Thursday, July 13, 2017

Amazing Calipers

In 1986, I was working at a shop in Burlington, Vermont as I was finishing work on a college degree. I was hired as a contractor by the US distributor of Look's new all-carbon framesets, which were being shipped out of the Geze ski binding warehouse on Pine Street. I believe that, by this time, Geze had sold its binding business to Look.  The issue was that they had been getting a lot of returns from dealers with the complaint that the frames were out of alignment. The initial KG86 model was made by the French company TVT and used carbon fiber tubes, bonded to aluminum lugs and fittings, using the proven adhesive technology perfected by ALAN on their aluminum frames. The KG86 was the first commercially successful carbon fiber frame, discounting the earlier Exon Graftek and the Raleigh SBDU experimental carbon frames, which all used similar technologies.

French on the left side, English on the right.
One design constraint of carbon fiber frames is that they cannot be cold-set after manufacture, to correct small variations in alignment. My task was to remove each frameset from its packaging, place the frame and fork in alignment jigs, and record its actual measurements. I used the New England Cycling Academy's portable frame alignment system jigs to make the measurements. This jig attached to the bottom bracket shell, as with a full surface plate, but then used precision bearings to allow a beam to provide a rotating reference from which to make measurements. I expect that it may not be quite as precise as a properly setup granite surface plate, but it was a well established system, so I believe the measurements were of acceptable accuracy.

As I recall, the results I got showed that the majority of the frames were, indeed, off by a few millimeters. However, the amount they were off ranged from 0 to a maximum of 4 mm, with most being in the range of 2 - 3 mm. The misalignment could be dropout spacing or centerline alignment, but dropout parallelism was generally quite good.  Main triangle alignment was generally fine. My sense was that almost all the frames were fine, with the amount they were off being too small to notice while riding. I would expect that the misalignment was understandable, considering the state of the technology used to make the frames. It was my sense that the real problem was shops that were pandering to an overly-discerning clientele, attempting to establish their elite status as "pro shops" by misapplying steel frame expectations to an emerging technology. After all, Gereg LeMond and Bernard Hiault were successfully duking it out in the Tour de France on these frames! I also think that the shops had found that the frames, which were about twice the cost of a quality steel frameset at the time, were not selling as fast as they would have liked, leaving Look USA to search for a way to either send a bunch of frames back to France, or to seek a discount.

When the project was finished, I had measured up over 400 framesets. The project took longer than anticipated and the folks who had hired me tried to wiggle out of their initial agreement, but I had kept excellent records of my hours and stuck to the rate under which I had been hired. As I recall, it totaled a bit over $400, which I estimated being about the wholesale value of one of the frames, so I made an offer that they simply give me one of them in my size. For some reason, they wouldn't go for this, so I took the cash.

Fast forward to this past winter when I found a KG86 in my size on eBay for a good price--much less than what I was paid for the project, even without factoring in inflation, which makes that $400 worth about $1,000, today. The decals have a few gouges and the fork was incorrect, but the frame apparently was ridden very little, and I'm happy with it. I expect I will eventually find the correct fork for it, but this one is quite close, in spite of the Litespeed yellow highlights.

Look KG86 - The first commercially successful carbon fiber frame
When I initially built the bike up, I decided to try to use as many French components as I could, and to use parts from the same time period as the frame. If I didn't have a suitable French component, the next choice would be European. I had some cool stuff I had accumulated over the years, so almost everything came right out of the basement.
  • Mavic: headset, crankset, hubs, rims (wheels are SSC, which is pretty much just some decals), stem
  • Sachs freewheel, 13-23, 7-spd
  • Sedisport chain
  • Huret titanium Success front and rear derailleurs
  • Simplex retrofrictions shifters, of course
  • Look pedals--certainly!
Non-French components
  • Campagnolo seatpost (25 mm diameter)
  • Sampson Stratics sealed bottom bracket. I'll probably swap this out for a Swiss Edco, once I replace its bearings.
  • 3TTT handlebars (they probably made the Mavic handlebars, anyway) 
  • Universal LS-1 (Italy) brake levers--just because they're composite, cool, and there has to be some reason I've held onto a brand new pair for 30 years
  • Concor saddle (Italy)--it doesn't fit my butt at all, so perhaps I'll swap it for my ultra-cool Ideal 90 with titanium undercarriage
  • Tape is Orbea, which is the current iteration of the Spanish Zeus company, no telling where it's actually made
  • The sewup tires are high quality, but I'll likely swap them out at some point for a pair of Michelin tubulars that I have, which are exceptionally nice
  •  Tacx (Dutch) bottle cages -- just because they're cool
It's the brakes that are the real point of this post, though. In the photo, the bike has Zeus Supercronos calipers, which were the only non-Campagnolo, European brakes I had that were both short-reach and recess-mount.  While they fit beautifully, and are like-new, I found that the springs are a bit too strong for the mechanical advantage of the levers, and the first ride showed that it was not a great combination.  While the brakes stopped quite well, they were difficult to modulate, as I needed to squeeze the heck out of the levers to get any braking at all, and I have a very strong grip.

I found a set of MAFAC's very last model brakes for sale in France. I had not heard of the LSX model before, but learned that it was a Hail Mary play by the French company in the 1980s. MAFAC was the first company to market cantilever (1946) and centerpull (1952) brakes, but once they had established these designs, they ceased to innovate and simply manufactured them for decades, with small modifications. This really wasn't much different from most bicycle component manufacturers at the time, but by the mid-1970s, MAFAC's products were very stale and the Japanese had become the innovators. The LSX was MAFAC's pull-out-all-the-stops attempt to produce the finest single-pivot sidepull, along the lines of the Modolo Professional, and I think they succeeded. The calipers extensively use high-strength aluminum pieces, superbly machined and finished.  The precision is impressive, the fit on all parts is perfect, and the features are both impressive and, in some cases, unique.
Mavic LSX brakeset
The mounting bolt arrangement is quite interesting, allowing the caliper to be fitted to either recessed or nutted frames and forks.  Different length bolts were available. This could present a bit of a challenge today, as it uses a fine thread. I've labeled the few steel parts with the letter S in the photo. The black piece next to the left arm is a special rubber piece designed to fit over the quick release nut to keep it from scratching the paint on the frame, similar to that used by other manufacturers at the time. The left arm has a nylon insert to keep the arms from rubbing together (some Japanese brakes also have this feature). The arms have brass inserts for the pivots and are separated by plastic washers, backed up by a steel bearing washer. The brake shoes are asymmetrically designed to direct water away and out from under the pads. Even though they are 30 years old, they stop impressively well--with similar grip required for dual-pivot calipers. They are, however, a weak part of the design, as their mounting bolts are designed to mate with the wheel guides and the arms have a recess for these guides. Since the holders are closed at both ends, replacing the rubber pads will be difficult at best, and good luck finding replacement shoes for these rare birds.

The levers are scratched, so perhaps at some point I'll file and polish them and fit them to the bike, but I'm in no rush as the Universals work fine. I'm just surprised that these fine brakes were not accepted better in their time, though I suspect that the cost of manufacture made these too expensive and low-margin to turn around MAFAC's faltering finances. Eventually, MAFAC was snapped up by Sachs, as it consolidated many other failing French companies, like Sedis, Huret and Simplex.  I may eventually swap the derailleurs over to Mavic, but for now the Huret units shift at least as well as any other non-indexed shifters I've ever used, while being even more rare than the Mavics of the era.

I've ridden the bike twice, on 25 and 45-mile rides, each featuring about 10 miles of gravel and with steep (17%) climbs. The ride is quite nice, very similar to that of a quality steel frame of the era, but smoother over rough surfaces and without any chainring rub under heavy load, indicating less flex in the bottom end. Otherwise, it's very similar to my mid-1970s Masi Gran Criteriums in terms of handling, as could be expected, considering the geometry. It rides nice enough that I expect it won't be a wall hanger.