If you’ll permit me a brief digression in this story about threading a heavy metal rod through my $220 bicycle and bending it out of its original shape on purpose, I would just like to begin with some good news and get to the more confusing stuff after.
My project was a simple one: Instead of buying a modern steel bike for something around $1,600, I’d buy an old one for cheap ($220 in this case) and upgrade it with modern parts. I wanted to be able to ride off into the woods and camp for a week or a weekend, snacks in a basket up front, sleeping bag rolled up in the back. I wanted to go bikepacking, and I didn’t want to blow a ton of money doing it.
The simple plan became immediately less simple once I actually started working on the bicycle itself, a 1989 Schwinn Voyageur I dragged out of someone’s basement. Whoever had it before me rode the thing into the ground just about, but the frame seemed fine and off I went, pedals wobbling as I rode home. Everything snowballed from there. To get a basket up front I needed wider handlebars. To get wider handlebars I needed a new stem. And new brake levers. And new cables. And new brakes to fit the new wheels that I needed to fit the new tires I needed to ride rough dirt tracks.
The good news is that after months of trial and error, I cleared the first big hurdle of this bike build. The brakes had absolutely wrecked my shit. I was sure that if I just re-adjusted the old cantilever brakes that were on the bike already, they would work fine. They did not. No matter how I set them up, they barely stopped the bike. Historically, cantilever brakes were made obsolete by v-brakes starting in the 1990s, so I bought some v-brakes I was sure would fit. I raced around town to get a set, only to find that my bike was built in such a way that v-brakes could never really fit. On the day before a long ride through the woods in New Jersey, I took a chance on a modern cantilever brake (a $10 Shimano CX50) I found on Craigslist. I was up until one in the morning cussing and adjusting the brakes to fit, and fell asleep before I had a chance to test them. The morning of the ride, I carried the bike downstairs to take it around the block. I got up to speed and grabbed full brake, just like I always did on my weak old setup. I nearly went over the handlebars. The new brakes worked, and they worked perfectly, all through that day’s ride.
(This may not be the same opinion shared by all the people on the ride, who watched me at one point get distracted, not look where I was going, and slowly guide my bike off a road and into a drainage ditch, where I bailed out into a soft pile of leaves. My bike and I were unharmed.)
Things were finally going my way. It wasn’t just that the bike was working; what I had done to the bike was working. All the hours I’d spent reading on the internet, all the hours I spent greasing and de-greasing parts on my kitchen floor, the blood and the sweat and the cost and the frustration, all that had come together.
Before I knew it, just about everything that I could add new to the bike I had added. The handlebars, the brakes, the wheels, the tires, everything. There was just one last piece that I couldn’t change by putting something on the bike. I had to change the bike itself. This is where the metal rod comes in.
Look at this very scientific setup, with my bike close to the camera, and the bike that it is trying to do the job of behind it. The bike in the back is good to go right out of the box, a Surly Straggler, but it also costs about $1,600 up front.
The whole crux of this project is that I could get what was a top-rung touring bicycle of the late ‘80s for what a Walmart bike costs today. What a top-rung touring bike got you in ‘89, though, has gone obsolete over the intervening years. Everything was made to be as slim and as light as possible. Skinny tires on 27″ wheels, laced to a narrow 126 millimeter wide hub. If I wanted modern gearing, modern wheels, modern tires, I had to switch not only my wheels, not only my tires, but also the hubs that they all mounted to. And the modern hub I got wasn’t 126mm wide. It’s 135mm.
It’s not like it didn’t fit. Steel is strong, a good and easy material for making a bike out of, but it actually has a decent amount of flex in it. That meant I could just sort of wedge the modern wheel with the modern hub in place and the frame itself would bend the nine millimeters difference. The only problem was every time I took the wheel out, the frame would spring back to its original shape. Changing wheels was a process of pulling and wheezing and cramming and I was kind of over it. If I ever wanted the new wheel to really fit, I would have to permanently bend the frame so it wouldn’t spring back. This is called “cold-setting” the frame.
To cold set a frame, you stretch the steel frame of the bike far past where it would have to go so that when it springs back in, it’s still wide enough to fit the new hub. Here’s a video where a guy stretches his steel frame from 126mm in the back to about 160mm just so that it can spring back to around 130mm.
There are tools to do this job, and it’s something any local bike shop with a bike stand could do. So I called all the local bike shops around here and nobody wanted to do it. I understood why. Who would want to bend somebody else’s old bike to the point where it might crack? One guy said he wouldn’t do the job even if I signed a liability waver, another told me that he’d tried doing it himself and admitted he’d cried when he cracked his frame.
I resigned myself. This wasn’t exactly a critical issue, just a minor annoyance. I stared out of windows at grey skies. My old frame accepted the modern hub, so why even bother futzing with it? These were the thoughts going through my head as I picked up a long metal rod out of the shed and walked over to my bike to do it myself.
If you want a full guide on how to cold set a frame yourself, the classic Sheldon Brown site has an excellent primer with photos included. The process, though, is as simple as it can be. Slide your makeshift lever (in this case a metal rod) through the bike, pivoting it on the seat tube. Pull up on the rod and you’ll bend what are called your chain stays up and out. Here’s how Sheldon put it:
There are a number of ways to do spread a frame. Probably the easiest way is to use a lever. A piece of 2 x 3 or 2 x 4 lumber, roughly 5-6 feet long works well for this:
Sounds pretty low key for the process of taking a lever and bending your bike out of shape! The way that these bikes are put together means that there are a lot of joining points and bends, all of which could, as far as I knew, bend too much and crack. I’d have to get them welded back closed—not a process I was rushing to. More realistically I could be trashing my frame, sending me back out onto the internet to find another old frame to torture.
I had watched I don’t know how many YouTube videos before diving into this, but I wasn’t prepared for the fact that this bike isn’t just a little bit flexible. It is very flexible. After one or two presses, the bike was set at 135mm.
If that seems like a short description of the process, it only represents how short the job was in and of itself. A few pulls and I was wide enough. The only reason it took more than a few minutes was that I checked and re-checked and re-re-checked all my measurements, trying to be even on the left side as on the right.
I had the bike now wide enough, but I knew there was one more step to do. Once the bike frame is stretched, you need to check the alignment. You need to see that the rear wheel still points straight. There is another tool for checking the alignment of what are called dropouts (the slots where the wheel itself mounts on the bike) and I drove the frame over to the bike shop in town for them to do the job for me.
As I pulled up, the owner of the bike shop came out. “I thought I heard an air-cooled pull up!” It wasn’t long before I was hearing the guy’s Volkswagen story (he has a Type 2 in a barn) and he took me into the shop to double-check what I’d done. After looking it over, he disappeared into the back of the shop. When he came back out, he had the alignment tool, and rather than do the work, he let me align the bike myself, holding the frame for me and helping me eyeball the thing and make sure it was straight.
You can see how this aligning works here, but basically you get two metal rods with wide ends and you bend them so the faces line up flat:
Nothing scared me about this bike project more than the prospect of cold-setting the frame. I had spent weeks worrying about it, and had everything but nightmares about cracking this frame. As it was, the whole thing took an hour or two and went better and easier than I ever expected.
I will admit at this point, that I have not been entirely honest with my reasoning for wanting to cold-set the frame. I had a sneaking suspicion that spreading the rear chain stays out to accept the wider frame would give me more clearance for wider tires. A few months back I had bought a set of somewhat wider gravel tires for the bike, as wide as I thought would fit. I replaced the 28c wide tires (nominally 28mm wide) with 35c tires with a knobbly tread. When I took the bike home from the bike shop fresh from its alignment, borrowed a set of 42c tires from my girlfriend and mounted them up to my tires. For a moment, it all looked perfect. The rear did seem a hair wider, and I had a bit of clearance side to side.
All that work was, basically, for nothing. All that fear, all that risk, all that worry. The bike “fits” my modern hub now, but if it fits a wider tire than it did before, I can’t really tell.
Am I bummed out about it? Not at all. I’ve ridden the bike around some of the four-wheeler trails up here upstate where my girlfriend and I have decamped for the past month, and the bike rides like a dream. The 42c tires do fit up front if not in back, and the bike happily plows through mud and slop. It feels trusty but fun, light but not like it’s about to fall apart.
It rode well before I cold-set the frame and it rides well after it. The only thing that made any substantial change was me, in that I now have more familiarity with another pointless fix.
As it was, the problem was not a big one. I was fixing, I guess I could say, something that wasn’t broken. Years of futzing with my old car had told me this was a bad idea. Don’t let the old car magic out. Don’t try to change what isn’t broken, lest you break it yourself. Mis-adjust the timing that was running fine, and you’ll strip a thread double-checking something that didn’t need it. Don’t take needless risks, especially when being able to reverse what you’ve done isn’t a given.
Post time: Jun-01-2020