Restoring a motorcycle over winter: where to start

The classifieds are full of projects that stopped at 60 per cent. A bare frame on a trestle, three boxes of parts, an engine on a pallet, a repainted tank, and always the same line: "no time, hence the sale". The seller didn't run out of time. He ran out of method, and he noticed too late to turn back.

A restoration almost never dies on a technical difficulty. It dies on three things: a budget spent in the wrong order, an unobtainable part that blocks everything else, and a reassembly that turns out to be impossible because nobody photographed the strip-down. All three are dealt with before you pick up the first spanner. That's what this article is about, not how to lift a cylinder head.

Why projects die at 60 per cent

The stalling point is nearly always the same, and it's logical. The first sixty per cent are enjoyable: you strip, you clean, you discover, and every evening produces a visible result. The remaining forty are reassembly, which is the opposite: every evening runs into a missing bolt, a part that was never ordered, an order of operations nobody wrote down.

By then three counters are already in the red. The budget has been eaten by the visible parts, the timescale has stretched because two components are out of stock everywhere, and motivation has collapsed because the bike no longer looks like a bike. That is the exact moment the advert goes up.

The answer isn't to work faster. It's to decide from the outset in what order you spend, and in what order you strip.

Three hours of diagnosis before the first bolt

Before dismantling anything, you need to know what you've bought. Three hours of diagnosis save months of going the wrong way, because they answer one question: is the engine good?

• Compression. This is the test that sets the whole budget. On a healthy four-stroke petrol engine you'd generally expect more than 8 bar, with less than 10 per cent variation between cylinders. The exact figure for the model is in the manual, but the spread between cylinders can be read without one. If the engine is weak, the entire project changes scale. The full method is in our guide to compression testing an older motorcycle.
• Frame and mounting points. A bent frame, a marked headstock or cracked engine mounts close the subject. Look at front-to-rear wheel alignment, the condition of the original welds, and any evidence of repairs.
• The paperwork side. Frame number legible and matching the V5C, and the machine's status checked — not recorded as stolen, no outstanding finance. A bike with no usable paperwork is a heap of parts, not a restoration. Our pointers on checking a two-wheeler's frame number and status apply to a project exactly as they stand.
• The electrics. A loom bodged over thirty years, with connector blocks and insulating tape, costs more in time than an engine rebuild.

Write everything down, photograph the machine as it arrived, and only then decide the scope.

Parts availability is checked before, not during

This is the step almost everyone skips, and it's the one that kills projects. Before stripping anything, draw up the list of the ten to fifteen critical parts for the model and check they can actually be had: not "that must be findable somewhere", but a seller, a price and a lead time.

The items that block things most often: model-specific plastics and covers, headlight and instruments, original silencers, unusual casing gaskets, original sheathed cables, decals, and gear selection parts. On a lot of machines from the 1970s to the 1990s, the mechanical parts are still around and the bodywork disappeared long ago.

That list changes the nature of the project. If the right-hand side panel exists nowhere, you know it before you've spent £1,300 on an engine, and you can decide to accept a used-condition bike rather than aiming for original. That decision is easy to take in October and impossible to take in April with everything in pieces.

The order of works: make it run before you make it pretty

The rule that saves the most projects fits in one sentence: get the bike roadworthy before you make it beautiful. It's counter-intuitive, because paint is the fun part — and that's exactly why most people do it first.

An order that works:

1. Active safety: brakes, tyres, steering, suspension, lighting. Nothing that follows means anything without these.
2. Running: ignition, fuel supply, carburation, cooling, clutch, final drive.
3. Reliability: seals, hoses, cables, bearings, fasteners replaced with new.
4. Ergonomics and use: seat, controls, instruments.
5. Appearance: paint, chrome, decals.

The benefit isn't only rational: at the end of stage 2, the bike starts and rides up the drive. That moment is what fuels the following six months. A bike that has never run during the project has nothing to hold on to the person doing it.

Budget is broken down by job, not held as one lump

"I've got £2,500 for this" means nothing. What matters is what each job costs and which of them can't be squeezed. Here are market ranges for a mid-capacity classic road or trail bike in the UK, excluding rare parts.

Two lessons come out of it. First, the mechanical work often costs less than the cosmetics, when popular belief says the opposite. Second, three jobs are enough to blow an envelope: paint, chrome and wheel building. They are exactly the ones to price first and commit to last.

Finally, allow 15 to 20 per cent contingency on the total. It always gets used, and not on what you expect: it goes on new fasteners, on special tools and on postage.

What to farm out without a second thought

Farming work out isn't giving up, it's avoiding six months of a project stuck on a skill you don't have. Three families go out:

• Anything needing a machine. Face skimming, reboring, disc refacing, pressing wheel bearings in and out, balancing.
• Anything needing a trade. Paint, chrome, upholstery, wheel building and truing. A bad paint job is redone from scratch; it's the area where false economy costs most.
• Anything safety-critical you can't verify yourself. A steering or brake reassembly you aren't equipped to check to the specified torque.

Conversely, keep the carburettor rebuild, the electrical circuit, routine servicing and the reassembly: that's where you build your knowledge of the machine.

Photos, bags, labels

This is the least glamorous point and the most decisive. Reassembly is prepared during the strip-down, not afterwards.

• Photograph everything before you touch it. An overall view, a close-up, then each stage. A hundred photos cost nothing and replace documentation that doesn't always exist.
• One bag per sub-assembly, labelled by hand. Not a communal jar. Write down where it goes, not what it is: "front mudguard mounting, 4 bolts + 4 washers" beats "M6 bolts".
• Note bolt lengths when one casing takes several sizes. That's the classic mistake that puts a hole through a crankcase on reassembly.
• Keep a notebook. Date, operation, part ordered, price, reference. It serves reassembly, budget and resale.
• Never strip two sub-assemblies at once. One job open at a time, closed before the next one starts.

Rolling milestones every two months

Set stages that produce a visible result: engine running on the bench in December, rolling chassis complete in February, bike moving up the drive in March, cosmetics in spring. A project without milestones lasts indefinitely, and a project that lasts indefinitely ends up as an advert. Build the MOT into that plan too: a machine going back on the road needs one once it's three years old, and lights, brakes, tyres and steering are exactly what the tester looks at — one more reason stage 1 comes first.

To get started concretely: three hours of diagnosis and photos, the list of the ten critical parts with prices and lead times, then quotes for the three expensive jobs before anything is dismantled. If the tank is rusty, treat it straight away — it's a long job because of the drying time and it governs the carburation: our article on ways to treat a rusty fuel tank sets out the options and what they cost.

And for everything to do with reassembly figures — torque settings, capacities, valve clearances — L'Atelier's assistant gives you the manufacturer's data from the make, model and year, which saves guessing the torque on fasteners it took you six months to find.