Every CB550 owner arrives at the same fork in the road eventually. The bike runs, the tune is honest, the brakes work — and it still feels like it's carrying about forty-nine horsepower's worth of ambition and rather less than that in delivery. So you start reading. And on the SOHC/4 Owners Club forums there are three threads waiting for you, none of which we've covered in this series, and which together add up to something more useful than any one of them alone.
The first is a fifty-four-thousand-view thread from 2007 in which one member takes a Dremel to a cylinder head on HondaMan's advice and touches off a debate about port finish that's still going. The second is a hundred-and-forty-reply argument about a $120 eBay big bore kit that promised to turn a 550 into a 600. The third is the oldest and most-read of the lot, and it isn't about making power at all. It's about a ten-inch steel rod that costs almost nothing, is no longer available from Honda, and quietly ruins the riding experience of every CB500 it breaks in.
Read together, they describe the whole arc of chasing performance on one of these engines: what actually works, what you're really buying, and the humble part that can make all of it feel like nothing.
The Dremel thread: what porting a 550 head actually means
The thread starts almost by accident. A member had pulled his wife's '78 CB550K apart to fix leaking cylinder head oil-feed O-rings, mentioned it in passing while asking HondaMan about something else, and got a suggestion back: while you're in there, clean up the intake ports, polish the intake valves, match the runners. His first reaction was the honest one — I've never done that kind of work and we've been happy with the bike as it is. Then he picked up the Dremel anyway.
What he found when he cleaned the head is the part every 550 owner should know, whether or not they ever touch a grinder. There was a genuine mismatch between the intake openings in the head and the inner walls of the intake runners. The insides of the runners were coarse from the sand-casting cores. There were casting marks in both the intake and exhaust ports, and machining "ledges" left in the aluminum around the valve seat pockets. None of this is wear. It's how the engine left the factory.
HondaMan's explanation of why is the most quietly interesting thing in the thread. These are late-1960s designs from a Honda that was not yet the financial giant it became, and the engines were drawn up around using as little metal as possible for cost reasons — thin walls, barely adequate fins, and castings done too hastily for real accuracy. He's since seen enough 500 and 550 heads pass through his hands to report that roughly a third of them show factory-matched intake spigots, and you can always tell, because whoever did it ground them without smoothing them. Post-1976 heads got more of that hand attention than early ones.
The work itself, as documented, is less exotic than it sounds. High-speed steel hemispherical bits for the actual metal removal — preferred over carbide because the larger flutes clear aluminum better — then small flap wheels, medium followed by fine, then a Scotch-Brite wheel for the final pass. The main structural change is narrowing the valve guide boss by about a quarter and blending rounded valleys either side of it, feathering the top of the boss into the taper of the guide. On the exhaust side he deliberately did almost nothing but remove casting marks, on the reasoning that the surrounding metal is there to carry heat away from the valve guide and the stock exhaust is restrictive enough that opening the port up gains you little.
Two details are worth stealing regardless of how far you go. The intake valve backsides can be polished by chucking the valve in a drill and working the back with emery cloth, taking great care not to touch the sealing face or neck down the stem. And when he went to match the runners to the head, he found slop between the mounting holes in the runners and the studs — enough to throw off the concentricity of the whole joint. He made small sleeves from thin-wall aluminum hobby tubing to take up the gap and hold the runners in one repeatable position before matching anything. That's a five-cent fix for a problem that would otherwise make careful work pointless.
Safety notes from the thread, unglamorous but real: eye protection and a dust mask, because the cutters throw aluminum slivers and the flap wheels make a great deal of metal dust; a wrap of heavy duct tape around the valve seat so a slipped bit doesn't nick it; and a thorough wash of the head in hot water and detergent afterward, because grit left in a port is grit headed for a cylinder.
Smooth or rough? The argument that never resolves
The thread had barely started before someone objected that a polished intake is the wrong goal — that you want a degree of roughness to keep fuel from separating out of the air, and that it's the exhaust you want slick so carbon has less to cling to. That's a widely held view and it's backed by real reasoning: a very smooth wall accumulates a stagnant boundary layer, and fuel drops out of suspension against it, entering the cylinder less thoroughly mixed. Members cited engine builders who glass-bead their ports rather than polish them, and one posted photos of intake ports cut with fine spiral threads — a commercial treatment sold on exactly that boundary-layer logic.
HondaMan's response is the reason the thread still gets read. He didn't dispute the physics. He disputed the transplant. Thread-cut ports, he pointed out, debuted on high-winding, computer-managed, multi-port fuel-injected engines, and every successful implementation he knew of depended on injection tuning. He knew three drag racers who'd paid to have the treatment done and then paid again to have it removed, because the power came in peaky and uneven. Rough walls genuinely help in engines with large, slow-moving intake tracts — big V-8s, Harleys, lawnmower engines. But a CB550 is a small, asthmatic, 1960s design breathing through two-jet carburetors, and in his experience smoother simply flows better in these, even without opening the ports up at all.
Another respected voice in the thread settled the anxiety more bluntly: the member had removed material from the right areas, which is the part that matters. Wall finish is where porters argue; port shape is where the power is. And years later, when a newcomer revived the thread wanting to do the same job to his '75 CB550F, the advice had compressed into a single sentence from one of the long-timers — don't try to port the head as such, just knock off the obvious casting lumps that don't belong and fix any mismatch you find between the head and the manifolds.
That is, for most of us, the whole answer. You are not going to out-flow a professional head shop with a rotary tool. You are going to correct manufacturing sloppiness, and on these engines that is worth real money.
The cam sprocket, five degrees, and a rat-tail file
Buried in the same thread is a modification that costs nothing and gets recommended more often than porting does: slotting the cam sprocket to advance the camshaft a few degrees.
HondaMan's description of the effect is refreshingly specific. Power comes on sooner and falls away sooner at the top — above 8,000 rpm it drops off at five degrees of advance, though not severely — with noticeably more torque in the 6,000 range. He ran three degrees on his own bike and wished he'd gone to five. The problem he was solving is one every 550 owner will recognize: in stock trim the bike feels flat between 55 and 75 mph, and this is the change that perks it up for fourth-gear passing on two-lane roads.
The execution splits along the lines you'd expect. The clean way is a rotary table in a mill. One member reported doing it with a die grinder, slotting only the hole he needed since the sprocket is centered by its flange anyway. HondaMan's contribution to the how-to is characteristically deflating: he did all of them with a rat-tail file in the 1970s, and only got access to a mill later. Blue Loctite on the bolts, and buy new ones, because the temptation to over-torque them is real.
One clarification the thread had to make twice, and that's worth repeating: advancing the cam and advancing the ignition are entirely different operations. One is breathing timing, the other is spark timing, and doing the cam properly means a degree wheel, a dial indicator, and finding true top dead centre rather than trusting the marks on the spark advancer.
As for fuelling, the original poster's own carb outcome is a useful data point. He went up from 90 to 100 main jets after his pipes started turning gold, tried dropping the needle clips a notch to richen the midrange, found the bike fell on its face under a handful of throttle, and went back to the centre clip with the larger main. Plugs came out a proper grey. He also flagged something easy to overlook while the carbs are off — check whether the screw plugs sealing the vacuum ports protrude into the carb throats, because if they do they disturb airflow and want grinding flush.
The $120 big bore kit: what you're actually buying
Five years later a different thread opened with a much simpler question. Someone had spotted an eBay listing for a CB550 600cc big bore kit at around $120 and wanted to know whether anyone had tried it.
The first objection was technical and immediate: the pistons were 61.5mm, and the received wisdom on the board had always been that 61mm was as far as you could safely go on stock sleeves. The answers that came back were candid. Yes, there's metal there — you could physically bore further — but thin cylinder walls warp under heat, and nobody had put enough miles on a 61.5mm 550 to say how it ages. One of the more experienced engine men added a warning that has nothing to do with this particular kit and everything to do with old motorcycles: use the standard oversizes, because if you jump straight to the limit you have thrown away your ability to rebore the barrels ever again. Have the shop final-hone each bore to its own piston and number them.
Then came the quality question, and the honest answer turned out to be "good value, not finished." A member installing the same manufacturer's 836 kit for a CB750 reported flash inside the piston skirts that needed smoothing, oil holes needing deburring, every accessible edge needing chamfering, top and second ring end gaps close but requiring matching to within a thousandth, and oil rings that were uniformly too tight. His verdict was the one to carry into the purchase: what you save in money you make up in time, and you should measure everything before assembly. Someone else who wrote to the seller looking for boring specs and jetting guidance got back a cryptic reply from someone clearly working in a second language, and no technical support at all.
The most useful argument in the thread is the one about whether to bother. One member put it sharply — if you're already doing head work, fitting a cam and upgrading the ignition, why spend the effort on big bores that don't raise compression? The correction came quickly, that unless the pistons are dished the compression ratio does rise with the larger swept volume, and the discussion moved to how far you'd want to take it: planing the head to bring things toward 10:1, checking chamber volume and valve clearance afterward, and the practical ceiling of roughly 10.2:1 on modern pump gas. Someone noted a member running 12.5:1 on the street, which drew the best line in the thread — that he had the luxury of fuel that hadn't been force-fed corn.
The defence of doing it anyway is unglamorous and pretty convincing. Your bores are worn already. These pistons cost about a quarter of first-oversize Honda parts. You're paying the machine shop for the labour regardless, so the incremental cost of a bit more displacement is close to zero. And if you live somewhere ethanol-free premium is a two-hour round trip, a little more compression is not what you're chasing.
The thread also killed a tempting shortcut. There's a well-known write-up on the board about modifying standard 61mm CB750 pistons to fit a 550 — cheap displacement from a common part. A retired engine builder consulted by one member pointed out that the domed CB750 crown intrudes far higher into the chamber than the 550's flat top, likely landing somewhere around 12:1, and that you can't machine much off the crown before there's nothing left. Someone who'd actually bought a set of the modified pistons confirmed the practical version: they weren't machined evenly and the skirts fouled the crank.
What settles the thread is the long tail of owners reporting back over the following years. The consensus on the budget kits landed at "genuinely fine, with clean-up" — sharp edges to relieve, quality otherwise good, with the expensive high-compression kits from the specialist suppliers acknowledged as better and priced accordingly. One member ran his with stock valves and the stock cam and had zero clearance issues, noting the pistons sit at the same height as stock and are simply larger in diameter. Another, two summers in, had skipped the kit's own rings and liners, added a 650 cam and a primary tensioner, rejetted the stock carbs and fitted a race baffle, and reported an instantly noticeable gain in acceleration and roll-on. Piston-to-cylinder clearance in the thread settles around 0.0012–0.0015 inches. Make sure the head is flat, and use a multi-layer steel head gasket in the correct oversize — at least one member ordered the wrong size first.
And now the part that undoes all of it
The third thread is the most-read of the three, and it opens in Houston in 2006 with a member stuck indoors during monsoon season, writing up what he'd learned taming the clutch on his '73 CB500. What follows is a small masterpiece of forum writing, partly because of what he got right and mostly because of what he got wrong.
His diagnosis of the failure chain is exactly right and applies to any of these bikes. The clutch lifter sleeve cracks — he found that nearly every one in his local junkyard was cracked in the same place. But the crack is a symptom. The clutch rod shortens with years of use, the lifter gets over-adjusted to compensate, the over-adjusted lifter puts the cable cinch where the drive chain can hit it, and the hammering cracks the sleeve. Fix the sleeve without fixing the rod and you have bought yourself another cracked sleeve.
Then he described the clutch rod as a two-piece assembly, and one of the board's ex-Honda mechanics gently corrected him: it's one piece, and it gets broken by a loose chain. The original poster's response — cheerful, self-deprecating, immediately asking to have his post edited so a moderator could copy the corrected version into the FAQ — is the reason that thread is still worth reading. The correction improved the record instead of ending the conversation.
And the follow-up explained something genuinely elegant about the part. The original rod isn't plain steel; it has an aluminium centre section, and the reason is thermal expansion. Aluminium grows at roughly six times the rate of steel, and as the engine heats up the lifter mechanism effectively moves away from the pressure plate. A rod with an alloy centre expands at a closer rate and keeps the clutch from dragging when hot, particularly if your adjustment isn't perfect. That is a lot of engineering thought in a component most owners have never seen.
Because the part was discontinued in 2007 and has stayed that way, the thread turned into a long, collaborative parts hunt, which is exactly the sort of thing a good forum does better than any catalogue. The dimensions were eventually pinned down at about 10⅛ inches — call it 257mm — and 3/8 inch in diameter, with the shifter-side end ground flat and square and the clutch-side end finished as a neat half-sphere. Members made replacements from drill rod or silver steel, with the strong recommendation to harden the ends, and a good suggestion to case-harden rather than through-harden so the core stays ductile and the rod bends rather than snaps. Others found substitutes: the early CB450 shares the part number, and one member successfully adapted a rod from an '04–'07 CBR1000RR, half a millimetre smaller in diameter and about 20mm long, trimmed to fit.
Three diagnostic details from the thread will save someone a weekend. The rod is directional — the domed end goes inboard toward the clutch, the flat end rides on the ball, and reversed it will skate off the ball and travel askew. The ball bearing inside the lifter is part of the effective length, and if it's missing the clutch will never fully disengage no matter how you adjust it. And if you can't withdraw the rod from the left case, do not conclude it's intact; one member did exactly that, pulled the right case instead, and found it in two pieces with metal shavings for company.
Two more findings are pure cafe-racer content. First: you cannot run an 18-tooth front sprocket on a CB500 to drop your highway revs the way 550 owners do. There is barely any clearance, and the chain will snap the clutch rod. If you've read our earlier piece on CB550 gearing, this is the same conversation arriving from a different direction — the drive chain on these bikes is quietly load-bearing for things that aren't the drive chain. Second: drag bars will make your clutch feel heavy. The original poster fought a stiff-but-smooth clutch for a year before working out that the sharp bends his low bars forced into the cable were the entire problem. Wider, slightly higher bars fixed it. If you're committed to the drag bar look, shorten the cable to suit.
The unglamorous closers are worth writing on the wall of the garage. The shifter shaft seal leaks on every SOHC/4 ever built, and it's a few dollars — replace it, because a wet clutch wants the oil level near the max line and you can't hold that with a weeping seal. Grease the lifter mechanism through its fitting as part of routine maintenance. And on the aftermarket "power levers" that promise a lighter pull: the thread's verdict, from the man who'd tried everything else, was don't bother.
The through-line
Put these three threads next to each other and a hierarchy falls out that nobody in any of them states directly.
The cheapest real gain on a CB550 is correcting what the factory didn't finish — matching the intake runners to the head, taking the casting lumps out, tidying the guide boss. It costs a Dremel and a careful afternoon and it is the one modification nobody in the thread regretted. The next cheapest is a few degrees of cam advance, which needs a file and some patience and transforms exactly the part of the rev range you actually ride in. Displacement is available, cheaply, and it works — but it's a machine shop job, the budget kits arrive needing an hour of deburring and measuring, and it's the only one of the three where getting it wrong costs you the barrels.
And all of it — every hour with the grinder, every degree of cam, every extra cubic centimetre — arrives at the countershaft through a ten-inch rod that a loose chain can snap and that Honda stopped making two decades ago. That's the real lesson of putting these threads side by side, and it's the same lesson the forum has been patiently repeating in every subject we've written about so far: on a fifty-year-old motorcycle, the fundamentals are not the boring prelude to the exciting part. They are the exciting part. Get the clutch working, the chain properly slack, the seal replaced and the oil where it should be, and a stock 550 will feel quicker than a modified one that's fighting itself.
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 the better part of twenty years grinding, measuring, arguing, correcting each other in public and publishing the results — including a fair number of expensive mistakes made so the rest of us wouldn't have to. If you own one of these bikes, those threads deserve reading in full, and the forum deserves supporting. There is no substitute for an archive that deep.
When the fundamentals are sorted and you're ready to build the bike you've been picturing, that's where we come in. The Thirteen Motorcycles CB550 catalog exists for the parts worth buying right — designed for this bike, made to work together, so the time you spend in the garage goes on the interesting problems rather than the ones somebody already solved.
0 comments