Cycling to the Opera

Dharma Dawg Blawgs

Chasing the Ghost of Charlie Miller

By Luis Bernhardt

This is a man who thinks with his heart, 
His heart is not always wise. 
This is a man who stumbles and falls, 
But this is a man who tries. 

From “The King and I,” lyrics by Oscar Hammerstein II


I had thought it would be much easier. In 2011 I had ridden Paris-Brest-Paris (PBP) in under 65 hours on a fixed-gear bicycle. It was not a continuous ride; I had started with the 84-hour group at 5 am and had stopped both nights, well before midnight, to sleep before heading out refreshed the next morning before 5 am. The parcours had some hills, but I didn’t recall too many or too steep, except for the one at about km 1200, the climb in the Forest of Rambouillet, where the grade approached 15%, but not for long. So completing the ride in 2015 in under 56 hours and 40 minutes – Charlie Miller’s time in the 1901 PBP - seemed a not too difficult prospect. I should even be able to do this on a basis more consistent with the 1901 participants – riding a single-speed*, and – in my opinion, the real hallmark of a Charlie Miller ride – doing it completely unsupported, contrary to his well-equipped European competitors, just as Miller did.

Of course, the conditions were totally different. In 1901, the course was largely flat, along what is now France’s N-12 autoroute, but between the Bois du Boulogne in Paris and the Café du Commerce in Brest. At that time, this was an unpaved road used largely by horse-drawn carriages and the first motorcars. The 2015 route followed various secondary roads connecting tiny villages circuitously from west of Versailles to Brest and then back to the start. The 1901 route covered 1196 km; the 2015, 1230, even though it started about 30 km west of its 1901 location. And today’s route is hillier, much hillier. It was also unclear if the 1901 route was marked with arrows, unlikely, or if it merely entailed following the main highway signs to Brest and back to Paris.  

What the course was like in 1901 is largely conjecture. The other half of the challenge was Charlie Miller himself. Miller is little-known today, even among bicycle racing aficionados, who can rattle off the names of the great riders of the start of the 20th century – Zimmerman, Major Taylor, Bobby Walthour, et al. But who was Charlie Miller?

In August of 1893, an 18-year-old from eastern Germany, in the area of Thuringia or Saxony,[i] arrived in the United States. Klaus Mueller likely landed on Ellis Island, opened the previous year, where the immigration official must have anglicized his name, as he is known to us today as Charlie Miller**. He must also have been one of the thousands of adventurers to have made the journey from Europe to New York in search of opportunity. He continued westward to Chicago, and he must have been interested in the latest technological developments of that progressive age, as one of his first stops was to the Chicago World’s Fair, also known as the Columbian Exposition in honor of the 400th anniversary of Christopher Columbus’s arrival in the New World.

Amongst the many attractions at the fair – which included the first use of alternating current electricity to power the exhibits, the original Ferris Wheel, built by George Ferris himself, and a popular dancer known as Little Egypt, whose suggestive adaptation of a middle-eastern belly dance was known as the hootchy-kootchy – amongst those attractions, there was the increasingly popular sport of bicycle racing. The Columbian Exposition Bicycle Racing meet was being held on a track at 35th St. and Wentworth Avenue. This had to have captured young Charlie’s imagination, because he soon found a job in Chicago as a grocery clerk, saving $5 a month to put to his first bicycle.[ii]

He showed some potential at this, because soon “he was spotted by scouts from the old Morgan and Wright bicycle tire manufacturing firm while racing as a recreation in Garfield Park. They sponsored him for his first race in Louisville, KY, and he won in a breeze.” [iii]

Soon, Charlie Miller became arguably the best of the long-distance riders of his day. He won the last two single-rider six day bicycle races held at Madison Square Garden, 1897 and 1898. These were the demanding marathons where each rider rode the entire six days – all 142 hours between midnight Sunday when the starting pistol was fired through 10 pm on Saturday when the winner crossed the line, many riders having been unable to finish, or finishing as hallucinating wrecks. But Miller outlasted them all.

Near the end of the 1898 MSG Six-day, with Miller well in the lead, his handlers arranged for a publicity stunt to pump up the paying crowds. They set up a trackside wedding with his fiancé, 17-year-old Genevieve Hanson, who had been hanging out at the velodrome all week just to glimpse Charlie whizzing by lap after lap.

“I knew he would win. I stayed beside the race track day and night, until I was almost worn out with watching. Charlie always wears a button with our heads photographed on it. Well, I used to watch that button spin round and round the track until I was dizzy.”[iv] The promoters brought Miss Hanson’s mother in by express train from Chicago, and they interrupted the racing for a half hour on late afternoon Saturday to perform their wedding ceremony, charging the spectators twice the normal admission to witness the event! Beyond being a promotional scheme, the two remained steadfast; they celebrated their 50th wedding anniversary in Chicago in 1948 amongst many of the old-time bike racers.[v]

The following year, 1899, in response to growing public concern over the inhumanity of the single-rider sixes, the New York City ordinance was changed so that no competitor could ride more than 12 hours each day. The counter-response to this came from the innovative impresario at Madison Square Garden, William Brady. Bill Brady had produced over 250 Broadway plays, managed world heavyweight boxing champions “Gentleman Jim” Corbett and James J. Jeffries, and is credited with starting the acting careers of Douglas Fairbanks Sr., Humphrey Bogart, and Helen Hayes. He came up with the idea of turning the 6-day into a two-man team race, with one rider racing while the other rested, the two usually alternating two-hour shifts (unlike today’s sixes, where changes are continuous), and neither racing more than 12 hours each day. Thus was born the Madison, known in France as l’Americaine.[vi]

In 1899, the first year the two-man six-day was held, Miller won again, partnered with Frank “Dutch” Waller. The press referred to this team as “The Flying Dutchmen,” a popular moniker for Germanic riders of that time, though neither was Dutch. Waller was originally from Munich, and both lived in Chicago. The two had covered 2,733 miles + 4 laps on the tiny 10th-mile (160-meter) MSG track, a record that would stand until 1908.

In addition to winning the Houston individual six-day in 1898 and the San Francisco Six Day earlier in 1899 (the last of the “single-rider sixes” ever held), Miller made annual trips to France in April and May. Newspaper accounts refer to trips he made in 1898 and 99. He rode and finished the Roubaix 96-hour continuous race (did it start at the famous outdoor velodrome with the stone showers where the classic monument Paris-Roubaix still finishes today?), and he won the continuous 72-hour race held in Paris. "When Charlie Miller won a three day race in Paris, France, in 1898, he took only seven five-minute rest periods in the 72 hours."[vii]  So Miller would have been known to the French, and he would have been familiar with the racing around Paris. And he was used to riding for ten hours straight!

I should have realized at this point that Miller’s PBP, even on relatively flat roads, was not to be underestimated, given what he was capable of achieving. Here he was, winning 142-hour continuous track races, completing or even winning 96- and 72-hour road races, and by contrast my longest continuous effort to that point had been 1,000 km over a mere 45 hours.

Miller’s “short rest periods” were key to his success and equivalent to minimizing time spent at PBP’s thirteen controls. I used a planning spreadsheet that took anticipated average speed entered during each segment, along with projected time spent at each control, and was dismayed that even with just 10-minute stops (with one 30-minute stop in Brest), I would have to be able to maintain a 22-23 kmh average into the final three or four stages, even given faster 28 kmh early stages, if I were to have any hope of finishing in under 57 hours. This would be an issue if PBP turned out to be hillier than I had remembered. This was becoming a real challenge.

The initial PBP in 1891 was a race open to French riders only, and was ostensibly a test of equipment over a long distance. Charles Terront, a notable French professional riding for Michelin, won over a field of 207 riders, of which 99 completed the distance. It was sufficiently difficult to organize that the promoters – Pierre Giffard and the newspaper le Petite Journal – decided to run it only every ten years.

The second edition of PBP, the one Charlie Miller had entered, was sponsored again by le Petite Journal, but with the added resources of Henri Desgrange and his publication, L’Auto-Velo. Two years later, Desgrange would go on to found and promote his own race, the Tour de France. In the 1901 PBP, riders from outside France were admitted, and for the first time, so were amateur “touriste-routiers,” the harbingers of today’s randonneurs.

Charlie arrived in Paris after a forgettable couple of seasons. He had crashed in a 6-day in September of 1900 and had to be taken to hospital suffering a concussion and broken collarbone. In December he had ridden the 6-day at Madison Square Garden with Bobby Walthour Sr. and had retired early in the race.[viii] His 1901 results were sketchy, and PBP was likely the key to salvaging his season. It would have attracted an endurance rider used to racing in Paris, such as Charlie, despite the odds against him. The 10,000 franc first prize would also be encouragement.

And the deck was definitely stacked. Charlie had arrived on his own, apparently sponsored by Cleveland Bicycles* but without the team support typical of his European competitors. He had no trainers or mechanics, no follow car with spare bikes, no pacers to ride in front of him and shield him from the wind. Support between controls was not banned until the 1911 PBP, and the teams took maximum advantage of this. He was riding against the top Continental professionals of the day. Maurice Garin, “the little Chimney-sweep” (le Petit Ramoneur) was a two-time winner of Paris-Roubaix. The odds-on favorite, Lucien Lesna, had been the French and two-time European motorpaced (demi-fond) champion, as well as having won Paris-Roubaix earlier that year. Gaston Rivierre was the three-time winner of Bordeaux-Paris.

I had arrived in Paris on the Friday before the Sunday start. I had over 15,000 km in my legs to this point in the year, including a couple of continuous hilly 600 km rides, and had come off two weeks of easy tapering. I knew how to peak for road and track races, but ultramarathon cycling was still imbued with trial and error. It did not shake my confidence. Like Miller, my background was on the track, and my training had included speed work on the indoor 200-meter velodrome across the hill from my home in Burnaby, Canada. I was pleased to see that this year’s PBP would start and end next to the new indoor 250-meter Velodrome National in Saint-Quentin-en-Yvelines.

The 1901 PBP was supposed to start and finish at the Parc des Princes Velodrome in the southwest of Paris. Today, the big 48,000-seat soccer stadium is there, but in 1901, it was a very shallow, roundish, egg-shaped 666.6-meter velodrome (later changed to a 450-meter oval, banked track). However, rain on the Friday morning of August 16 caused the start to be moved to the posh Chalets du Cycle, “run as an establishment of good company, and located near the bridge to Suresnes.”[ix] Evidently, the Chalets were a “cyclists’ resort!”

There were 16 controls for PBP in 1901. The eight outbound controls were:

- Dreux – at the Hôtel de l'Epoque
- Mortagne-au-Perche - Café des Arts
- Pré-en-Pail - Café de la Mairie
- Laval - Café de l'Ouest
- Rennes - Café de l'Europe
- Saint-Brieuc - Café Tardivel
- Morlaix - Café de la Terrasse
- Brest - Café du Commerce[x]

These towns are still situated on the N-12 today. They were revisited on the return, with the finish at the Parc des Princes Velodrome.

Shortly before 5 am, the 25 professional riders set off. They charged over the bridge and into Suresnes, then over the hills west of the Seine to Dreux and beyond. Over a hundred tourist-routiers started 17 minutes later. Lucien Lesna and Hyppolyte Aucouturier were the first to Dreux (km 81), having averaged 34 kmh. Miller was in 8th when he entered Dreux only 15 minutes later.

My own start was at 5 o’clock in the afternoon. We set off in warm sunshine, the fifth and final wave of the 80-hour starters, buoyed by cheering crowds lining the barriers next to the velodrome. We spun quickly down the hill to the giant shopping mall in St. Quentin-en-Yvelines, a right turn and through the parking garage, and then along winding roads and continuous roundabouts until we settled into the lovely French countryside. I casually avoided two crashes near me, and still felt fresh as the sun disappeared ahead of us and we entered the first night.

Lesna was on a mission. By Saint-Brieuc, km 451, he had 10 minutes on Maurice Garin and 12 minutes on Aucouturier, who was prompted to jeer, “He must have forgotten that we have to turn at Brest and return to Paris!”[xi] By now, Miller was more than an hour behind Lesna, but in 7th place at this point.

Our route in 2015 was further to the south and over a jagged sawtooth profile of hill after hill. Although both Miller and I passed through Mortagne-au-Perche, our paths diverged at this point, with Miller headed east to Pre-en Pail and me headed a bit further south and climbing to Villaines-la-Juhel. Our paths then criss-crossed twice, me heading due west to Fougeres, Tinteniac, and Loudeac, Miler south to Laval and Rennes, then north to Saint-Brieuc, skirting the hills.

There was much more climbing on my route than I remembered from 2011, and on a continuous ride on a single-speed, the sustained tiredness was becoming a companion, especially on the long, 8% grade of the Roc Trevezel, especially after being softened up by the two sustained climbs that preceded it. I reached Brest on Monday at 6:30 pm, an hour and a half behind my best schedule, but still full of hope, knowing I could lose several hours and still reach my target. I was starting to have trouble keeping up with faster groups, and I was starting to see 80-hour riders coming back the other way, but my being on a single-speed gave me an excuse to remain optimistic.

The flying Lesna reached Brest at 3 am on Saturday, having covered the first 598 km in a shade over 22 hours, a remarkable average speed of close to 27 kmh. Garin and Aucouturier were two hours behind when Lesna started his return. But then things started to fall apart.

Despite the rain at the start, it had been a sunny and hot two days, and Lesna, riding with bare arms and legs, had developed bad sunburns, his skin becoming blistered and charred. At his return past Saint-Brieuc, his lead over Garin was down to 1h45. In Rennes, he was so uncomfortable that he stopped for a bath. Although he had left before Garin arrived, by Laval the headwind had picked up and his lead was down to just 7 minutes on his fast-charging chaser. Garin caught and passed Lesna in Mayenne as night was falling on Saturday, but Lesna immediately latched onto the faster wheel and the two rode together for the next 20 kilometers. But the effort of the two days and the sunburn were too much for Lesna. By Pre-en Pail, he lost 50 minutes in no time to Garin, who rode off alone into the darkness. Rivierre was still two hours behind, and Miller was even further back with mechanical issues, but Lesna had had enough. Before reaching the 1,000 km point in Alencon, he announced his retirement from the race to the controllers.[xii]

When I reached Fougeres after midday on Tuesday, I had 309 km left to ride and some recalculation to do. I would need to ride at close to 25 kmh and spend close to no time at the controls to finish inside Miller’s limit. My heart was sinking, as a 20 kmh average was more likely in my state of tiredness and with the remaining hills, including the wall of Rambouillet. My mouth was completely dry, and just the act of spitting would cause a wave of nausea and dry heaves. After 40 hours on the road, I was at risk of hallucinating, but riding in broad daylight was at least holding this off. Who knew what darkness would bring?

Rennes is southwest of Fougeres, and 346 km from the 1901 race finish. It was here, or near here that Miller’s bicycle broke.[xiii] There is no record of where or what part of it broke or how; we only know that when Miller finished, he was on a bicycle borrowed from his friend and motor-paced rival, the Welshman Tom Linton.[xiv] I can only surmise that Miller had something go wrong with his bike that made riding difficult but still possible. And fortunately, Linton had ridden to Rennes to check on the progress of the competitors, including his American friend Charlie. Or perhaps Linton had agreed to help Charlie in Rennes, the strategic place to have someone provide food and spares as it was the control closest to half-distance between Paris and Brest. And perhaps Charlie had been nursing a broken frame or component since the previous control, knowing that he could borrow Linton’s machine in Rennes?

While Miller had mechanical difficulties, and Lesna his sunburn to contend with, I was encountering very chilly nighttime temperatures as I neared Dreux. While only 75 km past Mortagne-au-Perche, it took me nearly four hours to get there. It was dark, and the dampness that hung over the farm roads made riding bitterly cold. I was in two jerseys, arm warmers, knee warmers, rain jacket, and reflective vest, and I was still feeling the biting chill. In 2011, arriving in Dreux had been straightforward. In 2015 though, there was an endless succession of straight roads across dark fields leading to more straight roads over more dark fields, over and over. I was beginning to think we were going in circles. The darkness also brought on early stages of disorientation. I began to wonder just why I was here, what the rules were for this ride. Didn’t I need to get a card signed at each control? Why didn’t we just drive to the next control?

By the time my small group reached Dreux, it was 11:20 pm. I had until 1:40 am to make the finish, 2h20 to go 65 km. I’d have to maintain an average speed of 27 kmh! I stoically conceded; inevitably, time ran out somewhere between Dreux and St. Quentin. Further misfortune arose when I lost the route within 10 km of the finish, easy to do with the arrows posted higher than the reach of my headlight beam. It took me an hour, in my confused and hallucinating state in the darkness, to finally get back on course and somehow manage to finish inside 60 hours.

Miller arrived at the Parc des Princes Velodrome about four and a half hours after the winner, Maurice Garin, who had finished in a record time of 52 hours 11 minutes. Garin had been followed two hours later by Rivierre , then Aucouturier and Michel Frederick finished together about 45 minutes after Rivierre. Because this was a huge 666-meter track, the riders would have had one and a half to two laps to ride after they entered the velodrome, and they would be timed over the final paved lap and a half, the last kilometer. On entering the velodrome, Frederick had sprinted ahead of Aucouturier, but he mistook the bell for the end of the ride and eased up. Aucouturier flew past him and held the lead to the finish a lap later, with Frederick only 10 meters behind.[xv] Miller finished fifth, an hour and 50 minutes behind these two. Showing off his track speed, Miller also set the fastest time for the last kilometer, a scintillating 1:27.8, after nearly 1200 km and on the borrowed bike! The first touriste-routier, A. Rosiere, finished in 9th place with a time of 62h26.

My failure to qualify for the Charlie Miller Society was not a discouragement. On the contrary, it clearly demonstrated to me just what a superhuman effort Miller must have made to ride the 1196 km unsupported, and on a single-speed bicycle, and to ride fast enough to remain within a reasonable distance of the leaders, given the roads, food, and knowledge of training and nutrition in existence over 100 years ago. My admiration for Miller and his achievement only grew as a result. I plan to try again in 2019; I’ll be 68 going on 69, but I think I will use gears to lessen the effect of the continuous hills; Henri Desgrange even said I could*. But again, like Charlie Miller, and as a respectful tribute to his achievement, I will ride unsupported.

Miller quit bicycle racing “for a rest” in 1901. Two years later, he bought a car, a Thomas Flyer, and went into the taxi business. This became his career. He continued to drive, but for the larger taxi companies, until his retirement in 1945. Then he and Genevieve moved to Franklin Park, selling their old Chicago home on North La Salle St., and there they lived until his death in 1951 at age 76.[xvi]


* Although Miller used a fixed gear on the track, he had always been interested in the latest technology. Like many riders, he became interested in motorpaced racing (“demi-fond,” or middle-distance), behind motorcycles and multi-rider tandems, and in 1899, he and fellow rider Tom Linton had brought over three motors from France to stir up North American interest in this discipline. He had also collaborated on the development of a motorized tandem. The freewheel had been invented in 1899/1900, and I’m sure he would have seen its advantages on the open road and had access to one.

** “Klaus,” of course, as in “Santa Claus,” is short for “Niklaus,” or Nicholas, so he should have been Nick Miller!

* Cleveland was a brand of H.A. Lozier and Co., one of the biggest bicycles companies of the 1890’s. It had been bought out by and was owned by Albert Pope at that time.

* Desgrange, of course, famously stated, “I still feel that variable gears are only for people over 45,” in an interview with L’Equipe. “Isn’t it better to triumph by the strength of your muscles than by the artifice of a derailleur?”



[i] says Miller was born in Hildebrandshausen (Thuringen), while the Chicago Tribune, Aug. 15, 1943 says that he came from Saxony.

[ii] Chicago Tribune, Aug. 15, 1943


[iii] Chicago Tribune, Dec 26, 1948.


[iv] San Francisco Call, Volume 85, Number 88, 26 February 1899.

[v] Chicago Tribune, Dec 26, 1948.


[vi] Peter Joffre Nye (with Jeff Groman and Mark Tyson), The Six-Day Bicycle Races, Van der Plas Pulications, 2006


[vii] Alec D. Kennedy, quoted in the Chicago Tribune, Aug. 15, 1943




[ix] Jacques Seray, Paris-Brest-Paris 120 Ans, 1,200 Kilometres, 2011, Jacques Seray


[x] Audax Club Parisien,




[xii] Seray, 2011


[xiii] Bill Bryant , A SHORT HISTORY OF PARIS-BREST-PARIS ©, 1999; or so the legend goes. But there doesn’t appear to be any documented evidence that Miller suffered any punctures, and there is only the implication that his bicycle broke because he finished on Tom Linton’s bicycle.


[xiv] New York Times, Aug. 19, 1901


[xv] New York Times, Aug. 19, 1901


[xvi] Chicago Tribune, Dec 26, 1948

Experimenting with a Front Disc Brake, Part 2

Last year, I had installed a mechanical SRAM/Avid BB-7 front disc brake on my fixed gear road bike. I used the existing Campag carbon fiber brake lever, added compressionless cable housing, and bought a new Wound Up carbon fiber fork designed for disc brake use. The fork was semi-custom-built, as I required a 367mm axle-crown distance to match the ENVE road fork I had been using during the summer.  There were many other carbon fiber forks designed to handle disc brakes, but with one exception they were designed for cyclocross, with 380-400mm axle-crown dimensions. This would have thrown off the bike’s trail measurement, affecting its handling. The one exception was the ENVE carbon disc brake fork, with 367mm axle-crown, but it only came with a tapered (1 1/8” to 1 1/2”) steerer tube. I also investigated having a lightweight steel fork built, strong enough to handle a disc brake, with the standard Post mount, but the weight and cost of such a fork was pretty much a wash with the Wound Up carbon fork, both in the $500/650 gram range.


Figure 1: Avid BB-7 mechanical disc brake

The result was very positive. I went thru only two sets of disc brake pads last winter, the original Avid pads and a set of Brake Authority replacement pads (available at Mountain Equipment Co-op in Canada for just over $21). Each pad lasted between 3,400 and 3,900 km, and I think there’s still some life left in the Brake Authority pads. Both sets of pads wore smoothly, and because I was almost exclusively on the front brake (rear brake used only on snowy/icy days), the rear caliper pads lasted all winter. When I switched back to the ENVE summer fork with Campag caliper brake in March, the front wheel was still absolutely pristine. No wear at all, of course, on the Argentine-made aluminum rim. The sealed SRAM hub still spun smoothly, and I only had to replace one front spoke (fortunately one that laced to the outside on the side opposite the rotor, so I didn’t have to remove the rotor to replace the spoke).

Braking performance was very consistent regardless of wet or dry weather; the only complaint was brake squeal on rainy days, but this seemed to disappear with greater use and water being removed from the rotor. I was so pleased with the performance of the disc brake that I built a second front wheel, based on the Shimano 525 front hub with 6-bolt rotor mounting (rather than their proprietary threaded center-lock method) laced to a cheap Chinese-made carbon fiber rim. I was curious to see just how compatible hubs from different manufacturers were with respect to the position of the disc caliper – would it need readjustment? I also wanted to build a winter wheel with standard cup and cone bearings, as I think cartridge bearings are a technology misapplied to bicycle hubs, which are subject to thrust, as well as radial, loads. Bikes that are rocked side to side on climbs are more subject to thrust loads on the hubs, so it’s less of an issue for me now, as my climbing now is more of a forward-back motion, but back in the 70’s when I climbed Burnaby Mountain every day to get to school, using lots of side-to-side bike rocking to drive the cranks, I could go thru a set of Phil Wood hubs in a month of winter commutes.  I also resolved to upgrade to a hydraulic front disc brake this winter in order to confirm (or deny) the accolades given to such brakes over their mechanical counterparts.

Last year, SRAM was busy fixing their hydraulic road brakes, which had a problem with their seals at low temperatures and had to be recalled. Shimano had also introduced its hydraulic road disc brakes (R785, RS785). Both, however, were intimately connected to their brake lever shifting systems. In order to get the brake, you needed to get the lever and then remove the shifting crap if you rode a fixie. (Note: you can now obtain the RS785 caliper separately, and for a very reasonable price.) At least SRAM provided a right/front option. Most manufacturers assume that your left lever goes to the front brake. This is an idiotic convention that permeates US bike culture, and like “lawyer tabs” on the forks assumes that bikes are ridden by idiots.

“Lawyer tabs” assume that you are incapable of properly adjusting the front wheel quick release so that the wheel doesn’t fall out while you’re riding. In over 42 years of riding, I have never had this happen, and I am always gentle with closing the front QR. The first thing I do whenever I buy a new bike or fork is to file off the dumb tabs, new UCI regulations be damned. I have not done so with the disc brake fork, hearing wild tales of front wheels being ejected by rigorous application of the front disc brake (which is reportedly in the exact position to maximize wheel ejection). From my experience so far, covering over 10,000 km on the disc brake fork, I am inclined now to go ahead and file off the tabs, obviating the need to spend the extra time to unscrew the “quick” release to get the wheel out. So far, my QR has come nowhere near the tabs after long days of riding and heavy brake application using the front brake only.

Left-front braking assumes that you are right-handed, and that in an emergency (often downhill) braking situation, you will slam on both brakes hard. Because your dominant hand is usually stronger, the story goes, if the front brake lever is under the right hand, you will presumably lock the front wheel and be catapulted over the bars. This has never happened to me. In fact, if you are right-handed, you want the right lever to go to the front brake because the front brake does most of the braking, and you need to ease up on the front brake if you feel the rear tire skidding (easing the rear brake does absolutely nothing and in fact will lengthen the stopping distance, so you want to maintain rear brake pressure, even as the rear tire skids). You want the hand with the best motor control to be over the front brake!

I knew that TRP made a hydraulic brake, the HY/RD, but this was hydraulic only at the caliper. Between the lever and the caliper is a mechanical cable. This cable is subject to dirt and resulting cable binding, to which a hydraulic line is immune. The primary reason for using a hydraulic disc brake is the mechanical advantage you get between the master cylinder in the lever and the pistons at the caliper. At Jan Heine’s October Bicycle Quarterly “Un-Meeting” in Packwood, WA, I was standing at the evening campfire chatting with some VERY knowledgeable cyclists, and one of them mentioned the TRP Hylex, a hydraulic brake designed for road bikes that don’t need brake lever shifting. Rumor has it that this brake was going to become an aftermarket hydraulic disc alternative for Shimano Di2 systems, but Shimano’s own hydraulic Di2 quashed this.

The large, sculpted levers of the TRP Hylex are more similar to Shimano than to Campagnolo, so there could be aesthetic issues when paired with the older Campagnolo (10sp) lever on the same handlebar, but that didn’t deter me. It’s possible to purchase a set of similar TRP mechanical levers (even a set in carbon fiber) if matching levers are desired. The first issue was that the brake came as a complete lever/caliper assembly connected by the pre-bled hose, but since I wanted right/front braking, I had to order a rear brake to get the righthand lever. Fortunately, front and rear brakes use the same caliper, and I already had the Post/ISO front fork adaptor from my previous disc brake, so the connection to the fork would not be a problem. However, I would need to shorten the hose to the front cable length (approximately 80 cm), which would require bleeding the system, something road cyclists have never had to do (until now).

I purchased the TRP bleed kit, then used a sharp knife to cut the hose, removed the end from the caliper, installed the correct hose into the caliper using the parts supplied in the bleed kit, then hooked up the bleed kit to the brake assembly. The TRP bleed kit had only one threaded fitting to fit the fluid input and output hoses to the caliper and lever, so instead of waiting for TRP to send the second threaded fitting (which they did for free), I used a technique called “burping” to top up the fluid level in the system, while making sure there was no air in the hoses. Fortunately, TRP’s system uses mineral oil as the brake fluid. This is far safer to handle than the caustic DOT brake fluid used in some hydraulic disc brakes, which is a factor because you do spill a lot of brake fluid when you’re doing your first bleed.

With the system closed up and installed on the bike, use of the new hydraulic disc brakes was pretty seamless. It didn’t feel awkward at all to be using two different lever shapes. Indeed, with a front disc, all the braking is securely performed with just the front brake (unless it’s snowing or icy).

On my daily ride to work, the first descent is a 300-meter 12% grade where I’ll typically nudge 50 kmh before braking for the slow turn onto the bike path. I used this section to wear in the new pads, which felt similar to the previous BB-7 on initial use, with sufficient power to slow for the turn at the bottom. Even on rainy days, this was easy to control with just the front brake.  On the flats, I soon became used to one-finger braking from the top of the brake lever hood. The TRP lever curves outward, inviting the middle finger to apply all the pressure needed. I had read that Avid BB-7’s lacked braking power, but I only noticed this as the speed approached zero. The Hylex brakes did not feel significantly more powerful than the previous mechanical disc brake, but it could be that the pads took a while to condition the previously-used rotor.

On very wet days, the brakes squealed loudly unless I started braking early enough to clear the rotor of water (or warm up the rotor). But continued braking caused the squealing to stop. If I didn’t brake for a while, the squealing would start again, but at a much softer level.


Figure 2: TRP Hylex hydraulic disc brake

The original pads that came with the Hylex lasted less than 2,500 km. When I pulled the pads, one of them was completely worn down, with scraped bare aluminum showing on about a third of the pad. The brake did emit a scraping sound and was a little less effective on descents, but I thought this was maybe road grit on the pads.


Figure 3: Worn pads - OEM TRP (top); Shimano M-05 (for M515, bottom)

Replacing pads on a hydraulic disc brake is a little different than a mechanical. This is due to the self-adjusting feature of hydraulic brakes. As the pads wear, the pistons reposition themselves closer to the rotor so that the amount of lever movement required for brake actuation stays pretty constant. This is why you always insert a tiny block between the pads when you remove the wheel for brake bleeding. If you pull the lever with nothing between the pads, the pistons will reposition so that the pads stay together. Which is the same situation when you install new, thicker pads. So before the new pads are installed, you have to take a plastic tire lever and force both pistons back into their wells. Once the pads are replaced, they must again be burned into the rotor.

I replaced the original OEM TRP pads with a set of Shimano M05 resin pads (for use with their older mechanical M515). TRP has standardized on this pad, one that more knowledgeable cyclists regard as an archaic design, across most of their range of disc brakes. The Shimano resin pads provided adequate braking in dry weather, but wore at an alarming rate in the rain. The pads were down to bare aluminum after only 418 km. This is less than a good set of rim pads that cost less. The Shimano pads were replaced with a set of Brake Authority “burly” sintered metal pads, and they are currently in good shape with over 1,500 km, although I have switched the pads around, as the inside pad seemed to be wearing faster than the outside, and the pads appear to be identical.

When I replaced the Shimano pads, I also switched to the wheel with the carbon fiber rim, and with the rotor from the TRP brake. The TRP rotor is heavily scalloped and lightened. After wear-in, I still detect little difference in amount of power required vs the mechanical disc brake for hard downhill stops. All of my braking has been from the top of the lever, using two fingers at most, with no rear brake input. I am happy with the self-adjusting feature of the hydraulic brake, though.

I have been making slight adjustments with the caliper. I started off bolting the caliper directly to the Post/ISO adaptor used for the BB-7, and noted the rotor was wearing a bit far to the inside, so I added the concave washers from the BB-7 to move the calipers more to the outside. I’m not sure how or if the adaptor designed for the Hylex would differ from the adaptor from the BB-7.

The next step will be upgrading the Hylex caliper to the more highly-regarded Shimano RS785 caliper. There doesn’t appear to be anything on the internet that relates to compatibility between the Hylex lever and the Shimano road caliper, or even if the TRP hose will connect to the RS785. More on this later.

So far, the main conclusion I can draw from my disc brake experiments has been that you don’t need to buy a new bike to enjoy disc braking on the road. This is because you really only need a front disc brake. Unless you are racing cyclo-cross, a rim brake is perfectly adequate for the rear, and is much lighter than a disc brake. All existing bikes can benefit from a front disc by merely swapping to a disc brake fork and buying the appropriate front brake setup. These usually come pre-bled, with lever matched to caliper and shifting system. It would make sense for aftermarket fork vendors to make disc brake forks available in 367mm axle-crown spacing for road discs, as well as in the current 380-400mm cyclo-cross forks.

Another observation is that equivalent pads seem to wear out faster with a hydraulic disc brake. Whether this is due to pad/piston design, or to the additional power of the hydraulic pistons is so far undetermined.



Equipment for the “Heroic Bike” – Cino Heroica 2014

The classical stuff wasn’t all it's cracked up to be

By Luis Bernhardt

Currently trending are “retro-rides,” where you show up on a classical/vintage bicycle, dressed like a classic/vintage rider, to ride over conditions prevalent in the heroic dawn of cycling. L’Eroica in Tuscany (over the fabled Strada Bianca) is the best example. Closer to home, there was the Cino Heroica in Kalispell, Montana this past September (, where you ride from Kila to Hot Spring over 90 km of mostly gravel roads, then return over 80 km of gravel and one long climb the next day. Halfway thru each ride, there is an Italian potluck lunch (“pranza”) where you can drink beer and wine and eat salami and cheese and whatever other robust delicacies participants may bring. And on the Saturday evening, everyone dresses up like Fausto Coppi in Italian designer suits and all sit down to an al fresco gourmet meal.

Aside from all the pretty bicycles, these retro-rides have a downside. They underscore everything we didn’t realize we lacked as cyclists back in the 1970’s when we first seriously got on bikes. Like clipless pedals. I put old school pedals with toe clips and straps on my steel-framed fixed gear Rodriguez in order to comply with the “heroic bicycle” requirements specified for the Cino. I forgot just how much I hated those old style pedals, and after completing the two-day ride, I couldn’t get the pedals off fast enough. Combined with the old-style cleats on my modern cycling shoes (the cleats had screws that fit into two of the three Look-pattern bolt holes in the sole of the shoes), they had to be tightened with the strap, else they’d pull out at the worst times in the middle of slippery gravel climbs. And when you wanted to pull out, you’d have to remember to pull back, not sideways. And even then, the modern Velcro shoe straps would get caught by the clip or strap, delaying the disengagement slightly. I can see why clipless pedals would be such a hit amongst off-road riders; it’s a safety issue!



What constitutes a heroic bike? The Cino Heroica organizers require your bike to comply with any three of the following features:

-        Built in 1987 or earlier

-        Steel frame

-        Non-indexed shifting

-        Downtube shifters

-        Toe clips and straps

-        Fixed gear

-        Single speed

-        Tubular tires

My 2009 Rodriguez had the steel frame and the fixed gear, so I could only add the toes clips and straps. I have tubulars, but they’re for track use and I didn’t want to destroy them on the gravel roads. The bike also had a carbon fiber fork, brake levers, and handlebars, an aero carbon fiber seatpost, compressionless cable housing, a crank-based powermeter, and a Garmin as well as a standard bike computer, but they got thru inspection with no problem.



Wool jerseys on a hot day were another irritation. I recall riding with a small group of Berkeley Wheelmen very early in my cycling apprenticeship. We were in the East Bay (across from San Francisco) on an early-season training ride, it was a chilly day, and we were bundled in our wool jerseys and long tights. One of the juniors remarked that he “couldn’t wait until it got warm enough to wear the silk track jerseys!” Wool jerseys might be comfortable on cool days, but they’re scratchy and uncomfortable on a hot, dry climb! I wore a full-zip lycra jersey on the second day (after spending the first in an old short-sleeve wool jersey with a short front zipper) and was MUCH more comfortable, thank you!

The domestic bicycle clothing industry back in the 70’s was nothing to write home about, either. Yes, you could get wonderful Italian tube-knit wool jerseys from Sergal, and comfortable wool shorts with thick chamois from Vittore Gianni, but you could also get badly-styled and/or ill-fitting crap from Kucharik and other US clothiers who knew zip about cycling. Kucharik made the ugliest leather hairnets in that era (lots of padding, but looking like you had sausages on your head). One guy at the Cino admitted to wearing his Kucharik shorts from the 70’s on the ride. They looked pretty bad; they fit him as loosely as baggy shorts, with serious sagging at the crotch.

I wore modern lycra shorts with synthetic chamois and was perfectly comfortable and stylish. I remember those natural chamois of the 70’s. They were great until you washed them. After drying, the chamois would turn to rough cardboard, and you had to knead and tug them smooth, applying liberal amounts of chamois cream. Today, you just chuck the synthetic shorts into the washer, hang them to dry, and put them on without even touching the synthetic chamois.

Ardent cyclists will give rapturous lip service to how beautiful the bikes were prior to 1987 (the generally-accepted cut-off date for heroic bikes), but aside from top-level components by Campagnolo, most of the equipment used on bikes of that period were poorly-finished crap. “Ergonomics” had not yet been invented, so the most egregious examples of industrial design could be found in the brake levers of the period. Although some of the levers were covered with rubber hoods (Campag, Weinmann, Universal), most had no rubber hoods, or might have a strip of rubber across the top (Mafac). Manufacturers didn’t seem to acknowledge that cyclists rode mostly on the tops of the brake hoods, not in the drops, resulting in the unwelcoming shapes of the levers.



As the North American bicycle market of the 70’s became more prominent, bigger and heavier North Americans started stressing European-made bicycle parts that were still built to early 20th century standards, or were designed for smaller and lighter Europeans. I remember breaking two or three French pedal spindles made by Lyotard because I couldn’t afford the more robust pedals made by Campagnolo. I also broke those plastic Simplex derailleurs, of course. And tubular rims (on 36-spoke wheels) would dent at the mere sight of potholes.

When I first started riding in 1972, most racers rode on tubulars. If you used clinchers, you had to buy 27x 1 ¼” wheels, and the lightest clincher you could get was a wired-on Michelin 50. Then you had to re-adjust your brakes when you switched wheels from training to racing (or have two bikes). The tubulars were expensive (relatively about the same as now), repairs were sketchy (you actually had to sew them back together – with an awl if you were clever) – the thread would snap and the thin rubber tube inside would explode, unless the cotton sidewalls rotted first from rainy winter riding.

Then in the mid-to-late 70’s, as I recall, Specialized (then called SBI – Specialized Bicycle Imports) came out with thin 700C clinchers in 700x25 or 28, and you could find Euro 700C rims and build up a wheel that was brake-height-compatible with your tubular wheelsets. But brake spacing was often an issue with the pre-1980’s frames. Many of the British frames were built for 27” wheels, so if you wanted to race on them, you’d have to install a “drop-bolt” for the rear brake just to get the pads to hit the brake track on the smaller 700C and tubular rims.

Because the gaps in the frame were so big, brake calipers had to be longer-reach. Today, we’re reverting to seeing the larger clearances as a good thing, since this allows you to use wider tires, but this also means the brakes have less mechanical advantage. The top bikes used Campagnolo calipers, but they were single-pivot and nowhere near as powerful as the short-reach dual-pivot calipers available now for close-clearance frames.

When I set up my wife’s restored De Rosa, I installed an old Galli sidepull caliper on the front and an old single-pivot Campag Record caliper on the rear. These were connected to the standard Campag Record levers of the day, which allowed for either internal or external cable routing. I should have set it up for external routing, as the internal routing requires the cable to pass over a small plate inside the lever that redirects the cable to the internal housing. I think this adds a bit of friction to the system, making the levers a tiny bit harder to pull. I know from personal experience that the plate causes the cable to break at the bend that it creates after one or two seasons of use. But the major problem is that most levers of the time were made for medium-to-large men with medium-to-large hands, not for short women. She had to sag it after the first protracted gravel descent when her hands gave out.

Meanwhile, I was on the steel Rodriguez fixie with modern ergonomic Campag EC-RE600 Record levers (in carbon fiber) attached to dual pivot Campag calipers (Chorus front and Super Record rear). And my hands were still getting tired over the long, rough gravel descents!

My only inconveniences from the old days were those pedals with toe clips. I also had the fixed gear, but that was the point of the ride, to do it on a fixie. If I had a derailleur, I’m sure I’d be complaining about the downtube shifting (although I did install the paragon of downtube shifter design – the Simplex retrofriction – on Carole’s bike).

I think that we don’t realize just how good we’ve got it now. And just what crap most of those old classic and vintage bikes really were!

Crashing the Charly Miller Society

* After the first day of racing the August 17, 1901 New York Times reported, "The betting to-night is 3 to 1 against Lesna and 6 to 1 against Miller"

I find the legend of Charly Miller to be quite compelling. In the second Paris-Brest-Paris ever held, back in 1901, Charly entered the professional racers category. At that time, PBP was an actual race for the pro’s as well as for a second group of “touriste-routiers,” the randonneurs of those days. The race was held over a slightly shorter course, just under 1200 km, and on what was then the “main highway” between Paris and Brest, following today’s N12, which in those days was largely dirt or cobblestone and more suitable for wagons and carriages pulled by horses, although not as hilly as the current route.

For the long distance professionals of those days, it was quite common to be “paced.” So from the start at Parc-des-Prince, the rider would have a teammate ride ahead of him, over the bridge at Suresnes, and out to Versailles. At the top of the long climb, the first pace rider would pull off, and a second rider would take up pacing duties to the next relief point.

In addition, at each of the controles, the rider’s support crew would hand up food and replace bidons, much like the supported riders in today’s PBP. The crew would hop back in their car and drive to the next controle, and problems with the bicycles would be fixed by the mechanics on the crew.


Charly Miller had none of this. He was a track endurance rider from Chicago, who reportedly didn’t even speak French. God knows why he rode PBP in 1901, but he rode it completely unsupported (and reportedly underfunded). Despite not having pacers and not having support on the road, and despite suffering a number of punctures and having to replace his broken bicycle (he rode the final 350 km on a borrowed bike), he still managed to finish 5th, and against 112 of the top European professionals. To give you some idea, this particular PBP was won by Maurice Garin, who went on to win the inaugural Tour de France two years later. Miller’s finishing time of 56 hours 40 minutes has been enshrined by Randonneurs USA: any RUSA member (I’m not sure if they need to be US citizens as well, but it’s OK, I’m a dual citizen) who finishes inside this time becomes a member of the “Charly Miller Society.”

I find RUSA too lax in opening its membership to the Society to anyone finishing inside 56:40. To me, it denigrates the most heroic aspect of Miller’s fine accomplishment: he did it completely unsupported. He had to stop and find his own food. He had to fill his own bottles. He had to carry all his own clothing and spares, and perform all his own repairs. In those days, assuming Miller was riding pneumatic tires, a punctured tire could take an hour to repair or replace.

I would have to think that Miller was quite the resourceful bike racer, so he must have had the presence of mind to sit on the wheel of supported, paced riders. Garin’s finishing time of 52 hours shows that reasonable speeds (where drafting is possible) were possible on the dirt roads of the time.

A less influential factor was Miller’s gearing. In those days, pro’s would have been on a single gear, probably with a freewheel (invented for bicycles around 1899), maybe even fixed, as Miller was a trackie. I would say less influential, since there was far less climbing than in today’s more rural PBP.

For these reasons, I am announcing my goal of doing a “real” Charly Miller in 2015: Inside 56:40, on a fixed gear bike, completely unsupported.

I finished the 2011 PBP in just under 65 hours on a fixed gear, but this was with the 84-hour group, starting at 5 am, riding until dark or later (I rode until about 11:30 pm the first day), getting several hours sleep plus sit-down dinner and breakfast, and departing between 4:30 and 5 the next morning.  I had no drop bags and carried everything (pretty minimal – it all fit in my jersey pockets or I wore it), but I spent too much time at the controles trying to find food and supplies. I never got sleepy on the bike, though!

To set a good time, I think you need to start with the 80-hour group in the late afternoon. This ensures that you’re fresh enough to ride the entire first night, and then all of the daylight hours the next day take care of themselves. The first stop for sleep would be at one of the controles on the way back from Brest, trying to complete 7- or 800 km as darkness sets in for the second night. Here, sleep would need to be limited to just one hour before the final 16 to 20-hour rush to the finish at SQY. That’s the basic strategy.

Water bottles can be refilled at the controles, but for this ride I think it’s a good idea to avoid obtaining food at the controles. The lineups are too long, or the dining areas are too far from the bikes or too difficult to find. I think it will be much faster just stopping at “convenience stores” along the route. I was also amazed at how much they charged at the controles for the “energy drinks” that were advertised during the ride! I’m sure I can do better at the epiceries.

There is a downside; in fact, there are a few. The main downside is that I will need to ride about 30 hours before the first sleep break. On a previous 1,000 km ride (geared road bike), I was able to nap about 30 minutes at the 30-hour mark, then finish in 45 hours. But on the Hoodoo 500 (fixed gear), I took a sleep break at about 21 hours, then needed another sleep break at about 25 hours, and then a couple more later. On both these rides, I started hallucinating. On the 1,000, it was after 47 hours, while on the drive home from the finish. At Hoodoo, the hallucinations started at night after about 40 hours. The hallucinations usually involve seeing an overpass above me, like riding on the lower deck of a bridge. The hallucinations are extremely realistic. After the Hoodoo ride, large grease stains in the parking lot turned into discarded umbrellas. I could even reach down with my hand to pick them up, and my fingers would just go through them, but the hallucinations remained!

In conjunction with the hallucinations, there is usually a reduced “will” to continue. On the 1,000, I was ready to quit at the 30-hour mark in Leavenworth, as I still had to get up Stevens Pass, out to Monroe and Snohomish, then up Highway 9 to the finish at Abbotsford. But I wouldn’t be able to get a ride out until the next day, after the controle closed, so I decided that it would be quicker just to ride the bike back, so I finished the ride. And it was quicker than waiting for the car to take me.  At Hoodoo, I spent too much time off the bike after the 21st hour. Having the support car right there was not a good thing; stops were too tempting. As well, once you’re tired, you get disoriented, and that combined with the hallucinations makes you question whether or not you are even on the right route. This is where a very strong and sure crew chief becomes invaluable. Failing this, a working Garmin is the next best thing. For PBP, I’ll have the external battery pack attached to the Garmin. The nice thing about the Garmin 500 is that it lights up when the external power is turned on, so I can read it at night while it’s charging.

So, what do you get when you qualify for Le Societe Charly Miller? A medal? A certificate? No, you get nothing. Zip. Which is fine with me, because the highest accomplishments usually go unrewarded. Prizes are embarrassing; it’s enough to just have done it. And it’s not about me, it’s about Charly Miller, an American about whom hardly any Americans know anything. He is even less well-known than Major Taylor, who would be in Paris the following year during his European tour, racing the best European track cyclists (he had become a World Sprint Champion in Montreal in 1899).  But the legend of Charly Miller is an inspiration worth knowing.

Experimenting with a Front Disc Brake: Results

Disc brakes are pretty standard now on mtb’s, but for the road they seem to be a solution looking for a problem. They provide much better modulation, but mechanical disks don’t provide as much maximum power, and they are heavier than caliper brakes when you factor in that heavy steel rotor. They are a good excuse for using carbon fiber rims, but the major advantage is in saving the front rim in wet and dirty winter riding.

I have been experimenting with a front disc brake this past winter, and I am impressed with its capabilities. I would recommend a front disc for winter riding on the road, but would also conclude that a rear disc is unnecessary and superfluous, unless you are set on using carbon fiber wheels front and rear.


The benefits, after 4,000 km of winter riding in the Pacific NW (and some riding in California):

-       The front rim is absolutely pristine. In previous years, the front rim would be worn down enough to form longitudinal cracks, or even fail when the tire bead would bend the rim from 100+ psi of tire pressure, allowing the tube to expand and explode. This could be a safety issue. Typically, a rim would last 15-20,000 km before this would happen, but lately I have noticed current rims have started to “bow” outward after only 7,000 km. This suggests that the rim manufacturers are using less material.

-       My first set of disc pads lasted about 3,500 km. This compares quite favorably to the 500 km rim pads will normally last in a wet winter before needing replacement. It helps to use sintered metal pads rather than organic pads for durability.

-       Even with a mechanical disc brake (Avid BB7), modulation has been good enough to be able to use front brake only over snow. More on this later.

-       Because I’ve used primarily front brake only (especially on dry surfaces), the rear rim pads have lasted much longer than the disc pads.

-       For travel, or for long-distance riding, a set of spare disc pads is far easier to carry than a set of rim pads. The disc pads can also be replaced without tools (at least on a BB7).

For the past several years, I’ve been using a Rodriguez fixed gear bike almost exclusively on the road. The typical summer configuration uses an ENVE carbon fiber fork with Campag Chorus caliper brakes front and rear. In the winter, I typically change out the handlebar/stem and attached brake assembly to Campag Centaur and vintage Record. This past winter, I wanted to experiment with a front disc in order to address the front rim wear problem.

The first issue was obtaining a proper disc brake fork. Most authorities are adamant that a fork must be built to specifically handle the loads of a disc brake. They show pictures of broken forks that are bent just above the disc mount. Another issue is that the force of a disc brake tends to move the wheel downward in the dropout. One hears stories of disc brakes causing wheels to eject from forks. Finally (after years) this justifies the use of “lawyer tabs,” which I have always filed off on all my forks so that I don’t have to mess with the quick release to get the front wheel on or off.

The second issue was finding a fork built to handle a disc that had a 367mm axle-crown measurement. This is pretty standard for a road fork, but most aftermarket forks are built for touring or cyclocross, with axle-crown distances of 380 to 400mm. You would think that product managers would make 367mm disc forks readily available for riders wanting to upgrade their road bikes for front disc/rear caliper use. Especially when used with a hydraulic disc, this setup would allay fears of losing brakes entirely if the hydraulic fluid overheats and causes the brake to fade on a long, steep descent; at least you’d still have the rear caliper brake. But no, the only production 367mm disc fork is made by ENVE for a tapered (1 1/8” to 1.5”) steerer. Most disc forks are built for cyclocross use and provide plenty of clearance, but they also raise the front of the bike slightly (and mess with the resulting rake/trail).

Fortunately, I was able to order a Wound Up carbon disc brake fork, modified with a 367mm axle-crown distance, and with a 45mm rake. The Wound Up disc fork appears to be their standard fork with the attachment of a standard disk brake sleeve (which says a lot about the strength of their standard fork. Wound Up, by the way, is about the only company to make a tandem-certified carbon fiber fork). The fork is ugly as sin, the aluminum crown is still massive and provides just enough clearance to a 23mm tire, and the carbon steerer is reinforced with an aluminum liner so that a standard starfangled nut can be used, but the result is a 650 gram fork (compared to the 350-gram ENVE 2.0. At least, like ENVE, the Wound Up forks are built in Utah, and are not from some unnamed factory in China!

The other preparation required was building a new front wheel. I used the SRAM 506 front disc hub as the basis. Since I like to have the outward spokes radiating backwards (with the rotation), I had to build the wheel asymmetrically. The theory is that you want the outward spokes radiating against the forces being exerted next to the spoke flange (“wind-up”), so the drive side of the rear wheel will have the spokes radiating back. Because the rotor is on the left side, the outward spokes must radiate towards the front (against the rotation); thus, the asymmetrical front wheel, which must also be dished. 32 spokes are sufficient; I’m not sure disc hubs with less than 32 holes are commonly available. I typically build my front wheels (for caliper brakes) with 28 spokes (I weigh 170 pounds/78 kilos).

Because there is so little stress on the front rim, I have resolved to build a second front disc wheel, but it will have a carbon fiber clincher rim!

I was able to obtain compressionless cable housing (highly recommended), and installation of the disc was quick and easy. After mounting the rotor to the hub with the six torx bolts, you install the front wheel in the fork, then LOOSELY attach the disc caliper unit, tighten down the pads against the rotor, and tighten the caliper unit to the fork. This ensures that the pads are properly aligned to the rotor. You then attach the brake cable and adjust the pads so the wheel spins freely, but lever travel is minimal. I didn't even bother to attach the cable adjuster – its sole purpose is to take out any cable stretching. The instructions are adamant about not using it to adjust the lever travel (other than taking out the initial unproductive part). The lever is a standard Campag carbon fiber with the gearshift internals removed. Lever travel is equivalent to a well-adjusted rim caliper.

You then have to seat/wear in the disc pads. This is done by taking the bike down a steep hill and riding the brake for about 10 seconds. My first ride with the disc brake took me down Thermal in Coquitlam in the rain, which requires far more than 10 seconds of braking. I rode only  the front brake most of the way down with no problem. At the last stop sign, I put my gloved hand on the rotor, which immediately started smoking! The rotor does get hot! Later in December, when I took the bike to California and rode up and down Mt. Diablo, the rotor had changed to a darker color by the bottom of the descent!

Since I installed the disc brake, with very few exceptions all of my braking has been front brake only. This was initially to test the efficiency of the brake, but I was so confident with the result that I just continued to use front brake only. Yes, you do need to revert to rear brake only on icy roads, but on soft snow, I've even had the confidence to apply just the front brake due to the disc’s superior modulation. I think the greater modulation is due to the mechanical disadvantage of the disk vs a rim brake. On a standard caliper rim brake, the calipers are located at pretty much the same part of the circle where resistance (ground force on tire) is applied, so the forces tend to cancel each other out, and it is easy to apply too much force to the brake, which causes traction to be lost. With a disc, though, the braking force is applied closer to the center of the wheel circle, and the resistance of the ground is multiplied by the greater distance (moment arm) of the hub to rim distance vs the hub to rotor distance. But because braking force is reduced due to this difference in leverage, it is easier to make small changes to the braking force without significantly disrupting the tire traction. The loss of leverage makes the brakes less “twitchy,” enhancing the ability to modulate braking.

I think the advantage of a hydraulic brake has more to do with the ability to apply greater forces at maximum braking, since there is some leverage provided by the hydraulic reservoir. I think this would provide more uniform modulation across the entire lever travel, including at the maximum extreme, where the mechanical disc’s inferior leverage requires more hand pulling force. But we will find out when I upgrade to a hydraulic disc for next winter.

And if the front disc is good, wouldn't adding a rear disc be better? For the road, I think not, and for the following reasons:

-       The weight penalty of a disc brake and rotor vs brake caliper.

-       The rotor requires dishing. Combined with the dishing inherent in a rear wheel with cassette, this would result in a very narrow base for the rear wheel’s spokes, making for a weaker rear wheel.

-       If the front disc is hydraulic, a mechanical rear caliper would be a better backup option than a second hydraulic disc.

-       There is far less of a safety issue with the rear rim cracking due to pads wearing the brake track than with the front rim failing.

-       The rear brake does not require the same amount of modulation as the front. Basically, it just needs to be on or off, since the front brake does most of the work. In the worst-case emergency braking situation (assuming no ice), both brakes are applied at maximum. The resulting weight shift unweights the rear tire, causing it to start skidding. Rather than modulating the rear  brake, which will have absolutely no effect since the rear tire is nearly weightless, the competent cyclist will maintain the same rear brake pressure, but will ease up slightly on the FRONT brake. Although this will reduce deceleration slightly, it will also reduce weighting to the front, providing equivalent weight to the rear, and giving the rear tire more traction, causing the skidding to stop as the rear tire weights, and rear tire adhesion helps slow the bike. Pressure is then reapplied to the front, which is modulated as needed according to the above to control the rear tire skidding.

-       In my own particular case, with a fixed gear in a track end, a rear disk is impractical because it would be difficult to move the rear disc brake caliper in conjunction with the rear wheel as it is gradually moved further back due to the chain stretching.  As well, one of the big advantages of my particular fixie is that it can be quickly converted for track use by removing the brakes and handlebar/stem and replacing them with a track handlebar/stem. With a front disk, I must also include a fork change, but a useless rear disc brake caliper assembly mounted to the seat stay would make this harder to achieve!