Rotax 912 is misfiring, any ideas why?

[tigger_na]

Dear Forumites

I have a 980hr TTSN 912 UL in a new Hawk Arrow II. Something is seriously wrong with the installation, as per the RPM/EGT recordings attached. Right now I cannot rule out instrumentation errors, OTOH the EGT recordings are supported by physical evidence.

These are the symptoms:
1) For my first three flights, my cylinder 2 EGT is in the yellow (>840°C), dipping into the red (>880°C) occasionally. On at least two occasions when I throttled back slightly EGT2 shot into the red and then went back down to yellow. (Why does the EGT increase, even momentarily, with a decrease in engine speed?)
2) It feels like the engine is missing that final kick on full throttle, but I'm not used to it (this engine, this airplane), so maybe this is normal. I'm generally getting a rate of climb of about 400 fpm, loaded with say 220lbs pilot and fuel, from 4,300 AGL at say 27°C OAT. I expected more and this is borne out by other pilots on 912 powered Arrows.
3) The engine only revs to max 5,200 rpm at an IAS of 70 mph. I have a 2 bladed wooden prop, 73", 39° pitch, which was designed and supplied by Peter De Necker.
4) The spark plugs on cylinders 1 and 3 are sooty black, the spark plugs on cylinder 2 are more glossy black, the spark plugs on cylinder 4 look new. After 4 hours of mostly runway operation the insulator is as white as it left the factory!? The spark plugs are all new and I'm running unleaded fuel. (White means lean running, but cylinder three records normal EGT on full throttle. The EGT instrumentation channels haven't been mixed up, I checked)
5) With the spark plug cap removed, all 8 ignition circuits have a strong blue spark. I didn't check the timing.
6) The aircraft vibrates after throttling back on high finals and I felt the rpms aren't quite stable, but I thought they varied due to turbulence and the resulting changes in airspeed. If I look at the temperature records, I now know that cylinders 2 and 4 are basically switching off on throttling back. (Why would they only misfire only when I throttle back?)
7) Ignition test rpm drops @ 4,000rpm: 1 = 250rpm, 2 = 150rpm. (Within limits, but still high)
8 ) Fuel flow maxes out at about 18l/h, pressure at 0.4bar. I feel the fuel flow is low for full throttle, but I haven't calibrated the system yet, so the flow reading may be off.
9) Carburettor synchronization was done, but turning the idle mixture screw didn't really change the engine speed, so I left it at 1.5 turns out. The engine sounded normal during full power tests and synchronization.

My feeling is that I should look at the carburettor, because it is common to cylinders 2 and 4. All four spark plugs on cyl 2 & 4 are powered by four different coils, timing signals are shared with cyl 1 and 3, and these are running normally. But then why are the spark plugs on cyl 4 white as snow and cyl 2 black? How can cyl 4's EGT mostly follow those of cyl 1/3 when it's plugs are still white? If the carburettor is the culprit, why only on throttling back?

Back to you...

[justin.schoeman]

Could be an induction air leak on the #2 cylinder?

[tigger_na]

Yes/No. Checking for an air leak is my next step, but I wonder why this would only happen on throttling back? O, btw, my Odyssey MAP sensor is connected on this carburettor and the MAP readings have always behaved completely erratically. I always ASSumed this to be an instrument/sensor error...

[justin.schoeman]

Yes/No. Checking for an air leak is my next step, but I wonder why this would only happen on throttling back? O, btw, my Odyssey MAP sensor is connected on this carburettor and the MAP readings have always behaved completely erratically. I always ASSumed this to be an instrument/sensor error...

Manifold vacuum is higher, the lower the power - so it sucks more air through the leak. At full power, the carb is fully open, and the manifold is hopefully only about 0.5" or less below ambient - there is almost no pressure differential to suck air in (unless the leak is HUGE).

[KFA]

Whithout having seen the engine and only looking at the data it could be a leak in the manifold or an electrical proplem. I noted that when the the rpm at 12:27:22 drops the mixture gets richer in No 2 only and not in the rest telling me that it is possible that the cylinder was not sparking at that moment causing the rpm drop as well as causing a rich mixture reading on the EGT. Maybe a loose connection or plug problem. Also interesting to note is that the 2 cylinders furthest from the carbs are running slightly leaner than the ones close to the carbs.

There could also be a cable/link problem to cylinders no 2/4 if you look at the sudden drop in the EGT when the rpm is reduced.