Lean Surge and Hanging Idle After a Filter and Slip-On
A 2005 carbureted 600 cc inline-four comes in with a slip-on muffler and a drop-in high-flow air filter the owner fitted 800 km ago. Complaint: a lean surge and stumble at a steady 1/8 to 1/4 throttle between 60 and 80 km/h, an idle that hangs at 2,600 rpm for several seconds after blipping from 4,000 rpm, and occasional afterfire on deceleration. You find float height 13 mm (spec 13 ± 0.5), pilot screws 2 turns out (spec 1.5), #38 pilot jets, #98 main jets (stock #95), needles in slot 2 of 5 (stock 3), and carburettor sync within 1 cmHg. Turning a pilot screw in does change the idle, and the hanging idle returns within 30 km of resetting the base idle to 1,200 rpm. A propane test aimed at the intake boot of cylinder 3 raises idle speed by about 200 rpm, and a second pass at the other three boots shows no change. The spark plugs are tan on cylinders 1, 2 and 4 and pale on cylinder 3.
1. Interpret the propane result and the plug colours together. What do they establish, and how does a single boot leak produce surge, a hanging idle and decel afterfire?
Read the worked answer
Propane is fuel. When you direct it at cylinder 3's intake boot and idle rises by 200 rpm, you have enriched a mixture that was lean at that exact point: unmetered air is entering there. The identical test at cylinders 1, 2 and 4 produces no change, and the plugs corroborate it, tan on three cylinders and pale on cylinder 3. Leak confirmed at the boot, localised to one cylinder.
One leak explains all three symptoms because a boot leak is a fixed air supply that does not pass the throttle plate. At 1/8 to 1/4 throttle the total charge is small, so a fixed trickle of air is a large fraction of it and cylinder 3 runs lean enough to surge and stumble; at wider openings the leak is diluted and the symptom fades, which is why the bike feels fine at large throttle. At idle the same air bypasses the throttle stop, so the idle stop has to be closed further to reach 1,200 rpm; when the throttle shuts after a blip the leak air keeps the engine turning at 2,600 rpm until the mixture and the closed-throttle vacuum settle. Decel afterfire is unburnt charge igniting in a hot exhaust from the same lean, misfiring cylinder.
Note also that the plugs confirm this is a one-cylinder fault, not a global lean condition. Global lean would pale all four plugs.
2. Why is re-jetting now the wrong move, and what specific evidence in the setup shows the owner has already been chasing the wrong variable?
Read the worked answer
Re-jetting now would mean changing two variables at once and losing the ability to attribute the result to either. The leak is a known, fixable defect that affects one cylinder; jetting is a global change that affects four. If you richen the pilot circuit or raise the needles to cover a 200 rpm propane response at one boot, you will foul cylinders 1, 2 and 4 (currently perfect tan) and the leak remains, still admitting the same volume of air at small throttle. You would have masked the symptom and created a rich-running bike that fails emissions and soots plugs.
The setup shows the owner has already been doing this. Needles were moved from slot 3 to slot 2, which is one step leaner, and mains went from #95 to #98, one step richer. That is not a coherent correction for a high-flow filter and slip-on; it is a guess made against a symptom rather than a measurement. The mains matter only near wide-open throttle and are irrelevant to a 1/8-throttle surge, which is pilot and needle territory. Pilot screws at 2 turns out versus a 1.5-turn spec is another small compensating tweak, and turning them does not stop the hang returning within 30 km, which is the signature of a mechanical air leak, not a mixture setting.
Fix the leak, restore the stock baseline, then test. Only measured lean conditions at a defined throttle position justify a jet change.
3. Lay out the repair and post-repair validation sequence, including what you do with the needle position and main jets, and how you prove the fault is gone rather than merely hidden.
Read the worked answer
Repair. Replace all four intake boots as a set, or at minimum the cylinder 3 boot plus any hardened ones. A 2005 machine with a leaking boot almost always has siblings that are close behind, and a second failure after the customer collects the bike destroys confidence. Inspect the boot clamps and the airbox-to-carb seals while the rack is off.
Restore the baseline. Set needles to slot 3, refit #95 mains, set pilot screws to 1.5 turns out. This gives you a known reference from which any further change is measured.
Validate in stages. Re-run the propane test at all four boots, expecting zero rpm change. Sync the carburettors and set base idle to 1,200 rpm. Ride 30 km and blip-test the idle; a hang that does not return within 30 km confirms the leak is gone. Then read plugs after a steady 80 km/h cruise: expect four matching tan plugs.
Only now assess jetting. The slip-on and filter may genuinely need one size up on the mains, but that is verified by a wide-open-throttle plug chop or a lambda reading at WOT, not by feel. If the bike is still lean at 1/4 throttle after the leak is fixed and the baseline restored, move the needles one slot richer and re-read plugs. Document each change with the throttle position, the reading and the outcome so the next technician inherits data, not folklore.