Every cafe racer build begins as an act of subtraction. Before you have bought a single part, you have already — in your head — taken off the centre stand, the rear fender, the grab rail, the mirrors, the enormous taillight and the great steel mudguard hanging over the front wheel. What's left looks lean and purposeful and about forty pounds lighter than the thing actually sitting in your garage.
Three of the biggest threads on the SOHC/4 Owners Club forums are about that instinct, and we haven't covered any of them in this series yet. One is a 249-reply argument from 2011 that started with a simple question — what else can I get rid of? — and turned into the most complete inventory of removable weight on a small-block Honda four that exists anywhere. The second is a 272-reply, fifty-nine-thousand-view fight about the front fender, which is not really about the fender at all. The third is a twelve-page, eighty-thousand-view how-to on drilling your brake rotors that has been running since 2005 and contains, buried on page four, the one piece of metal on a CB550 that the veterans will tell you never to put a drill anywhere near.
Read together, they answer the question every 550 builder actually has: how much can I take off before the bike stops being better and starts being worse?
The inventory: what actually comes off, and what it's worth
The weight thread opens in April 2011 with a member who wants to lose his centre stand but keeps using it, which turns out to be the perfect framing for everything that follows. Almost every item on the list that gets built over the next ten pages is something that weighs real pounds and does a real job.
The centre stand leads the list because it's the biggest single lump of steel you can remove without a cutting wheel. On a late K it's close to ten pounds with the pivot tube and spring. The starter motor and its associated hardware is another ten or more. The stock four-into-four exhaust is heavy enough that several members named it as their single biggest saving, and a four-into-one drops a meaningful chunk. Then it gets granular: the rear fender with its light, plate mount and grab bar as an assembly; the stock seat pan, which is heavier than it needs to be; the chain guard; the big lens turn signals; the left-hand switch housing if you're willing to run a self-cancelling clutch lever; one mirror instead of two; a smaller battery; braided lines and deleting the brake pressure switch junction.
The most instructive contribution came from a member who listed everything he'd done in one season — integrated LED taillight, battery and battery box gone in favour of an aluminium plate carrying a regulator and a capacitor, a permanent-magnet alternator, fibreglass seat pan, no chain guard, lighter chain and sprockets, aluminium fasteners — and then admitted he couldn't say conclusively how much difference it made, because he'd done it gradually over months with nothing but his own impressions to go on. He then listed a second set of changes made purely to make the bike feel lighter: fresh bushings and bearings, rebuilt forks, tapered steering bearings, the seat profile brought as low as possible, and clubman bars. That distinction — between mass removed and mass you can feel — is the most honest thing in the thread.
The rotating and unsprung weight discussion is where the numbers get interesting. The aluminium rims from a GL1000 or CB750A come up early as a swap, and one of the racers in the thread pours cold water on the headline figure: they're only about a pound lighter apiece than the stock steel rims. But he immediately adds the qualifier that matters, which is that a pound at the rim is worth far more than a pound on the frame, because it's unsprung and it's rotating. A magnesium rear wheel was reckoned at sixteen to twenty pounds saved on a 750. Thinner discs from other manufacturers were worth about two pounds each. Wheels, tubes and tyres together were put at over forty pounds on a small block — which is why every serious weight-saver in the thread ends up talking about wheels within a page.
And the two centre-stand workarounds are worth writing down, because they resolve the thread's opening dilemma completely. The first is simply a pair of stout wooden blocks: put the bike on the side stand, slide one block under the frame on the right, lift by the bars until the bike is level, and slide the second block under the left. Thirty seconds, no tools, rear wheel off the ground. The second is more elegant. With the stand up and the spring under tension, you feed pennies between the spring coils with pliers until the spring is held stretched; put the bike on the stand, and the now-slack spring lifts off. Pull the cotter pin, loosen the pivot tube pinch bolts, and the tube slides out. The stand lives on a shelf in the garage, and whenever you need it — chain lube, rear wheel out, a tyre change on a trip — you hold it up to the frame, slide the tube back in and use it. The member who described it had been doing it that way for twenty years without dropping a bike. The spring, it turns out, exists only to stop the stand dragging.
The objection, and why it's worth taking seriously
What makes this thread more than a shopping list is that it has a genuine opposition, and the opposition is not stupid.
The first objection is arithmetic. Unless you're racing, one member argued, shedding weight does close to nothing you can measure — maybe a couple of miles an hour, probably not a full mile per gallon. His stronger point was that centre of gravity matters more than mass, and he offered a comparison that's hard to argue with: a GL1000 is comfortably a hundred pounds heavier than a CB750 and is nonetheless easier to tip into a turn, and dramatically easier to lift off the side stand, because of where its weight sits. The centre stand, he and others noted, is about as low on the motorcycle as it's possible to be. It is not the weight hurting your handling.
The second objection came from one of the board's most respected long-timers, and it's the one worth pinning above the bench. He pointed out that everyone in the thread was talking about flickability as though it were unambiguously a virtue, and that flickability is a two-way property: the same bike that responds instantly to your steering inputs also responds instantly to a crosswind and to the turbulence off a passing truck. He rides the other direction on purpose — fairing, bags, luggage rack — because he'd rather hold the line he chose. And he asked the question that nobody in the thread really answered: what's the longest ride you routinely do on your radically lightened SOHC/4?
The counter to both is simple and it also holds. A lighter bike stops better, turns in with less effort and is easier to pick up when you've stalled it on a hill facing the wrong way. One member with a basket-case CB500 documented the honest version of the whole project: 425 pounds when he started, 592cc and 385 pounds when he finished, with better brakes and better suspension, and — importantly — still comfortable to ride. Another framed it as a straight trade you make item by item rather than as a philosophy: for each removable part, is the weight worth the loss of convenience? One part won't matter. Fifteen parts will.
There's also a purely practical argument for losing the centre stand that has nothing to do with weight, and it's the one that convinced us. On a bike ridden hard, the stand is often the first thing to touch down in a corner. More than one member reported grinding one halfway through. If you're building a cafe racer and you intend to actually use the lean angle, ground clearance is a better reason to take the stand off than ten pounds ever was.
Where the thread lands is roughly this: weight reduction on a 550 is real, it's free, and it will make the bike more pleasant. It will not make it a modern motorcycle, and anyone selling you that is selling you something. Do it because you like the way a light bike rides, not because you have a plan involving a stoplight and a CBR.
The fender: the removal that isn't free
Which brings us to the one item on the list that isn't a straight trade — and to a thread that started with a member simply reading a product description and asking whether it was true.
He'd been looking at a fibreglass replacement front blade, and the seller's own copy carried an unusually blunt warning: Honda designed these forks around the heavy steel brace under the fender, don't run them without it, severe wobbling will result. Their suggested method was to drill the rivets out of your stock fender, keep the steel brace and hoops, and rivet the light fibreglass blade onto the original hardware. His question was reasonable — he'd seen plenty of bikes running no fender at all, so was the warning real or was it lawyers?
The answer arrived from about six directions at once and it did not fully agree with itself, which is why the thread ran for eleven pages.
The sceptics had a decent case. You can flex that pressed-steel piece with one hand. It bolts on with small screws. The axle is clamped solidly and the triple trees are substantial, so how much can a bit of tin halfway down the legs really matter? And — the strongest version of the argument — the road racers of the early seventies routinely ran fibreglass fenders on simple side plates or no front fender at all. Harley's KR and XR racers had neither. The BSA and Triumph triples ran without. Photographs of Dick Mann's Daytona-winning Honda in 1970 show a fender that isn't hung from anything resembling a brace. Were all of those teams ignoring physics for the sake of looks?
The rebuttal came in two parts, and both are worth understanding rather than just accepting.
The first was a matter of understanding which direction of flex you're testing. Yes, you can bend the brace fore and aft with your hands — but the forks aren't trying to go that way, so that flex costs you nothing. The load that matters is twist: the fork lowers rotating relative to each other about the steering axis. With the brace bolted at four corners, twisting it is very hard. Without it, the only things holding the two fork legs in register are the axle and an inch or so of aluminium clamped by the pinch bolts at the bottom of each leg. On a CB550 this is worse than on a twin-disc bike, because the single front rotor loads one leg and not the other every time you touch the brake, which is precisely a twisting input.
The second part was the historical correction. Braces kept getting added to those race bikes later, and every subsequent generation of motorcycle has arrived from the factory with a proper one. Vintage racing organisations now typically forbid fitting a brace to a machine that didn't originally have one, on the grounds that it's an unfair advantage — which tells you what the sanctioning bodies think it does.
The CB550-specific detail in this thread is the best part, and it complicates both sides. A member with a '77 550K pointed out that his brace is mounted to the fork legs through rubber grommets — six of them, and they're right there in the parts fiche — which makes it very hard to call it a rigid structural member. Another owner with a '72 CB500 initially insisted his was bolted solid, then checked and conceded that his had been modified by a previous owner and the grommets were factory. So the factory piece is a compromise: a fender mount with useful bracing properties, deliberately isolated so the fender doesn't rattle or crack.
And then a member who happened to own both a '76 and a '77 CB550K just went and tested it. He pulled the fender and brace off the '76, stood in front of each bike with the front wheel gripped between his knees, and twisted the handlebars. The '77 — brace in place — was substantially harder to twist. He noted the caveat honestly, that the '77 lowers are longer and may be stiffer on their own. It's not a laboratory, but it's the only direct CB550 A/B comparison in eleven pages and it points the same way as the theory.
The thread's practical conclusions are the useful part, and they're not "keep your ugly fender."
If you want the bare-fork look, take the fender off the brace and bolt the brace back on. Powder-coat it black and it disappears against the tyre. You get the whole aesthetic and lose nothing structural. If you want to do better than stock, fit a real fork brace — the kind that clamps the top of one fork lower to the top of the other, making a rigid rectangle out of the two legs, the axle and the brace itself. That's a genuine upgrade over the factory piece, and several members reported it as the single most noticeable handling change they'd made to these bikes, most obvious when trail-braking into a corner.
What you should not fit is the other thing that gets called a brace: the bar that clamps around the fork tubes up near the triple clamp. That's a tweak bar, it exists for extended chopper forks, and on stock-length legs it has to sit directly under the lower triple clamp, where the triple clamp is already doing that job. A brace works best down near the top of the wheel, midway along the legs.
Sourcing one for a 500 or 550 was, and remains, the sticking point — the popular off-the-shelf makers cover the 750 and not the small blocks. The thread's answer was to have one made; a member with a '78 550K got as far as a custom fabricator willing to build one given accurate fork dimensions.
And if you're going to run without any brace at all, the thread's most reasonable voice had the right framing: that's a legitimate choice, and plenty of people ride that way for years. Just make it knowing what you've traded, and don't tell a newcomer that the brace does nothing.
Drilling rotors: the good version of taking metal off
The third thread is the oldest of the three and the most cheerful, because for once the subtraction is nearly all upside.
It opened in 2005 with someone asking how to drill the rotors on a 750F, and the first substantial answer set the tone for everything that came after. The early Honda rotors are big slabs of stainless — around four pounds each on a 750 — and they're hard. The best way to do it is on a mill with a rotary table, which lets you place holes precisely by degrees and keeps the material distribution even so the disc stays balanced. The classic pattern is the alternating two-hole, three-hole arrangement. Countersink each hole slightly to knock the burr off and to reduce the chance of cracks propagating between holes, and don't get ambitious with hole diameter.
The DIY route works too, with caveats that come up over and over across twelve pages. Use a sharp, decent-quality bit and a drop of cutting oil for every hole. Go slow. Step up through sizes rather than trying to punch the final diameter in one pass. Clamp the disc down onto something solid so the bit isn't breaking through into air. A pre-drilled rotor from another bike makes a perfectly good template if you don't want to lay one out, and CAD templates and printable patterns circulated freely in the thread. One member drilled all three of his rotors with a single cobalt bit on a mill and chamfered thirty thou each side afterwards, which suggests the "you'll destroy a dozen bits" warnings are worst-case rather than typical.
The single best piece of advice in the whole thread costs nothing: check your rotor for runout before you drill it. One member spent an hour and a half drilling a second rotor for a dual-disc conversion and only then discovered it was warped and unusable.
The countersinking question never fully resolves, and both sides have a real argument. The pro-chamfer camp says it removes burrs sitting below the disc surface that sandpaper can't reach, reduces cracking, and stops the holes holding water. The anti-chamfer camp points out that factory drilled and slotted discs are made with sharp edges and no chamfer at all, apparently on purpose, because that sharp edge scrapes water off the pad. Both sides agree that if you drill properly — sharp bits, stepped sizes, right speed, plenty of fluid, backed by something solid — you barely get a burr to argue about.
What do you actually gain? Honestly, less than the internet implies, and the thread is refreshingly candid about it. Around half a pound comes off a stock disc, which is unsprung and rotating and therefore worth more than half a pound elsewhere. Several members said flatly that they'd never felt an improvement in dry braking and that they did it because it looks right. The one benefit that came up repeatedly and convincingly is wet weather. A solid disc in the rain can be genuinely alarming — nothing happens, and then everything happens — while a drilled disc is merely a bit weak for the first fraction of a second. Squeal drops too, because the holes keep the surface cleaner.
There's a companion trick from the same era that's worth more than the drilling and takes ten minutes: these Honda discs glaze over, and breaking the glaze with wet-and-dry paper reduces squeal and improves bite. If your 550 stops poorly and the rotor looks like polished chrome, that's your first move, not a drill press.
And if you don't have a mill, a drill press, or the patience — the thread turned up multiple small shops running rotor exchange and drilling services at very reasonable money, including resurfacing with a non-directional swirl finish rather than a concentric one, which stops the pad crawling across the face and putting unnecessary load on the caliper. Twenty years on, the specific businesses have moved on, but the specification to ask for hasn't.
The line the veterans won't cross
Which brings us to page four, where a member with a '74 CB550 asked the obvious next question. If drilling the front disc helps with cooling, gas and weight, why not drill the rear brake surface too?
The answer came from one of the board's ex-Honda mechanics, and it stops the conversation cold. The rear drum on a CB500 and CB550 is not a simple steel or alloy casting. It's a cast iron liner cast together with the aluminium hub — and those hubs are notorious for cracking on their own, without anybody's help. Replacement hubs, when you can find them, cost real money. He'd had a couple salvaged by having the old liner machined out and a new one shrunk in, on the machinist's explicit condition that the wheel never see a race track. There is a maximum service diameter stamped inside the drum, and once you're past it the hub is scrap.
That's the boundary. On a CB550, the front rotor is a slab of stainless with material to spare and a well-documented century of people improving it. The rear hub is a known crack risk holding your wheel together, and there is no version of drilling it that ends well.
The through-line
Put the three threads next to each other and a rule falls out that none of them quite states.
Removing weight from a CB550 is one of the few genuinely free improvements available to you, and the forum's inventory is the best list anywhere: centre stand, starter and its wiring, stock exhaust, rear fender assembly, chain guard, seat pan, battery, signals, switch housings, and — the biggest single opportunity — the wheels. Do it because a light bike is nicer to ride, and be honest with yourself about the fact that convenience is what you're spending.
But two of the things you'll want to remove aren't paying for themselves in weight. The front fork brace weighs almost nothing and is doing structural work that gets more important the harder you ride and the more you trust that single front disc. Take the fender off it if you like the look; leave the brace, or better, replace it with a real one. And the rear hub, which looks like it has metal to spare, does not.
It's the same lesson the SOHC/4 forum has been patiently repeating in every subject we've covered in this series. The parts that make these bikes good are rarely the parts that look exciting, and the fastest way to build a 550 that disappoints you is to remove something that was quietly holding the whole front end honest.
Credit where it's due
None of this is ours. It belongs to the members of the SOHC/4 Owners Club community, who have spent two decades weighing parts, twisting fork legs between their knees, ruining drill bits and arguing about it in public so the rest of us don't have to learn it the expensive way. Those threads are worth reading in full, and the forum is worth supporting. There is no substitute for an archive that deep.
When you've decided what's coming off and you're ready to think about what goes back on, that's where we come in. The Thirteen Motorcycles CB550 catalog is built for exactly this stage of the project — parts designed for this bike and made to work together, so the weight you take off doesn't cost you the way the bike rides.
0 comments