At last I am back in the air. Thank you to all of you who responded to
my question about possible broken wires in the ignition system: I did
not find any. The culprits were the electronic modules which have aged
to the point were they are unable to produce reliable sparks at normal
cranking speed. I replaced one of them (cost me 928 Euros) and the
engine starts very well now.
To summarize the history of this problem:
- the symptom was: impossible to start engine in spite of a normally
charged battery. It happens once in August last year and twice in the
past few weeks. There were no problem after the engine was finally
started, never an ignition miss during run-up.
- I checked the carbs first and found nothing wrong. I then suspected
the ignition
- I installed dummy spark plugs to allow to visualize the sparks, and
a counter to mesure crankshaft rpm when driven by the starter motor.
- I confirmed that most of the time there were no spark at normal
cranking speed (I was able to vary crankshaft speed from 270 to 460
rpm. ) I noticed that the higher the cranking speed, the better the
probability to get sparks, but this was not repeatable all the time.
According to the Rotax manuals, the ignition cut-in speed is 150 to
220 crankshaft rpm. Most of the times I had no sparks below about 400
rpm.
- I made all the measurements and checks as suggested in the Rotax
Heavy Maintenance Manual and found nothing wrong.
- I then replaced one electronic module with a new one and noticed
that it was able to produce seemingly more powerful sparks even at the
lowest cranking speeds. The other module was still unable to spark.
Permutation of the two modules confirmed that the new one was always
producing sparks while the other one was not.
- subsequent engine tests showed confirmed that starting was ok and
both ignition systems were fine when the engine is running. Moreover,
the bad ignition module is still able to start the engine most of the
time.
One benefit of this mess: now I have the Soft Start system on both
channels as Rotax has it incorporated into the new module.
Some lessons learned or re-learned:
- do not believe that a redundant system cannot fail completely: in
this instance it did. However there were no safety issue as the
failure was effective during starting only.
- electronics failures are intermittent most of the time and therefore
difficult to troubleshoot.
- the lack of precise data and testing procedures from Rotax make
troubleshooting more difficult: for example what voltage the charging
coils should deliver vs crankshaft speed? I was able to measure this
voltage but had no reference to compare with, so there were no way to
be sure that the stator was performing ok.
- it is easy to measure cranking speed using a 15 Euro bicycle speed
counter. With the Airmaster prop, you install the sensor on the brush
bracket and sense the magnet included in the prop flange. The right
programming of the unit allows to read directly the crankshaft speed.
- cranking the engine with one spark plug removed from each cylinder
is a bad idea: as there is no blow-by anymore, the oil pumped into the
engine is not returned to the tank and you may end up with a full
crankcase and an empty tank.
The Rotax is ok when it works. When it doesn't it's a real nightmare
to work on. It really is not design with maintenance in mind. If I had
to pay somebody to do do the maintenance it would be tremendously
expensive. Anyway it is more expensive to maintain than a Lyconental.
Remi Guerner
F-PGKL, 830 hours
PS: now I have the alternator red light coming occasionally. I have
ordered a new regulator today (just another 110 Euros!)
|