2026 Q3 Review - Hillclimb Car

The weather looked to be wet and windy for the duration of the day as we set off for a convoy lap of the Abingdon Carnival sprint course in the hillclimb car.

Despite being confident in the Impreza's abilities in the wet, some caution was still needed! The day's practice runs were approached with caution by both drivers, but an intermittent misfire started to occur, firstly when the car was pulling through the gears, and subsequently when idling. Due to the standing water on the course, this didn't affect our times as much as it would have done in the dry (as the car would have been on full throttle for a longer amount of time), but it was nevertheless frustrating.

The coils were swapped over to a spare pair during the lunch break to see if this would make a difference. With the EJ25 still misfiring on the third timed run, the spark plugs were (quickly) swapped back to a hotter set before the final timed run of the day, to see if this would change things. Strangely, this seemed to work, with the EJ25 pulling strongly, and the Impreza winning class A6 against a Golf R, and a Jaguar F Type R.

Our next sprint was to be at Castle Combe on the first of August, with the hot weather likely to give the tyres a hard time, and to make cooling the intake air a priority. It was the plan to give the WRX's map a check over on the Friday afternoon before the event, with some colder plugs gapped down to 0.6mm fitted prior to driving the car. With everything in place and the ECU working nicely, we set off to check that all was well before Castle Combe. Unfortunately, when the engine got up to its operating temperature, the misfire occurred again. Cylinder 3 seemed to be misfiring, as the exhaust manifold branch for cylinder 3 was cooler than the others. The hotter spark plugs were swapped back in, but again the misfire occurred. The coil packs were then swapped between cylinders 1 and 3, but yet again the misfire prevailed. Unfortunately at this point, time was running out, and the hatchback stepped in so that we could still compete in the sprint (see the hatchback blog to see how that went!).

A few weeks later, the investigation continued, with various theories conjured up in the interim about the problem's probable cause. We used a borescope to firstly inspect cylinder 3's piston surface, valves, and cylinder head bowl, and two other cylinders were inspected in the same manner for comparison. It seemed that the issue could be an ignition related issue, as the injector spray pattern could be seen more distinctly in cylinder 3, compared to the others. Both coil and injector tests could be undertaken via the Link G4X ECU, and with the injectors clicking, and the spark plugs sparking as they should, more investigation was needed.

The first theory was that the valve lashes had closed up for either the intake or exhaust valves in cylinder 3, which had subsequently closed the lash enough when the valves heated up so that the engine would lose compression when hot (although when cranking the hot engine via the starter motor, the regular noise indicated that compression was present in all cylinders). The RHS cam cover was removed, and the lashes were duly checked with the trusty feeler gauge. The lashes were within the required clearance window, so it was time to test the next theory, namely that one of the intake valve stem seals had dislodged, causing oil to enter the combustion chamber during the intake cycle, fouling the plug and causing the misfire. For this, both the intake valves stems in cylinder 3 were inspected after removing the beautiful Process West intake manifold, but no oil seemed to be present, and neither was there oil pooling at the bottom of the intake port. The next theory was that a break in the earths for either the coil or the injector plug for cylinder 3 caused the misfire. For this, the engine bay wiring loom was inspected, and continuity of the earth wires were checked, with all seeming to be ok. The last thing (which was a bit of a guess) tried was to swap injectors 1 and 3 over, to see if that made a difference.

The car was then run briefly to check if the issue was solved. From approximately 10 minutes of gentle driving, the missfire did not occur. However the real test would be to check how the engine performed under boost. We will be checking this soon so stay tuned to see if the issue is solved!