Bikes are everywhere. Bike parts aren’t.

Ever have one of those days where you’re occupied for hours getting nothing done? That’s most days in bike repair.

Diagnosing a bike repair requires multiple steps. What system is malfunctioning? Can it be adjusted, or is something outright broken? If something is broken, can it be fixed? If not, can it be replaced exactly? If not, can some other part fit? Is the exact part or the substitute part actually available? How long will it take to get here? What will it cost? Can I fake it with salvaged parts or widgets in the various boxes and bins we've accumulated?

Multiply the process by the number of broken parts on the bike. Add one more repetition for every additional part that turns up while you’re working on what you already identified.

A large shop, perhaps part of a chain of shops, in a heavily populated market area might manage to have a phenomenally well-stocked parts department. If a rider never leaves such an area, or only does so briefly -- and is ridiculously lucky -- they might not run into a problem with parts availability. But lots of riders live in smaller population centers or travel outside of the zone that Big Bicycle considers worthy of their attention. There are no parts stores, like NAPA, O'Reilly, VIP, for bicycles.

Mountain bikes often show up encased in dried mud. Their riders tend to delay maintenance and repair until the bike is completely unrideable. I had one this week that looked like it had been buried in a salt marsh for a couple of years. These modern marvels of trail mastery have lots more moving parts than their ancestors did in the 1990s, mostly so that their riders can propel them with less caution at higher speeds under the influence of gravity.

Road riders don't tend to bash their bikes as hard and frequently. Their bikes show up with overuse injuries because they don't take hard hits that show dramatic symptoms instantly. Shift cables quietly fray under housings and bar tape. Chains wear. A broken shift cable can jam an entire shifter. Worn chains wear rear cogs too badly to accept a new chain. Gravel bikes borrow from both road and mountain categories.

Even casual recreational bikes can be disasters. People bring in a bike they bought in 1998 and say that it's only 15 years old, and that we just worked on it recently. A check of our extensive records might show that "recently" was three years ago, and the problem it had then was completely unrelated to the one it has now. Bikes are taken for granted until they fail too completely to ignore. Sort of like cars, only without the built-in weather protection of body work.

Bicycles have always challenged mechanics with different dimensions and standards applied to overall mechanisms that operated the same way. Into the beginning of the 1980s, these were mostly nationalistic variations in thread pitch and some tubing diameters. From the late 1980s onward, these differences were mostly corporate-driven, related to indexed shifting systems. These affected whole drive trains, as companies messed with cog spacing to match proprietary click shifters.

Initially, the click systems used modified progressive levers. The lever would stop in a different, distinct position for each gear. This meant that the rider usually had the option to switch to friction shifting if the synchronization went out. So the companies had to mess with cog spacing to make the stops adapt only to their patented parts. By the time SRAM beat Shimano in an unfair trade practices lawsuit, Shimano's unfair trade practices had already given it market dominance, so cog spacing became more or less standardized on their pattern. The other format was Campagnolo's, but Campy has always been a luxury brand.

Index-only shifters make perfect adjustment and synchronization essential. A bike that was high end when new from the late 1990s through today might have eight, nine, ten, eleven, or twelve (sometimes 13) cogs on the rear hub. Drive trains have to match all the way through by brand on the more recent bikes, as SRAM and Shimano have kept the Shifter Wars raging. As with every war, the civilian population suffers much more than the actual combatants. Do you have three, two, or one chainring? By extension, front derailleur or no front derailleur? Well into the 21st Century, the sheer number of speeds was a selling point. Three in the front and nine in the rear makes 27. Three in the front and ten in the rear makes 30. But as chains got skinnier and shifting systems had to handle more chain angle, the industry singled out the front derailleur as the source of all evil. Your high end bike now will have only 12 speeds and damn proud of it. In other words, we're back to the same gear range we had in 1980 with two in the front and six in the back, only it all costs at least three times as much and is far more failure prone. Progress!

Riders mostly don't pay attention to any of this. They buy a new bike and treat it they way they have always treated a bike, expecting the same longevity and reliability. A younger rider who has only ever known finicky index-only shifting will have worse "good old days" to look back on compared to an old geezer who remembers friction shifting and well crafted simplicity, but they both can share the realization that things have gotten steadily more costly and fall apart sooner.

We haven't even talked about suspension, disc brakes, or tubeless tires yet. A guy came in with a sheared off alloy spoke nipple on his mountain bike wheel. With a tube-type tire, it's a quick and simple fix. With a tubeless tire, its a time-consuming, messy, costly process that involves completely redoing the rim tape. An air-tight seal is absolutely essential to tubeless tires. You can't maintain that if you peel back a section of rim tape to drop in a replacement spoke nipple. The guy bought a handful of brass nipples and went off to try his own luck with it. He is free to try cutting a hole and patching it afterward, and then dealing with the almost inevitable failure of that patch, leading him eventually to redo the tape completely. That requires completely cleaning and drying the rim before meticulously applying your tape of choice and remounting the old tire or replacing it because you discover that the sidewalls are too broken down to reseal. Even applying a patch won't work unless the work area is perfectly clean and dry. 

An inner tube will press a rim tape repair into place, while also not depending on it. A tubeless tire does not have that advantage. Air pressure alone will not press the patch more firmly where you want it. Air pressure alone will work its way into any area of weakness and turn it into a leak.

Every customer who comes in interrupts the flow of work already in progress -- or the treasure hunt for needed parts so that work can continue. It's not their fault, it's just how things go. Their questions need answers. Their bike or bike part needs preliminary diagnosis. It's important to share information and knowledge, but in the meantime the bike on the stand is just sitting there. And the new work probably triggers more treasure hunting for some part we either hadn't bothered to stock or just ran out of.

Sheet metal cogs and tinfoil chains

An article on Ars Technica extolling the "improvements" in bike technology since the 1980s has been popping up on social media today to give my ulcer a workout. Observations by technophiles who aren't mechanics help fuel the headlong rush of the bike industry into fragile technology that is a pain in the ass to maintain, and often impossible to fix.

People love gizmos. Bikes started out on the forefront of mechanical innovation in the 19th Century. As they were superseded by the automobile, they languished in neglect, evolving only a little until influxes of cash and public interest in the 1970s and '80s inspired a rising curve of development that really foamed up in the 1990s with the mountain bike boom and the expansion of the aero/tri segment.

The 1990s brought engineers off the sidelines as hobbyist users and shoved them deep into the design process. Much of this was fueled by the demand for mountain bike suspension systems that would work with the heavy and inefficient human engine, but once you get a serious case of engineering it spreads far and fast. A public conditioned to crave expensive new technological things and accept that they are junk within a year or two at most was ripe for such changes in the bike industry.

For the first 100 years or so, bicycles were evolved machines, not designed machines. Even now, the designers have to study what has been working and figure out why it did before they can screw with it. A hundred years of trial and error honed the vehicle in its various forms to meet a variety of needs. A couple of decades of technological promiscuity have led to some genuine improvements and a lot of expensive and unhelpful complications, not to say downright handicaps. People who think in absolute terms will laud the vertical dropout and its offspring, the through-axle. People who, on the other hand, understand the value of a variable rear wheel position will be stockpiling old steel frames with long horizontal dropouts. They're not just for fixies. 

I would go through the article and dismantle it point by point, but if I read it for more than a few seconds my head explodes, so I'm not going to do that. I have to accept that the battle for public perception was lost a long time ago. The best I can do is put better advice out there for the few people  who will still appreciate the versatility and freedom that a simpler bike offers to the average underpaid toiler. As it was in the 1880s, is now and ever shall be, a bike is a good investment for a working stiff, as long as it is well chosen.

A Trek Fuel full suspension mountain bike last week provided a nifty example of how a designed weak link, intended to fail under stress to preserve more important and expensive parts of a system, can be bypassed when those more expensive parts become more fragile in the greedy quest for more "features." The rear derailleur hanger of old steel and aluminum frames used to be part of the dropout on that side. It would occasionally get bent on a road bike, in a crash, or if an improperly adjusted derailleur shifted into the spokes. With the coming of mountain biking, bent hangers became common because riders would pick up sticks or other debris in the chain and drag it into the derailleur as they continued to pedal. Steel frames could be straightened from some pretty alarming looking deviations, but an aluminum dropout was usually ruined. Aluminum frame builders started providing replaceable hangers. The concept spread to become the norm. Derailleur hangers are now one of several jigsaw puzzles that repair facilities have to solve on a regular basis.

Some early replaceable hangers were made of such soft alloy that they actually bent just from the ordinary stress of shifting. The first run of hangers on Specialized Stumpjumpers in the mid 1990s were notorious for this. If they weren't, they should be. The Big S eventually started making hangers out of steel, and has now evolved functional alloy versions, but it actually took them a couple of years to face their blunder. Denial was as big a force as "innovation" in the bike industry in the 1990s.

Having finally hit the right level of fragility in the hanger for the drive trains of the time, the industry looked at it no more. They moved on to adding as many cogs as possible to the rear cassette and making chains as thin as necessary to fit the stack of saw blades that now makes up a modern gear cluster.

One dozen sheet metal cogs, so thin that they have to be pinned together with little rivets to keep the damn things from folding under load

Compare the 8-speed chain (top) with the 12-speed chain (bottom) While either can be bent with sufficient stress, the force required to tweak the tinfoil specimen is far less.

The cassette in the top picture is a 12-speed 10-50 SRAM Eagle. Ten to fifty. The biggest cog is the size of a chainring. It's the thickness of a chainring, too. They can't get away from that. Anything that big has to have the strength to support itself, connecting rivets or not. 

By killing off front derailleurs, the industry avoids the biggest bender of chains, but I have ridden double and triple front chainsets for 43 years and never bent a chain. Meanwhile, to provide some vestige of the former range of available gears, the industry has to cram more cogs in the back, over a range that requires a rear derailleur like a crane to cover the span and manage the chain wrap. That's a long cage, my friends. And a skinny, skinny chain. The chain line sprawls so widely that the chain is vulnerable to bending just from the normal deviation. Add a traumatic factor like a stick in that dangly derailleur cage and you'll be lucky to get away without trashing the chain along with whatever else gets mangled.

In the case of the Trek Fuel, the tech who checked it in saw the twist of the rear derailleur and stated that the bike needed a new derailleur hanger. But when I looked at the hanger it appeared pretty straight. The alignment gauge confirmed this. This is after we had to buy a new alignment gauge, because the old Park DAG-1 doesn't have a long enough nozzle to get into a derailleur hanger buried beneath rear suspension pivots. Our old, old Campy gauge worked, but the DAG series has a more sophisticated system for checking alignment around the circumference of the rim. It's tweakier to use, but gives a more fine-tuned look at any problems. Or, in this case, lack thereof.

The robust replaceable hanger had stood firm. The evolution of the drive train has moved the weakness in the system out to the $125 derailleur instead of the $32 hanger.

Once I replaced the derailleur, the drive train made a very slight clink noise while running the chain on one of the cogs down on the high end of the cluster. Thinking that this may have been the gear in use when the stick jammed, I wondered whether the sheet metal cog had picked up a slight kink. Because cog teeth normally exhibit a cyclic pattern of offset teeth, a slight but larger deviation is hard to assess. Because the clusters are all riveted together, if any single cog had been rendered unusable, the rider would have had to purchase an entire cassette at $215 retail. If it was only the chain, that's a mere $42. But you can see how the cash cost of a minor mishap can add up very quickly. The fail-safe part of the system was the least affected by the accident in which it was intended to take the most damage. That stress was distributed to the rest of the parts. 

The damaged derailleur was bent in the middle, so that the upper and lower pivots were no longer parallel. The rider said he stopped pedaling the instant he realized that the stick was in there, but the derailleur still went into the spokes. With a tubeless tire, spoke repairs can extend to dismounting the tire and possibly having to replace the sealed rim strip. Fortunately, the bends in the spokes did not appear too sharp, which would create a stress riser. I was able the true out the minor deviation in the rim without racking up uneven tension.

An autopsy on the bent derailleur revealed that the four pivot pins of the parallelogram were bent, making the derailleur twist increasingly as the rider shifted toward the larger cogs. You can get some replacement parts for the SRAM Eagle 1x12 derailleur, but not those parts. Removing them is a destructive process, because they are riveted. Just like SRAM Double Tap road shifters, the part you can get is not in the area that actually breaks.

I'll keep taking the money to work on this crap, but I will never stop pointing out that it's expensive, ephemeral bullshit: the exact antithesis of everything that made bikes a great piece of technology for decades. The mythical free market demonstrates time and again that the consumer's taste for excruciatingly engineered junk sucks all the money away from simple, durable items of lasting quality.

The Economics of Expertise

In ancient times -- you know, the 1970s and '80s -- bicycle technology was mostly simple and accessible. Beginning with Shimano's introduction of Rapidfire in 1990, the industry has been systematically destroying that, but you can still find it.

Someone who has evolved with the bicycle over the past several decades will understand what the new stuff is trying to do, based on its ancestors. The processes are simple. The new crap just uses more complex internal parts to achieve the same basic objectives.

Most sophisticated modern componentry can't be repaired. And why would you? The stuff you bought this year has already been thrown on the scrap heap by the manufacturer. And yet, people persist in keeping their bike equipment for as long as they can, and trying to get it fixed when it malfunctions.

There are plenty of YouTube videos of people fixing things. Personally, I would rather have an exploded diagram and a brief, written synopsis of what's inside something than watch someone who may or may not know what they're doing demonstrate it in a video, but I'm old, cantankerous, and independent. But I digress...

Through the efforts of intrepid people doing what they've been told they can't, the independent mechanic community has discovered some of the inner mysteries of modern widgets that have accidentally been made accessible enough to open up, explore, and attempt to repair. Sometimes the repair depends on having donor organs available from another specimen that may conveniently have failed in a different way. Other times, the inner mysteries just need a good cleaning.

When bikes were truly simple and accessible, a rider could perform complete disassembly and reassembly in a small room with simple tools. Some of them had to be bike-specific, but they would all fit into a small toolbox. You could fake a work stand and a truing stand in various ways, although those items do make even simple maintenance luxuriously easier. When I was briefly well funded in the late 1990s, I bought shop-quality versions of both of those. But without them I still managed to build wheels and perform complete overhaul after complete overhaul through the years.

I don't say you need to be so ambitious. I only point out that it was possible. My riding buddies and I did invest in a truing stand that we owned communally and passed around so that we could each build wheels for our commuting and racing bikes. It was a beat-up old thing, not a real shiny Park, but it got us started. Other than that, we got by with improvised facilities wherever we lived. Our apartments always had telltale handlebar scuffs on the walls of the sparsely-furnished living room. In one apartment, my crazy friend Mark painted the rear stays of his racing bike fluorescent orange, leaving the outline on the wall. It was a pretty slummy apartment. I'm not sure whether we even lost our security deposit over that. The owner of the building may have thought she owed us some compensation for the cockroaches and the lack of heat.

Fast forward to the present day: I was digging around inside a SRAM Rival brifter that had suddenly stopped responding to commands. SRAM lets you right in there. I've even seen at least one video in which the rogue mechanic shows how easy they are to service. "Easier than Campy," he says. Even now, we still relate high end road stuff to Campagnolo. The thing is, Campagnolo provides parts and service instructions, or at least they did a couple of cogs ago. In the past several seasons, I have not had to delve into anything later than very early 10-speed from the Italians, if anything. I have not bothered to keep up, because I need brain space for more common issues.

SRAM does provide ample access to the interior of their brifters. After that, you're on your own. Since the simple task of changing a shift cable can involve removing the handlebar tape and exercising a large portion of your profanity supply, anything deeper than that will take a lot of time and require that you keep track of a lot of little parts. And therein lies the economic problem.

A shop has overhead. Back in the 1990s, mechanics learned as quickly as they could to fix as many esoteric problems as possible, because an active riding population was constantly challenging us to prove that we knew our stuff. The industry steadily added complexity and buried the mysteries under more layers of concealment, but the changes were incremental enough, and the equipment remained mostly simple enough for a good wrench to maintain some level of stature from season to season. And it made financial sense for the shop to project this image of expertise. Riders would tell their friends where to get the good work. They might buy bikes by price and brand image, but enough people were really riding to understand the value of reliable repairs.

As complexity reached ridiculous levels and the industry had simultaneously managed to wedge rider groups firmly apart, all-around expertise became harder to maintain. Riders, particularly in hard-core mountain and road categories, now want to go where they recognize the nonverbal signs of group allegiance. The divisions were always there. Now they are more starkly prominent. But, beneath it all, certain mechanical concepts still make everything work. So a shop that may not appear obviously in with the in-crowd may still have a person or two on the staff who can get into a mechanism, figure out what ails it, and fix it or definitively declare it dead.

Unlike the hospital, our shop tends not to charge the customer if the patient dies. This means that we have to make our assessment quickly enough that we haven't lost a lot of billable hours. And, if we think it might be repairable by someone more in tune with a particular bike type, we say so. In the case of the SRAM brifter, we ordered and installed a new one. Then I started an autopsy on the old one. The ratchets are not engaging correctly, and probably cannot be stabilized. Donor parts from another unit might work, but the experiment would take hours, at shop rates.

It makes no sense for the customer to invest in that or the shop to spend the time. But a bike owner with mechanical aptitude and spare time might get in there, diagnose the problem, scrounge up the parts, fix the unit, and announce triumphantly that shops are ripping you off when they say your brifter has to be replaced. Total cost in commercial shop hours might have been $300 for a part that retails new for $114, but that part of the calculation is ignored.

What the industry deems unfixable they will make sure really is unfixable. Until some mad scientist develops a little machine for home mechanics to clone repair parts using recycled plastic bottles and metal cans, powered by scampering cockroaches or little solar panels, true repairs will be the exception.

Dual Drive earwax

One of the early repairs this year is a Cannondale hybrid with SRAM Dual Drive rear hub. The owner said the bike had been stored for a couple of years. He reported no functional problems, only requested that we figure out how to raise the handlebars.

Dual Drive uses internal hub gears combined with a freehub body to provide a set of external gears shifted with a derailleur. This particular bike had a three speed hub.

A single unit controls both internal and external gears

The control linkage for the internal gears has been disconnected in this picture. 

While I had the wheel out for truing I noticed that only one of the internal gears was engaging. I remounted the wheel to confirm this by reconnecting the shifter. Yep. It was totally air pedaling in low and medium gear ranges. This was not just an adjustment problem. 

With a little Internet mining I got to a tech manual on SRAM's site that showed me how to break into the thing, but I saw nothing in the schematics that suggested an obvious flaw that would lead to this kind of gear loss. The troubleshooting guide made no reference to it. Based on some residue on the end of the hollow axle I suspected congealed grease in the "lifetime lubricated" inner recesses. I flooded the hub with light oil and left it overnight. 

This morning it works.