A Better Way to Adjust Toe-In


Compass centerpull brakes now come with special washers to adjust toe-in. These washers are easy to retrofit on older Compass and even classic Mafac brakes. With this system, you to adjust toe-in only once, when you initially set up the brakes, and then never worry about it again.


What is toe-in? It means setting up the brake pads so that the fronts are closer to the rim than the rears. This is important for cantilever brakes, because the fork blades twist when you brake, and the brake pads rotate in relation to the rim. With the right amount of toe-in, the pads will be parallel to the rim when you brake hard.

How much toe-in do you need? There are no firm values because it depends on how much your fork blades twist. It’s trial-and-error. Too much toe-in just makes your brakes work less effectively. Toe-out is worse: Your brakes don’t work well and they howl and squeal. When in doubt, go for a little toe-in rather than too little.


Centerpull brakes attach close to the fork crown, so the fork blades don’t twist significantly. This is why centerpull brakes offer such consistent brake modulation. It’s also the reason they don’t require toe-in. Centerpull brakes work best when the pads are parallel to the rim at all times.

When we introduced the Compass centerpull brakes, the toe-in was not adjustable. In our testing of prototypes, we found that the brakes might squeal for the first few rides, but they became quiet as the pads wore until they were parallel to the rim. Most of our customers have had similar experiences. Riding in the rain helps, because it wears your pads faster.

However, we’ve found since that for a few riders, our brakes squealed, and continued to squeal longer than was acceptable. If the brake pivots on the frame and fork are brazed on at a slight angle, it’s possible that your pads initially have significant toe-out. Rather than wait for the pads to wear away, you’d want to compensate for the misalignment of the post by adjusting the toe-in.

We experimented with numerous ways to adjust toe-in, from the Shimano system of the late 1980s with its wedge-shaped washers to the modern spherical washers. We found all of them hard to adjust, hard to keep adjusted, and they all make changing brake pads a big pain. Basically, you have to hold the adjustment of the brake pad in three directions while you tighten the bolt:

  • angle of pad to rim (seen from the front)
  • alignment of pad to rim (see from the side)
  • toe-in (seen from above)

As you tighten the mounting bolt, all these alignments tend to move. You never get it right, and after a while, you just give up and accept whatever you have as “good enough”. Every Bicycle Quarterly test bike with cantilevers, even those set up by the best pro shops, has had inconsistent brake pad alignment.

What is silly about this process is that, at least on centerpull brakes, the toe-in should be set only once. It compensates for slight misalignment of the brake pivots, and those will remain the same for the life of the bike. Why would you design a system that requires riders to adjust the toe-in again and again, every time they adjust the brake pads? Wouldn’t it be nice if you could adjust the toe-in once, and then it have it remain the same forever? Adjusting and replacing brake pads would be easy!


So we adopted a system that does just that. It’s really simple: The washer that goes under the posts of the brake pad has one groove that is deeper than the other. This puts the brake pad at a 2.5° angle, which gives you about 2 mm of toe-in. There is a dot on the reverse (invisible once installed) that shows which groove is deeper. Install the washer in the orientation that you want your toe-in, and you are done. The washer remains in place as you adjust or change the brake pads, so you never have to think about it again. Your toe-in always remains the same.

And setting the adjustment is easy – you just substitute the new washer and then set up the brake as usual. Since the toe-in is pre-selected, you’ll get it right every time. With our post-style pads, tightening the mounting nut only tends to rotate the pads in one direction (alignment of pad to rim), which is easy to compensate. The other movements aren’t affected as you tighten the bolt, which makes it easy to adjust the brake pads. While we offer a special tool to hold the brake pad during this adjustment, it’s not really needed. Usually, I don’t bother with it.

Worried about adjusting your toe-in in smaller increments than 2.5°? As the pads wear, they’ll take care of any small differences. Realistically, if you get within a degree or two of your target value with modern spherical pads, you are doing very well.


The new washers are included with all Compass centerpull brakes (in addition to the washers without toe-in for bikes with perfectly placed pivots). They also are available as a retrofit for Compass and classic Mafac brakes.

Unfortunately, this method for adjusting toe-in won’t work on most modern brakes, which use “bolt-on” rather than “post-style” pads. For those, somebody should design a system where you adjust the toe-in once and for all, rather than having to fiddle with it every time you adjust the brake pads. (Some cartridge brake pads allow you to replace the pads without disturbing the holders. Unfortunately, this doesn’t help when you have to adjust the pads as they wear, so they still hit the rim in the right place.)

One more piece of news about our brakes: They now are sold individually, to give you a maximum of flexibility when spec’ing your bike.

Click here to learn more about our brakes or to order your set of washers.

Posted in Brakes | 26 Comments

The “Friend”, an Affordable Touring Bike


During a recent cyclotouring trip in Japan, we stopped at an onsen hot bath. As we locked up our bikes, I noticed an interesting touring bike, chained to a lamppost.

“A few decades ago, a bike like this was every boy’s dream,” my friends remarked. Looking over the bike, I can understand why. It’s a smartly designed, yet affordable, campeur in the French tradition. As a teenager, I would have dreamt of a bike like this, too.


The frame looks nicely made, with Nervex “Professional” lugs. The seatstays cross over the seat tube. This is intended to stiffen the frame at the seat tube, useful when you carry heavy loads on the rear rack. The seat tube is less likely to act like a hinge that has the bike flex in undesirable ways. The lack of logos is a marked contrast to the mountain bike next to it.


The seat tube sticker reads “Friend – since 1933 – hand made”. The pump peg on the seatstay is a nice touch. It frees up the frame’s main tubes to carry bottle cages. And it keeps the top tube clear, so you can shoulder the bike and carry it. A true adventure probably includes portaging the bike…


The front rack is made from tubular steel. Its large platform allows carrying a big handlebar bag, or even a small bundle of firewood. The clamps are for a flashlight – useful to illuminate the road when riding after dark, and easy to remove to use while setting up camp. In addition, there is a generator-powered headlight on the other side.


The fork and the rack stay both have attachments for low-riders. The kink in the rack stay is an interesting design feature: When using a low-rider, it spreads the attachment points further apart, and yet the rack remains triangulated.


There are low-rider attachments on the seatstays as well. Mafac cantilever brakes provide plenty of stopping power. Wheels are 650A (590 mm, rather than 650B/584 mm).


The bike is prepared for Rinko, with a split rear fender. The generator that attaches to the bottom of the seatstay is missing.


The SunTour Vx derailleur was one of the best-shifting derailleurs of its time. Together with the triple cranks, it provided a wide spread of gears. There is even a rubber strap to protect the chainstay from getting scratched by chain slap on bumpy roads. Vertical dropouts make it easy to remove the rear wheel even with fenders – important for Rinko.


Another nice touch is the seatpost: a Japanese SR copy of my all-time favorite, the classic Simplex seatpost. The Brooks Professional saddle looks like a later upgrade to me. It shows that somebody has loved this bike.


The headbadge finally gives away the maker of this well thought-out machine: Leopard. A quick Internet search turned up a bike maker with that name, but it was founded in 2004, not 1933 like the seat tube sticker of this bike proclaims. So it must be another company, unrelated to the maker of this bike.

I didn’t meet the owner of the bike. I would have liked to know more about its history, and especially find out how it rides. It’s probably great. The relatively steep head angle and a good amount of fork offset result in a low-trail geometry perfect for precise handling with a load, but also handles well unloaded. Even in this very small size, it doesn’t seem to have toe overlap. Whoever designed this bike knew what they were doing.


In fact, the entire bike would fit right in with the machines we ride today, after our decade-long “Journey of Discovery” has taught us the merits of fully integrated bikes, aluminum fenders, low-trail geometries and wider tires on smaller wheels. To think that in Japan, as a teenager, I could have bought a bike like this off the shelf…

It’s really a shame that bikes like these no longer are available, except as expensive custom bikes. Cyclotouring is resurgent, and this bike once again could be the machine of many dreams. Today, it’s not just boys who dream of taking to the road by bike, but girls (and women), too.

Just imagine if you could buy a bike like this at an affordable price: A nicely made, lugged steel frame. Low-trail geometry. Wide tires and good components. Even comfortable handlebars. (The shape is remarkably similar to the Compass Maes Parallels.) Aluminum fenders, lights, and a front rack as standard equipment. Front and rear low-riders as optional extras that you can add later, when you are heading out on a tour. And most of all, the “Friend” looks purposeful, yet elegant and beautiful.


Touring by bike is a dream, and bikes like these made the dream more attainable. Before and after the grand tour, the “Friend” would be fun to ride around town and on weekend jaunts.

Who will be the first to make a bike like this again? I have no doubt that it will be successful.

Further reading:

Posted in Testing and Tech | 43 Comments

Bicycle Quarterly Un-Meeting Next Weekend!


The Bicycle Quarterly Un-Meeting is next weekend. Join us for two days of riding, camaraderie and exploration! We’ll meet at the General Store in Carson, WA, on Saturday, June 25, 2016, at 9 a.m. There is only one store in Carson, so it’s easy to find.

Carson General Store

From there, we’ll ride up Panther Creek Road to Trout Lake. Faster riders can explore some side trips – I’ve been intrigued by the Guler Ice Caves for some time. We’ll have lunch at the diner in Trout Lake that is famous for its huckleberry shakes. Then we’ll return to Carson. The round trip is about 70 miles (115 km), with about 1/3 on gravel.

Unfortunately, Takhlakh Lake (top photo) is a bit far, but the scenery in that part of the Cascades is spectacular anywhere. Usually, we get great views of Mt. Adams on our rides in these parts.


I hope to have time for a bath at the famous Hotel St. Martin with its volcanic hot springs. Dinner is at the brew pub in Carson. Most of us probably will stay at the Panther Creek National Forest campground, where we’ll have a campfire that night. There are hotels for those who prefer a real bed.


On Sunday, we’ll return to Portland via the Bridge of the Gods and the Old Columbia Highway (above). It’s going to be a fun weekend, and I look forward to seeing many of you there!

About the Bicycle Quarterly Un-Meeting:

  • Everybody is welcome.
  • There is no registration, no fees, no services.
  • We simply publish a time and place, and anybody who wants to participate shows up and joins us.
  • Please be prepared to ride your bike self-supported. (There is no sag wagon!)
  • No need to have a special bike – we have everything from commuters to racing bikes participate.
  • It’s all about having fun and making friends.


Posted in Rides | 10 Comments

Suspension Losses Confirmed


Recently, Bicycle Quarterly’s experiments on suspension losses have been replicated and confirmed: Higher tire pressures don’t result in faster speeds – even on smooth pavement. Replicating results is a crucial part of science, which makes the new results an important milestone in the understanding of bicycle performance. No longer is it just Bicycle Quarterly talking about suspension losses and lower tire pressures – the science is becoming widely accepted.

When Bicycle Quarterly’s tire tests (below) showed that higher pressure didn’t make your tires faster, few people believed it. Back in 2007, everybody “knew” that pumping up your tires harder made them faster.

We had doubts, too. So we tested again and again, and our results always were the same. We concluded that it was true, even if it went against the accepted wisdom of almost 100 years of cycling knowledge.


Looking through the literature and talking to experts like Jim Papadopoulos, we found a mechanism that could explain this: suspension losses caused by vibrations. As the tissues in the rider’s body rub against each other, friction turns energy into heat. And that energy must come from somewhere: It is taken from the forward momentum of the bike. Your body vibrates, and that slows down the bike. (The bike also vibrates, but it’s not as significant, since it’s mostly made from hard materials that don’t generate much friction.)

The next step was to prove that these vibrations could cost significant power. We went to rumble strips on the shoulder of a highway (photo at the top), because they allowed side-by-side comparison between smooth pavement and a “standardized” rough surface. The results were surprising: Riding on the rough surface took up to 290 Watts more than riding on the smooth surface (below).


Where did those 290 Watts go? After testing various pieces of equipment on the rumble strips all day, I knew where the energy went: My body was sore all over. I had experienced suspension losses on my own body!

Careful testing is only a first step. Real science demands that all scientific experiments are repeatable and replicable.

Repeatable means that if you run the same experiment twice, you must get the same result. We did that multiple times: Each configuration was run at least three times. And we ran the same equipment at the beginning, in the middle and at the end of the test, to make sure that conditions (wind, temperature, etc.) did not change and affect the results.

Replicable means that others must be able to do the same experiment, and get the same results. We published our methodology for testing suspension losses in Bicycle Quarterly. That was back in 2009, and we’ve been waiting for others to replicate them. We are excited that now Joshua Poertner has done similar test, also using rumble strips. And his results are similar to ours:

rumble strip test web

The blue line at the bottom shows the old-style steel drum tests: Higher pressure makes your tires faster. But that is true only if you don’t have a rider on board. (No rider = few suspension losses)

Once you put a rider on the bike, things start to look very different: The green line shows brand-new asphalt, the yellow line coarse intermediate asphalt, and the red line are the rumble strips. You can see that resistance increases beyond a certain pressure. This is the opposite of the old wisdom, which is expressed by the blue line.

It’s important to remember that the green, yellow and red lines are real-world testing. The blue line is done in the laboratory. And when laboratory tests don’t match the real world, then they are useless.


The article doesn’t mention Joshua Poertner’s methodology. I am a bit surprised that the dropoff in performance at higher pressures is so large. Our own testing (above) – on very smooth pavement – showed that very high pressures actually resulted in the same performance as lower pressures – not worse performance, as Joshua Poertner’s data seem to indicate. In the future, we’ll have to figure out which is correct. Or perhaps it’s a simple matter of Joshua Poertner’s “smooth” asphalt being rougher than ours…

However, everybody now agrees that higher pressures do not make you faster. We also agree that when things get rough, higher pressures are actually slower.

For riders, what matters most is how you can make your bike faster. And Joshua Poertner’s advice mirrors what we’ve been saying for years: “It turns out that it’s much better to be 10 or even 20 psi lower than the ideal tire pressure than 10 psi higher.” And: “Here’s the next thing you have to think about. As tire width increases, tire pressure decreases. So a wider tire performs better in terms of rolling performance.”


Looking into the future, Poertner said: “I remember when wheels went from 19 mm to 23 mm. It was a very gradual process. And then we went from 23 mm to 25 mm. Now we’re seeing 28 mm wheels. Where does it stop? I don’t know.”

And we all agree that wider tires are faster because they can run at lower pressures over a mix of surfaces. Joshua Poertner is comparing identical tires at different widths. It is understood that to offer good performance, the wider tires must be supple, otherwise, you lose too much energy to flexing the tire casing at it deforms with each wheel revolution.

In other words: On most roads, and especially on rough ones, a 32 mm Compass tire will be faster than a 26 mm Compass tire. But a 42 mm Schwalbe Marathon will be slower than both, even though it’s wider – because it’s so stiff that its casing absorbs way more energy.

Here is what it means in practical terms:

  • Run the widest tire that fits your frame, at least within reason. Bicycle Quarterly’s tests have shown that 32 mm tires roll as fast as 25 mm even on very smooth asphalt, and faster than 23 mm or 20 mm. On rough roads, the wider tires are clearly faster. Since we measured this at 22 mph (35 km/h) with a rider, this takes into account the wind resistance at typical “spirited” cycling speeds.
  • Run your tires at a relatively low pressure that still offers good handling. You don’t want your tires collapse under hard cornering, but beyond that, there is no benefit to adding more air. Experiment with different pressures, but don’t be afraid to let out some air.
  • Select the most supple tire for the best performance.

It’s taken almost a decade, but it’s nice that our results finally have been replicated and confirmed. What once was controversial is becoming universally accepted. And as Joshua Poertner points out (“Where does it stop?”), there is more research to be done. Fortunately, Bicycle Quarterly is already working on this!

Further reading:

Posted in Testing and Tech | 80 Comments

Another Road for the Collection


A few years ago, I wrote in Bicycle Quarterly how I collect roads. Others collect bikes, or rare components, or photos of racers. My collection is more esoteric: I collect roads. Not the physical roads, but the experience of riding on them. Like all collectors, I have criteria of which roads are worthy for inclusion in my collection, and which are merely conveyances that transport me where I want to go.

During the recent Golden Week cyclotouring trip, I was excited to add another road to my collection: Road 327 near Matsumoto in Japan. I explored this incredible road during a solo early-morning ride. And I don’t regret getting up at 5 a.m. to ride it.


The previous night, I had seen a valley in the distance, leading up to snow-covered peaks. If a road went up there, it could be spectacular. It seemed worth exploring…

A quick lock on Google Maps showed a promising road up the valley. The road dead-ends at an onsen hot springs high up in the mountains. “Dead-end” means little traffic… “High in the mountains” promises great climbs and descents.


The next morning, I headed out before breakfast. And the road fully lived up to my expectations.


The climb started with a set of amazing hairpin turns. The map and photo above show that first part of the road. The road wasn’t so steep that I was struggling, and it was fun to push myself on this stretch. I gained elevation quickly. Looking back, I could see several levels of the road below me.



The next section was more open as it ran along a steep cliff.  This part offered great views of the valley below and of the mountains ahead.


I rode past waterfalls and across little bridges, even traversed a short tunnel.


The entire time, I saw no more than three or four cars, as well as a very short tour bus. Its driver clearly had driven this road many times, as he took very confident lines around the tight turns.


After an hour of climbing, I reached a small pass and realized that it was time to return, if I wanted to eat breakfast. Later, checking the maps, I realized that I had climbed 800 m (2500 ft) in just 14 km (8 miles). The maps also show that the road continues for another 3 km, with another 250 m of elevation gain. I’ll have to come back!


The descent was even more fun than the climb. I quickly gained speed on the wide-open stretches. My brakes got a workout as I approached the tight hairpin turns at high speed. As I leaned the bike hard, I could feel my wide, supple tires bite into the pavement. It’s nice to get feedback that there is grip in reserve.

The best part of the downhill was the section along the cliffside. With hardly any braking, I threw the bike right, left, right again. Japanese mountain roads have useful mirrors that allow you to see around the corners: It’s important to make sure that there is no oncoming traffic when you take the best line on a single-lane road!

In twisty sections like that, I appreciate a bike that handles with precision and corners without reluctance. Descents like these are the reason we spent so much time studying front-end geometry, think endlessly about bike handling, and optimize tire construction. On this empty road, nobody could see the smile on my face, but it was huge.

And I made it back for breakfast (almost) on time.

Further reading:

Posted in Rides | 8 Comments

Look-Compatible MKS US-L Pedals

Compass bicycles_5982 copy

Looking for LOOK-compatible pedals for your travel bike? Here is the solution. The MKS US-L pedal uses a similar retention system as the LOOK Keo. In fact, the cleats are interchangeable. Unlike any other LOOK-compatible pedal, the MKS US-L is available with the EZ-Superior Rinko system.

Like all MKS pedals we sell, the US-L is a top-of-the-line pedal with ultra-smooth cartridge bearings. You really have to turn the spindle with your fingers to believe how smooth these pedals spin. And they still spin smoothly after years of use…


That system works similar to an air hose coupling: The part on the left remains screwed into the crank. Turn and push the outer ring, and you can remove the pedal. No tools required, and it only takes a few seconds.

The photo above shows the US-B Nuevo pedal, which is compatible with the Time Atac retention system. The Rinko system is the same on all MKS pedals in the Compass program.This has the added benefit of making it easy to swap pedals: for example, if you want to ride the same bike with platform pedals during commutes, but with clipless pedals during spirited weekend rides.

Compass bicycles_5970 copy

All MKS Rinko pedals come with a neat bag to carry your pedals while traveling.

Compass bicycles_5999 copy

The MKS US-L pedals also are available in a standard, non-Rinko version. The release tension is adjustable with a 3 mm Allen wrench (included) in three steps. This makes it easy to match the tension between right and left pedal.

The retention system is split, so that only one side has to open to release the shoe. This means that the release is relatively easy – you only have to overcome half the spring tension that holds the shoe when pedaling. No longer do you have to choose between safety during all-out pedaling efforts, and safety when stopping!

Click here for more information about MKS pedals.

Posted in Pedals | 22 Comments

Golden Week Cycling


Golden Week is one of the biggest vacation times in Japan. It’s a combination of one-day holidays that result in a little over a week of time off. And it’s springtime, so virtually every cyclist takes to the road. This year, we went on a ride in the Japanese Alps with a group of friends.


Spring in Japan is a great time for cycling. It’s warm, but not yet hot. The skies are blue, and the fresh green of the forests looks especially vivid in the bright sunlight. The rice fields are being flooded. It’s the Japan you imagine in children’s picture books.


The best roads of Japan go through the mountains, and this pass was especially spectacular. The cliffs were so steep and loose that the road was built into the mountain, with avalanche galleries protecting it from falling rocks (and snow in the winter). At the top, we exited a tunnel to see a spectacular view of the Japanese Alps (top photo).


This area really deserves the name “Japanese Alps”, as the steep mountains and broad valleys look remarkably similar to Switzerland. So do the small fields, and even the ski slopes.


We cycled on tiny roads past bucolic lakes.


The roads rarely were flat, which made the cycling more interesting.


The pace was unhurried, with plenty of time for exploring…


… visiting local shrines…


…and even a farm where wasabi (Japanese horseradish) is grown in the shade of a little valley.


We avoided large roads as much as possible, preferring little byways and even dirt paths.


Our ride ended in Matsumoto with its magnificent castle.


After some more sightseeing, we Rinko’d our bikes and returned to Tokyo. Thank you to our friends for organizing this great trip!

Photo credit: Natsuko Hirose (top photo).

Posted in Rides, Rinko | 10 Comments