|
Hi Tim,
As we all know, all things break. Generally the
simpler the component the less likely hood of a failure - but still
possible.
The Crank angle sensor has:
1. Gear drive
2. Shaft
3. Two timing wheels
4. Two magnetic reluctance type coils.
5. Electrical Connector
It would appear that the areas most like to fail are (in
my opinion) in order of likelihood:
1. Electrical connector external to the crankangle
sensor housing (as related to the skill, knowledge and
craftsmanship involved in making the interface between the aircraft's
electrical system and the crankangle sensor).
2. Gear Drive
3. Shaft
4. Reluctance coil
5. Timing Wheels
The device is about as simple as you can get. The
coils have no wear (but could possible corrode), the wheels do not rub against
any component so no wear, so the highest likelihood of failure (integral to the
crank angle sensor) in my opinion would most likely be the drive
gear. External to the sensor the most likely
failure mode is the electrical connection. I personally use the Mazda
stock connectors and then added a strain relief strap (plastic electrical tie)
across the two connectors holding them together. If I need to take the
connection apart, I just cut the plastic strap and replace it when I
reconnect.
I personally have never heard of a crank angle sensor
itself failing - could certainly happen, but must be fairly rare. However,
it is a single point of failure so a candidate for redundancy.
But, in my opinion, there are so many other
components with a higher likely hood of failure - that I am unable to get
alarmed over the possible failure of the crankangle sensor. If someone
should develop a flywheel/pulley add-on sensor that can be integrated with the
EC2 for a "reasonable" price, I would be interested - otherwise
not.
Just my 0.02 worth
Ed
----- Original Message -----
Sent: Saturday, June 10, 2006 9:52
PM
Subject: [FlyRotary] Re: Ignition
Failure
Thanks Perry!
That was what I was looking for.
I didn't mention names because I didn't want
things to get personal, but I guess everything is out in the open now...
Every conversion I have seen so far has some
non-redundant items. One example on the Subaru side, I have seen
backup ignition controllers, but all the engines I have seen only have one
sparkplug per cylinder. Is redundancy needed there? I'm not
certain. Normally I would say no. But I just had a plug failure
strand me while riding my motorcycle. (First time ever for me... and no,
it's not a V-Twin) So now it raises a red flag... there have been
other red flags, but this was just the first thing to come to mind.
I do know that a 91 Porsche 911 engine had dual
plugs, dual distributor, dual coils. Maybe replace the distributors with
electronic ignition. Parts would be pricey though... typing
out loud again.
So now I have confirmation that a crank angle
sensor failure will shut down an engine. But it also appears that a
crank angle sensor failure is extremely rare (difficult to quantify, but there
seems to be a general consensus on this). Something else that would be
interesting to find out is if failure usually occurs during operation
(stress), or during idle time (corrosion, etc.). If the primary failure
point is before startup, then the risk is relatively benign (not happening in
the air).
I hope everyone here (and other mailing lists for
that matter) continue to report their successes and failures. I have
been following (and cheering on! :-) ) John Slade for a while now. I
just hope you don't give up on the A/C system John, now that you have
moved out of the tropics and north of the Mason-Dixon line!
Back to lurking!
-TP
----- Original Message -----
Sent: Saturday, June 10, 2006 11:49
AM
Subject: [FlyRotary] Re: Ignition
Failure
The crank angle sensor is a single point of failure. There is
only one, so it's not redundant even if you have two controllers. However
from my experience driving RX-7s (70 miles round trip to work for the last
10 years) I've never seen one fail. Haven't seen an ECU fail either.
Occassionally one of my cars wouldn't start, I would unplug and then reseat
the crank angle sensor connector and it would then start. I've only seen
this phenomenon on one car. Probably just needed some contact
cleaner.
Generally I think any problem with the crank angle sensor
would mean the engine wouldn't run at all.
Might be good to have one
controller running on the stock crank angle sensor and a second running from
an independent sensor on the e-shaft front pulley.
You can't have
redundancy everywhere. I have one cranks sensor and one ecu and my engine
has never quit running in 600 hours. If you really want redundancy build a
plane like a twin-engine Defiant, that can truly fly on one engine.
Perry
-----------------
I understand that it's totally natural to feel defensive in this
situation. I know you may find this hard to believe, but the goal of the
post is to help others reduce risk. It's not a personal attack. There's no
rotary reference in it anywhere. As far as the readers know it's a piston
engine. But I hope there's at least one person on this list that
sees value of looking at the other causes.
The essence of the post is accurate. Doesn't matter if you are near
home base or not. Ignition failure is very high risk item. An ECU that was
better at self diagnoses would have greatly reduced your risk. On your car,
it would be recognized immediately. Ecu would say "Hey, my cam sensor just
went thru it's normal 50 pulses per revolution with the normal 5 and 8 ms.
gaps. But I didn't get the normal 20 pulses and signal gaps from the
crank sensor. Turn on warning lamp and crank code. Use cam sensor for timing
info." You don't see value in discussing stuff like this? If you guys just
used toothed wheel with pulse gaps, then the ECU could easily self diagnose.
You know what was cool on the other list? Guys came back with
"confessions" of how this parallels a problem they encountered. 2 guys
said, "yeah, I had partial failure, thought it was x, took off only to find
it was y". Then other guys described changes they made to system that
totally eliminated the risk. Some really creative stuff. Then we discussed
just how risky crank sensor really is...we see one ever 1 1/2 to 2 years.
Etc. Etc. Very positive experience that may save someone's life. That's the
goal.
Hi Tim,
Isn't it
wonderful how stories get twisted around as they fly from list to list.
Obviously this particular twisted story relates to my recent experience, so
let's dilute the fun a little by adding some
truth.
|