This column is not an endorsement. It is a description. There is a difference, and we would like it noted.
The factory service manual is a document written by engineers for a hypothetical mechanic working in a hypothetical dealership with a full set of factory tools, an assistant, a torque wrench that was calibrated this year, and no particular time pressure. It assumes you have the correct service stand, the correct oil filter wrench, the correct valve adjustment tool, and the specific spanner that exists for this one procedure on this one model and which you can order from the factory for a sum that will surprise you. It assumes you are doing this job on a new bike, or at least a recent one, where the bolts have not had forty years to become one with the surrounding aluminum and the gaskets have not compressed themselves into a permanent opinion about their situation.
This is not a criticism of the manual. The manual is right. If you follow the manual exactly, with the correct tools, in the correct sequence, on a bike that has been maintained according to the manual's schedule since new, the job will go the way the manual says it will go and take roughly the time the manual implies it will take and produce the result the manual promises. This is a real outcome that happens to real people. We have heard about it.
What follows is a parallel document. Same jobs. Different column. We offer it not as advice but as companionship — confirmation that the gap between the manual and the garage is not a personal failing but a structural feature of the situation, experienced by everyone and acknowledged by almost no one in print.
| The manual says | What is actually going to happen |
|---|---|
| Remove the four bolts securing the side cover. Remove the side cover. | Remove three bolts. Discover the fourth has been cross-threaded at some point in the past and is now decorative. Spend twenty minutes on the fourth bolt. Remove the side cover. |
| Using the special service tool (part no. 07716-0020100), remove the valve adjustment locknut. | You do not have part no. 07716-0020100. Nobody you know has it. It is available online for sixty dollars and will arrive in four days. Fashion a substitute from a combination of a 10mm socket, a screwdriver, and optimism. It will mostly work. One of the locknuts will disagree. |
| Replace the gasket with a new one (part no. 12391-300-000). | The new gasket is no longer available. There are two alternatives: a pattern part from an overseas supplier with a three-week lead time, or the old gasket, which you will clean up and reinstall after convincing yourself it still has some life in it. It will weep slightly. You will fix it properly next time. |
| Torque the cylinder head bolts to 35 Nm in the sequence shown in Figure 4. | Torque three of them to 35 Nm. The fourth will reach approximately 28 Nm before you feel something that might have been the bolt or might have been your imagination. You will stop there. You will think about it for a moment. You will move on. |
| Allow the engine to cool completely before draining the coolant. | Wait forty minutes, decide that's probably close enough, proceed, burn your hand. |
| Check and adjust valve clearances every 8,000 km or 12 months, whichever comes first. | Check valve clearances when the bike tells you to, which is when it starts making a sound you haven't heard before, which is not the same schedule. |
| Apply a thin, even coat of assembly lubricant to the camshaft journals before installation. | Apply assembly lubricant. Slightly more than a thin coat because the can is hard to control. Acceptable. Continue. |
| Installation is the reverse of removal. | Installation is approximately the reverse of removal, subject to the bolt you set down somewhere sensible that is no longer there, the clip that went a direction you didn't see, and the gasket that was symmetrical until you started reinstalling it and discovered it isn't. |
The special tool question deserves its own treatment. Every service manual contains at least one procedure that requires a tool that exists only to perform that procedure, costs more than the part being serviced, and is stored in exactly one place in the country, which is at a dealership that closed in 2009.
The official position of the service manual is that you need this tool. The official position of everyone who actually works on old bikes is that you need this tool once, that fabricating or improvising a substitute is a reasonable response to a unreasonable situation, and that the factory designed the procedure around the tool they already had rather than the other way around, which means there are usually several ways to achieve the same result if you think about what the tool is actually doing and work backward from there.
The senior editor has a collection of improvised tools that have accumulated over the years on a shelf above the bench. A bent rod that does something specific to a specific Honda that he can no longer recall without context. A piece of exhaust pipe that serves as a driver for a bearing that a proper driver would cost forty dollars to press once. A wooden dowel wrapped in electrical tape that he used for something delicate. None of these are in any manual. All of them worked. Some of them worked better than the specified tool would have, because they were made for the specific bike rather than for the category of bike, which is a distinction the factory cannot make and you can.
"Installation is approximately the reverse of removal, subject to the bolt you set down somewhere sensible that is no longer there."
— the parallel manual
We believe in torque specs. We want to say that clearly before what follows. Torque specs exist because the engineers calculated the clamping force needed to keep a joint sealed under operating conditions, factored in the thread pitch and friction coefficient of the fastener, and arrived at a number that achieves that force without exceeding the yield strength of the bolt or the material it threads into. This is real engineering with real consequences. Fasteners that are under-torqued come loose. Fasteners that are over-torqued strip threads, crack cases, and create problems that make the original job look simple. The torque spec is correct.
What the torque spec assumes, however, is a new fastener in undamaged threads with a calibrated wrench and nothing unusual about the joint. What you frequently have is a forty-year-old bolt with a forty-year-old relationship with its hole, a torque wrench whose calibration you are taking on faith, and a gut feeling about what the fastener is telling you that the spec cannot account for. The spec says 35 Nm. Your hand says this bolt has been in here since Reagan and something is slightly not right about the way it's moving. These are both pieces of information. The spec is more reliable in general. Your hand is more reliable about this specific fastener right now. Experienced mechanics use both simultaneously, which is a thing the manual cannot teach you and which takes years to develop and which is, honestly, most of what the skill actually is.
What a bolt feels like when it's about to go: There is a feeling. It comes before the sound and well before the consequence. You learn it by almost making the mistake enough times that you recognize the approach. The manual cannot describe this. It can only tell you the number, which is correct, and trust you to develop the rest.
Which instructions to follow and which to adapt: Some steps in the manual are precise because they have to be. Others are precise because that is how manuals are written. Learning to tell the difference is the job. Valve clearances: follow exactly. Which direction to rest a component while you fetch a tool: use your judgment.
When to stop for the day: The manual has no section on this. It does not know that you have been at it for four hours and are now making decisions that you wouldn't make fresh. Some of the most expensive mistakes happen in the last hour of a long day on a job that was almost done.
The sound of a thread catching: Versus the sound of a thread not catching. This also cannot be described. It can only be learned, and it is worth learning before you learn it the hard way, which most of us did anyway.
None of this is an argument against reading the manual. The manual is indispensable and anyone doing serious work on an old bike without one is making things harder than they need to be. Torque specs, clearance specs, service limits, assembly sequences, the correct orientation of a one-way valve that looks the same from both sides — these things are in the manual and not reliably anywhere else, and getting them wrong has consequences that are directly traceable to not having looked them up.
What the manual is not is a complete description of the job. It is a skeleton. It tells you the sequence and the specifications. It does not tell you that the bolt in position three always corrodes worse than the others on this model, or that the o-ring in the secondary circuit likes to fold over on installation and will leak if it does, or that the workshop tip in the back of the Clymer is more useful than the factory procedure for that particular job and was probably written by someone who had done it on a cold morning and wanted to save someone else the same trouble. That knowledge is not in the manual. It lives in the forums and in the heads of people who have done the job and in the gap between what the manual says and what actually happened, which is where most of what you know about motorcycles was learned, if you are honest about it.
Follow the manual. Depart from it deliberately, with a reason, and note when you do. Learn the difference between a deviation that is fine and a deviation that will visit you three months from now on a wet road at an inconvenient moment. The manual is right. The manual is also incomplete. Both of these things are true, and working in the space between them is the job.