing idea…

  7. Zed says:

    With 20mm of fender clearance, can you still launch pebbles at cars, like Yehuda Moon does?

    • Pebble launching requires high pressures. We run our tires at 40 psi, so we roll over pebbles instead of having them dig into the tire and then catapult at cars. (But also far fewer flats, because glass and other sharp objects also don’t dig into the tire any longer.)

  8. Phillip Cowan says:

    May I ask what is the make of the frame pump on your “mule’. It looks very nice! I have an old frame that I’m working over and before I have it painted I’d like to add pump pegs to the chainstay. I would prefer to have the actual pump in hand before lighting the torch. I see lots of cheap $4.95 eBay specials out there but I wouldn’t bother putting on pump pegs for a low quality pump. I’m sure yours isn’t one of those.

  9. Murray says:

    A slightly off-topic question: is there an optimal yoke height for Compass/Raid brakes, or is it just a matter of taste? I assume the principles are the same as for cantilevers (higher yoke = less mechanical advantage), but I’ve heard that centerpulls are less sensitive to yoke height than cantis. I ask because a) your Mule has really high yokes while the “show” photo has a low yoke, and b) I just set up a pair of Raids (using Compass bushings, straddle cables and washers), and was wondering what the best arrangement was.

    • Straddle cable height matters really only for low-profile cantis. It’s not just the mechanical advantage, but also the effective arm length. With low-profile brakes, you get both a longer effective arm and better mechanical advantage with a lower straddle cable.

      With wide-profile cantis (like old Mafacs) and centerpulls, the lower mechanical advantage is countered by the longer effective arm length, so changing the straddle cable height makes little difference. So I set the front straddle cable to clear the rack, and the rear so that the hanger doesn’t obscure the taillight.

      • Frank B. says:

        I think, your explanation is a bit confusing. Mechanical advantage is proportional to the effective arm length, but MA also gets bigger with lowering the yoke. That applies to all brakes including low and wide profile cantilevers or centerpulls.
        So I would reformulate your explanation a bit:
        “However with wide-profile cantis (like old Mafacs) and centerpulls, the smaller mechanical advantage that you get from a higher yoke cable is countered by a higher mechanical advantage resulting from a longer effective arm length, so changing the straddle cable height makes little difference.”
        I must admit that I recently replaced a wide profile cantilever with a low profile brake! The WP cantilever had rather short arms and wasn’t very powerful. The LP canti has much longer arms and is more powerful, which is what I needed on this bicycle.

      • You are right – your explanation is technically correct. I should have said “mechanical advantage of the straddle cable” vs. “effective arm length”. The old explanation overlooked the “effective arm length” and thus recommended a lower straddle cable for all brakes, when it applies only to low-profile cantis.

        I realized the mistake when I lowered the straddle cable on my Urban Bike, hoping to increase the mechanical advantage of the wide-profile Mafac cantilevers. I was surprised to see no change, which is why I looked at the system more closely and realized that I had shortened the effective arm length.

      • Frank B. says:

        Benno Belhumeur did an exhaustive geometrical analysis of cantilever brake mechanics at http://www.circleacycles.com/cantilevers/
        He condenses the interplay between yoke height, cantilever angle etc. into a formula, that can be plotted to show the effect of various adjustments. I did a plot with gnuplot here for example:

        His in-depth analysis IMO surpasses the explanations by Sheldon Brown and those in the Bicycle Quarterly’s brake issue.

      • Benno’s analysis is great. Benno was inspired to look into this issue in details after the BQ brake issue showed how the previous analyses overlooked important components. It was fun discussing these issues with him.

  10. Tim Evans says:

    The last photo shows the brake pad not quite parallel to the rim sidewall (brake surface). Is that intentional, perhaps to mute squealing? Or, is it insignificant?

  11. wilfried531 says:

    The clearance of Compass brakes is clearly a great advantage compared to brakes made by other brand.
    Paul Racer are very stiff and powerfull but I confirme you have to deflate the tire completely and even pinch it to remove the wheel…
    Over time, It’s very annoying and regret a bit my choice.
    As you see on this picture of my bike, Paul Racer are very close to the fender without a lot of room to open :

    • Many brakes designed for wide tires have those issues. With many cantilever brakes, the pads are so long that they hit the fork blades/seatstays. With dual pivots, the quick releases don’t open the brakes wide enough, as they are the same design used on “racing” brakes intended for 23-25 mm tires. It seems that few makers really consider the need to remove the wheels.

  12. Michael says:

    I can get a 32 mm tire through my tech R559 brakes.
    For my next bike I would like a super light Boulder. I wonder if they will spec posts for the Compass brakes?

    • Boulder Bicycles has built quite a few bikes with Compass centerpull brakes (and also with Compass racks).

      • Michael says:

        Good! I’m looking forward to getting a 650 B Boulder skinny tube light weight lugged bike one day.
        I like the idea of having nothing between the tire and fork crown though like cantis allow. But sounds like braze on center pulls are more stable to use? Less chatter?
        Does the twisting of the fork blades from Cantis effect brake performance at all?

      • The twisting of the fork blades with cantis changes the toe-in of the pads. When you brake hard, the brakes suddenly can grab… Centerpulls have better modulation, since they attach near the fork crown, where the fork is strongest.

  13. Reed Kennedy says:

    Jan, thank you for the gorgeous pictures and detailed descriptions!
    What pump is mounted to the left seatstay of the blue bike in the last photograph? From the design it looks like a Lezyne, but I’m not familiar with any Lezyne pumps long enough to be carried there. I’ve long wished for one!

  14. Cynthia says:

    An article in BQ or in your blog about how to modify pumps with springs so they can fit frame pump pegs would be great, since more riders are customizing their bikes for “real world” riding.

  15. Rob says:

    A Zefal HPX of appropriate length will fit between two pump pegs on a seatstay if you drill a hole to locate the tip of the bottom pump peg off-centre to the lever side at the end of the pump. Just make sure that the inner edge of the hole is far enough out not to touch the o-ring seal. Remove all the clamp parts first, obviously!

  16. Rob says:

    Forgot to add that you must also file down a bit of the metal channel that holds the hard rubber strip (Zefal calls it “HPX Protection”). I think it looks neater if you shorten the rubber strip at an angle and retain it but it no longer has a practical function. All the current HPX production available here in the UK seems to be all black but you can remove the anodising quite easily to get a polished aluminium barrel.

Comments are closed.