Despite the fact that the contemplation of Quality and motorcycle maintenance led to the complete mental breakdown of a man much brighter than myself, I feel moved by recent event to broach the subject. Whereas one of Pirsig’s primary problems was in definition (e.g. What is Quality?) I am going to expedite the process (and preserve my mental state) by offering a working definition. For the purposes of this discussion, quality will be defined as “The ability to finish a fair number of races without material failure.”

An army fights on its stomach, the Army Of Darkness, however, is fueled principally by coffee. Rich black coffee. Properly roasted and prepared coffee has over 600 discernable flavors as well as the delightful stimulant caffeine. Coffee is of such central import to the team (the team responsibilities reasonably approximating a second job for most of us involved) that we have a special equipment crate in the race trailer for the maker, the cups, the filters and most importantly, the coffee.

As any coffee aficionado (or Starbucks dilettante) will tell you, there are a myriad of varieties of coffees with diverse flavors and characters, however, these coffees can be broadly grouped into two sets: Arabica and Robusta. Broadly speaking, Arabica coffees are the good stuff. They are more difficult to grow, have smoother flavors and have about half the caffeine content of Robusta coffee which allows for an addict (let’s just come out and admit it here) to easily maintain a happy medium between the productive, alert and motivated and the familiar state of the over-caffeinated: jittery, nervous, irregular heartbeat and an attention span too short to install a rear wheel, much less assemble an entire race bike from its basic component parts. Robusta coffees tend to have an overly acidic (instead of just crisp) or vegetal flavor and, on the whole, just taste bad. They have twice the caffeine of Arabica coffee and even a cup or two can send even the dedicated coffee drinker into the land of the irregular heart beat and that twitchy eyelid thing.

In the bad old days before good coffee came back into fashion in the early 90s, Arabica coffee was only to be found in exclusive gourmet shops hidden away in the dark alleys of major cities. Virtually all pre-ground supermarket coffee sold under the major labels consisted primarily of Robusta coffee. But it wasn’t always like that.

Going back to the beginning of the century, Arabica coffee was the major ingredient in most commercial blends and coffee was one of the most popular beverages in the country. However, the major coffee producers realized that they could mix a little Robusta coffee into the blend and the consumers would not be able to tell the difference. The cost of the blend was reduced (since Robusta coffee is cheaper) and the coffee producers were able to make greater profits. After this first experiment the coffee sellers began, each year, to mix in more Robusta coffee and less Arabica coffee with greater profitability being recognized each year.

This went on for a number of years before the law of unintended consequences reared its often ugly visage; people stopped drinking coffee. Sure some of the oldtimers were still choking down bitter watery brown liquid but the time honored generational traditional of passing down a coffee addiction was ebbing. Young people, who had not slowly been introduced to Robusta coffee but were being asked (by the coffee mavens) to accept it in one fell swoop, found that the repugnant brew was not preferential to the myriad of other beverages from which they could choose.

As sales volumes for coffee steadily fell, the wholesalers were then driven to substitute cheaper and cheaper coffee into the blends in an effort to maintain profit margins. This, of course, merely accelerated the trend.

Arabica coffee, of course, was still available but was not commonly marketed. The shrewd business folks at Starbucks determined that people would, in fact, start to drink coffee and, actually, pay handsomely for the privilege, if the coffee was good. The reintroduction of high quality coffee to the American consumer sparked a wildfire of local coffee bars and gourmet coffee shops. The recent improvement in consumer taste for coffee has actually created an enormous glut of Robusta on the world market.

The lesson that the Robusta coffee people learned the hard way is that, you might be able to make a quick buck by decreasing quality for a little while, but eventually folks are going to figure it out by taste and experience and the new kids from Seattle will eat your lunch when they deliver a better product. Even if the better product costs more money.

After the coffee maker, the second most important piece of machinery in the AOD pits is the race bike itself. After struggling for a few years in the mid-nineties with Yamahas (lesson learned, just because the 400 is good doesn’t mean the 600 is any good) AOD switched to Suzuki in 1997. With the Yamahas we had experienced failed cranks, transmissions, coils, brake rotors, brake calipers etc. Although it took us a year or two to get the horsepower output of the Suzuki engine up to the level we had achieved with the Yamaha, the huge increase in reliability and trustworthiness of the basic engine architecture helped to obfuscate the shortcomings of a few bhp. We were able to capitalize on the Suzuki’s durability and proceeded to quickly improve our race results to the point where, in 1999, we won the National Middleweight Superbike Championship. In 1999 we were using engines from Suzuki’s model year 1998.

Our A engine (a 1998) ran for the entire 1999 season (and remember, this is endurance) served as a practice bike for all of 2000, was then sold and sprint raced for 2001 and 2002. The valve cover has not been off that motor yet.

Not what you want to see when you remove the valve cover.

For the 2000 season we purchased a brand new 2000 GSXR 600 and converted it to an endurance racer. We had built up a tidy lead in the points until a valve spring retainer broke in the engine during a race and allowed the competition to regain some lost ground on us.

Collet groove of a GSXR 4mm intake valve that has been sawed in half
lengthwise with a low speed diamond saw, embedded in epoxy filled
with 600 mesh silicon carbide abrasive to prevent edge rounding of the
exterior, then ground with 600 SiC/water slurry on an 800 rpm cast iron
wheel. The ground surface was polished with 6 micrometer polycrystalline
diamond abrasive, 3µm monocrystalline diamond, and .3µ aluminum
oxide abrasive, all at 200rpm on a groovy proprietary nylon polishing
substrate. The polished surface was then etched in 2% nitric acid
in ethanol (referred to as nital etchant), and polished again briefly
on .3µ aluminum oxide. The cracks in the hard coating on the
valve surface are clearly visible. This valve is just about to break.
The one that resides next to it in the motor has already broken.

We did some analysis on the valve spring retainers and found that Suzuki had changed the way they were made. We did not know if the new ones were cheaper than the old ones but, after our second failure, we determined that you could not run them for longer than 15 hours. This, of course, made it very difficult to endurance race the bikes in a series that had a 24 hour event. We almost lost the championship when we lost the second retainer save for the fact that our competition was also racing a GSXR 600 and, so, their motor failed as well. It was telling that a Yamaha actually won that race. We replaced all the valve spring retainers at regular intervals at great expense and hassle. This was the first sign that some Robusta coffee was being stirred into the mix.

In 2001 we were delighted with the new fuel injection on the new GSXR but were concerned when we saw how fragile the new motorcycles were. Even a minor tip over ruined subframes, broke mounting tabs off frames and bent the frame geometry. This was a marked departure from the previous generation which could be throw down the road with a fair degree of certainty that a clip-on and a footpeg would put the bike back in the race.

We chalked this development up to the “light is right” build philosophy and hoped that eventually Suzuki would throw a pound of aluminum back onto the frame so that the mounting tabs (footpegs, subframes) would not fail at the slightest provocation.

At the end of the 2001 season a more insidious problem reared its head when the Neighbor of the Beast engine mysteriously failed at the last race of the year. Fortunately it failed before the race. The post event tear down revealed that Suzuki has dropped the ball on the whole valve train design and/or execution again. Conversations around the pits quickly determined that many 2001 GSXR 600s have suffered a similar failure.

Back in the AOD research cave we tried to determine why the valves were failing. On the 2001 GSXRs the valve cotters etch the sides of the valves. This begins micro fractures in the hard coating of the valve which leads to the tip of the valve breaking off with any number of expensive consequences. The valve falls into combustion chamber and then ruins the head, the cam, the buckets, the piston and, sometimes, the cylinders and thus, the cases.

When the tip breaks, the valve will fall,
the piston will hit it, smash cam, lifter and all

The valve failure seems to occur at around the 25 hour of operation area. It is usually cheaper to buy a used motor, replace all the valve components (at a cost of around $600-700 dollars) then to attempt repair of the damaged engine. You can, alternately, replace all the valve components (valves, retainers, collets) regularly (the secret Suzuki recommendation is every 600 miles) to avoid this failure. This problem is well documented with Suzuki (although understandably not publicized) and Yoshimura has aftermarket valve collets which are intended to prevent this failure. The aftermarket collets are, of course, illegal for supersport/superstock competition, are not widely known or available and, have an unknown durability.

This whole problem is a huge and expensive hassle for the racer AND means that any used race bike or used engine with unknown mileage MUST be torn down and rebuilt before use. A survey of rebuild intervals of the AMA supersport 600 found that the GSXRs need to be rebuilt 35% sooner than the other brands. This should severely hurt the resale value of the GSXRs.

Ruined heads. This is starting to get expensive.

The problem, as our team learned to our dismay, does not end with the valve train. The transmission, once a strong point of Suzuki, is fragile and prone to failure in numerous fashions including: shift drum bolt backing out, dogs on gears rounding off, and the shift drum and shift forks wearing so that the shift forks no longer allow for the gears to change.

Hard coating is already wearing off of many spots of this shift drum after
1,500 miles. In other motors, this was quickly followed by transmission failure.

In order, there is a small bolt which holds the shift shaft detent star onto the shift drum. This bolt backs out on most bikes unless the clutch is removed, the bolt removed, a tap run through to clean out the threads, all threads thoroughly degreased using contact cleaner and then the bolt reinstalled with red loctite. If these steps are not taken the bolt will back out which can cause everything from a lack of shifting, severe internal damage to the gears as they try to engage in multiple gears simultaneously, and/or broken primary and/or oil pump drive gears. At worst this can cause accidents (ie, bike stops shifting on highway or racetrack), race losses, and damage ranging from a couple hundred (just tightening the bolt) to thousands of dollars. All because Suzuki won’t install the bolt properly.

Oil pump showing severe scoring. This came out of an intact motor
with one season of racing. Scores in housing reduce oil pressure.

The dogs (the little things that stick out of the sides of the gears and engage the gear next to them) are very soft in the new transmissions and will noticeably wear down. As the dogs become dull the gear’s propensity to stay engaged is reduced and the gear’s propensity to pop out of gear is increased. The gears have shown premature wear on every engine we have taken apart in out garage be it a race engine or a street engine with only 500 to 1000 miles. The gears are expensive and require the case to be split for replacement.

The shift drum (the thing with the grooves that guides the shift forks) seems to have a hard coating on it. The shift forks themselves also have a hard coating on most of the fork and on the nub that fits into the shift shaft groove. The hard coating wears off of the shift fork in less than a season of racing. This is noticed by the rider as the transmission becomes subtlety, but noticeably, more difficult to shift. Eventually the shift forks wear a severe groove in the side of the shift drum and will begin to hang between gears. The dogs on the gear teeth will quickly be destroyed in this circumstance. The only solution is to pull the motor and replace, usually, the shift forks, the shift drum, the gears and, probably the transmission bearings. It is important to note that this is happening on street bikes as well as race bikes although most of the street bikes are making it through the one year warranty period before failing.

A galled nub that should be smooth and hard plated.

There are also a few other anecdotal indicators of quality problems at Suzuki. The most frightening one we have seen so far was the catastrophic failure of a stock Suzuki GSXR 600 front wheel. The wheel had not been involved in any previous accidents. One of the spokes developed cracks at one of the normal holes in the wheel by rim. This crack propagated until the spoke failed. The other two spokes failed quickly and the rider was viciously thrown to the ground. This wheel was being used on a Suzuki factory sponsored racebike.

This is supposed to happen in a crash, not cause the crash.

Polished parts indicate enduring crack.

Other miscellaneous but unexpected failures include an output shaft bearing wearing out after a single season and, much worse, the body of the cam chain tensioner actually failing. This did not, fortunately, result in losing another motor but, of course, it would have if we had not caught it in time.

New style cam chain tensioner failed after 1,500 miles.

All of this points to a conclusion that, although the GSXR is a fast bike, perhaps the fastest bike, the build quality has fallen off the face of the earth compared to the bikes Suzuki was producing only three years ago. Perhaps the appallingly dismal Japanese economy (which has sucked for fifteen years) has placed such severe financial demands on Suzuki that the $100 or so they save by not properly installing transmission bolts, using poor coatings on the shift shafts, and installing inferior valve train parts is required to meet their quarterly earning targets. Certainly the many thousands of dollars which any GSXR owner must spend on parts to constantly replace worn components cannot be hurting the bottom line although the warranty claims to replace motors damaged in these failures in the first year have to offset the profits from selling valves, gears and cam chain tensioners.


But it makes for a bitter cup of coffee. And bitter coffee, we know, won’t sell in the long term.

Tim gets a bad cup of coffee

Any of the following actions might occur with negative financial consequences for Suzuki...

  • A street rider will get hurt when his GSXR transmission ties up in traffic.
  • An NHTSA investigation will follow and find that the GSXR is unsafe, Suzuki will be forced to recall the bike for repairs and it will cost Suzuki far more money to replace all of the worn out parts, perhaps perpetually because the NHTSA doesn’t recognize one year warranties, than it would have to build the things right the first time.
  • Suzuki will get a reputation for building fragile bikes which wear out quickly. The prices for used bikes will reflect this making it much more difficult for primary buyers to resell their old bikes in order to buy new ones. This will decrease the number of new bikes that Suzuki will be able to sell.
  • And, as Suzuki gets a reputation for building low quality bikes with high attached maintenance costs, more people will buy Yamahas. Say 100 people purchase R1s or R6s instead of GSXRs next year because of quality concerns. Suzuki will have just lost about $750,000 in revenue. And, as Ducati, Harley and various major brands of coffee have found, once you get a reputation for poor quality it is very difficult to live it down.

Alternately Suzuki can come clean about the problems with the 2001 and later GSXRs, increase the quality of the replacement parts so the engines will only have to be rebuilt once, and commit to making quality motorcycles from here on out.

I am urging the latter because no one likes to be left with a bitter taste in their mouth.

This sort of thing led to an expensive lawsuit with
a rider named Dunlop at the Isle of Man.

Pictures, unless otherwise noted, – Army of Darkness –Ministry of Information. All pictures are of parts from the 2001 GSXR 600s.