While assembling the Mazda RX8 engine I noticed a blue tag on the front cover that I did not see before could it be that this is not the original engine; the tag and numbers on the Rotor Housings suggest that this motor is not the original and was most likely changed by Mazda. There is no information on the engine history of this vehicle so I’m assuming that this is probably the second installed by Mazda. So this not an old worn-out engine after all and it makes the hard start-flooding issue even more interesting because hard-start issues are usually from high mileage engines.

I’ve now completed the reassembly of the Mazda RX8 motor which originally had the low-compression/flooding issues. I was really surprised to see the small amount of wear on the Rotor Housings. The apex seals were measurably worn with no chamfering or distortion. Side seals had the large clearance which seem to more on earlier 8’s. I am still pondering how the compression could have drastically dropped on the old engine (causing?) it to hard-start and flood. It almost doesn’t make sense; the typical RX8 scenario is one day the engine refuses to start and spins over freely (low compression) like an electric motor. Unless some extreme measure is taken such as pouring a light oil directly into the engine, it might remain in that mode indefinitely. After struggling to restart it suddenly catches and if allowed to run and warm-up is back to normal again-really weird. I am taking a second look at old rotor housings which have areas of dark brown deposits suggesting the engine apex seals were not scraping against the Rotor Housing surface. (See picture) I was able to remove the stuff in sheets with a razor blade. Over the years the old bugaboo about Rotaries and synthetic oils is the lubricant’s tendency not to burn-off completely in the combustion process and leaving deposit-build up on the Rotor Housing surface. I have seen this many times over my long career building rotaries but could never tell if synthetics were responsible.

I removed the Mazda RX8 engine from car and dissembled it. I anticipate seeing something inside the engine that would explain periodic loss of compression when the engine goes into the flooded-no start mode. The condition of the Rotor Housings are excellent with minimal wear and minor grooving. There is some light carbon staining in some areas (photo) which i consider normal. All seal and springs look ok; apex seals show normal wear with no chamfering. Side seals are in good condition but have too much clearance where the butt against the corner seals. This is the only measurement that I would consider abnormal in the engine seals that could profoundly affect the engine’s compression. My overall impression after disassembling the engine is that there is nothing visible that could cause the compression to drop, causing flooding. I am thinking if any seal become stuck in it’s groove it could cause a loss of compression but all seals moved freely.

After removing engine from car I did a closer inspection of ignition coils only to find out that they were replaced by previous owner. The coil mounting plate shows the characteristic white spots where the old coils bleed-off voltage to ground. Mazda RX’s like other rotaries need exceptionally strong ignition for start-ups and normal operation some of the flooding-hard start issues in the RX8 are attributable to leaking coils. My solution is to fabricate an insulator (picture) to further isolate coils from ground (plate) reducing the chances of coil-arc. A piece of old plastic did the job here but ideally a piece of bakelite is best.

Coil isolator

I have removed the engine from my RX8 and have begun to dissemble it. Removing the engine without the transmission is no easy task and i think the automatic transmission model is a bit more difficult. My sole purpose for rebuilding this engine is to rectify the flooding-low compression issue the engine ran great otherwise with no smoking or any other abnormal symptoms. Thus far everything looks normal including the catalytic which (again) can be visually checked and shaken to see if there is any internal rattling in the honeycomb structure in the converter.