>>
An ECU that was better at self diagnoses would have greatly reduced your risk.
On your car, it would be recognized immediately.
I know I’ve taken a
little flack for my MicroTech computer…but on the plus side it did tell me a
couple of months ago that I had a “bad Crank Sensor”. Of course, it
turned out to be a bad connector - the Mazda factory pig tail coming out of
the sensor had a bad connection in the plug. It must have been
intermittent because the engine still ran fine for the most part – just a
little rough every now and again.
Good of you to share this Joe. Can save a life.
So here is what problem solving is all about.......but just a second, some
of you are going to have a hard time entertaining what follows. If so, just hit
delete, have a nice day. This is for the guys that responded to me
privately..and others interested.
A) When we first heard of Johns failure, we need to force ourselves to
entertain the idea that he crashed. Under slightly different circumstances it
would have happened. We do this to eliminate our natural tendency to be
complacent, to rationalize.
B) We force ourselves to seek the OTHER causes. Because often it's the
other causes that can lead us to lasting permanent solutions. I'll explain this
better below. So the other causes I can think of are the 5 items I originally
listed.
C) We look outside our circle. We ask:"Well, how do the automotive guys
handle this type of failure". "How do guys outside of the rotary group do it?
Etc. Now one of those solutions is absolutely PROFOUND. It has far reaching
implications that greatly reduces risk of many historical failures.
Let's back up a little bit. Johns ECU saw the time interval between crank
pulses drop in half. Meaning the rpm went from 6000 to 3000 in 100 ms.
Physically impossible. Only a crank sensor fault can cause this. Realistically,
the ECU probably saw this weeks before his incident. Maybe hours before. But
it's the nature of most wiring faults to occur that way. Not all of them, one of
the crank sensor failures resulted in crash with almost no warning.
There is a huge difference in flight risk if the ECU sees this fault, tells
you immediately. You land, you focus on the real cause, you don't have to think
"Maybe it's my muffler or mixture that gave me that little pop. Maybe it's a
spark plug".
A key universal truth, is that risk is always reduced if you shorten the
delay in notification. Each day Johns connection got a little worse and there
was no way for him to know.
So we decide "I want the ECU to be smart enough to tell me when the crank
sensor develops blip." Our risk is reduced substantially. Then we think about it
some more. Crap, you know that guy that had almost crashed a few months ago due
to loose buss connection? That was same failure pattern. His airplane likely saw
a momentary voltage drop minutes or hours before he headed to airport. So we
say:"Hey, I want my EIS to respond to dips in voltage, so I get advance notice
of those faults too.
So you put this solution in place, and you have made every plane less
risky. Even though some of the wiring faults will still occur. You've now
reduced flight risk.
-al wick
Artificial intelligence in cockpit, Cozy IV powered by
stock Subaru 2.5
N9032U 200+ hours on engine/airframe from Portland,
Oregon
Prop construct, Subaru install, Risk assessment, Glass panel design
info:
http://www.maddyhome.com/canardpages/pages/alwick/index.html
|