Old message:
Another good one.
Subject: Knock sensors and 13b REW notes:
Date: Thu, 01 Jul 1999 08:56:45 -0700
From: Paul <rotaryeng@earthlink.net>
To: zz <rotaryeng@earthlink.net>
David Carter wrote:
Paul wrote:
Turrbie@aol.com wrote:
snip
David,
I would be happy to send you a knock sensor or two, but If you plan to use
them on an Escort, they will not be optimized for the engine's knock
frequency. I believe that the knock sensor is just a big quartz crystal. When
it experiences the frequencies associated with detonation, it generates an AC
voltage. Kind of like the quartz sparkers used to start outdoor grills or
pocket lighters. The crystal is manufactured to a certain dimension so that
it gives a strong signal when spark knock occurs. Different engine blocks
resonate and propagate the energy form spark knock at slightly different
frequencies. Most knock sensors are tuned to the specific engine in which
they are to be used. I would expect that these knock sensors would work in
another engine design, but not as well as a sensor tuned for that engine. I
can round up several different sensors from Toyota and Nissan four bangers.
Since the block design is similar to the Ford 1.9 (which I think is a Mazda
engine, not sure) you may have better results.
Bruce Turrentine
Bruce,
Thanks. Before I say I want 2 - or only 1 - of the knock sensors to go
on my RV-6's 13B, I'd like to discuss and ask for feedback on a couple
of points:
First, the immediate reaction anyone should have in a car or in an
airplane, is, NOTE THE Manifold Pressure & simultaneously RETARD THE
THROTTLE until pinging (or worse) stops. No delays.
This might be a problem in an airplane as you are likely to be at
take off power and max engine heating due to slow air speed. You could
be heading towards trees or a hill so then would be a bad time
to retard the throttle. You might want to sacrifice the engine
to save your life.
Very mild pinging
maybe could be allowed to go on for a few seconds (3 or 4) while you
play with the throttle or try to tune your old tinitis afflicted ears
(like in my Ford right now if I were to use 87 octane and tried to
accelerate at 65mph, for example, and the wife says, "Can't you hear
that thing pinging?!")
What a coincidense. I have the same problem with my Ford Cougar and
my wife says the same thing. :-) If you have EFI on there try cleaning
the hot wire in the MAF sensor.
- but I don't believe there is any safe time to
allow pinging. As for Detonation: By definition, we are talking about
DESTRUCTIVE power blasts that IMMEDIATELY starts to damage something
(pistons and valves in piston engines, apex seals in 13B). I've never
had detonation in a car, as far as I know - only pinging.
Detonation is a function of a lot of things, but at the time it happens
in real operations, there is only one variable that the pilot has any
control over: Manifold Pressure (throttle setting). He can't change
spark timing, octane, compression ratio, or rpm (except by first
throttling back, which in itself will stop the knock).
So, the point of this "first" topic is: Maybe I don't want to have two
sensors, one per rotor, because I don't want to re-advance the throttle
back into the pinging or detonation range to do audio diagnosis -
because airborne is no place to be doing engine analysis in that detail
- I already know that I'm too lean and/or too far advanced on spark to
be able to use a the manifold pressure where the ping/det started, given
that day's temperature and humidity and the fuel I was using.
You should use the sensors to set up the engine in safe conditions
perhaps at pattern altitude or above so you can retard the throttle
and still land. Tracy's and other EFI systems allows tuning while
in the air. Start with very conservative boost and timing when
you first fly the airplane. Sneak up on everything else.
You should also note the intake manifold temperature when pinging
or detonation occures. This is why effective intercoolers are
essential. You could also use the sensors when running
the engine up on the ground of course.
Second point: If I had 2, and if I wanted to intentionally re-advance
the throttle to re-initiate this destructive condition, I don't know
that I could differentiate which of the two rotors (or if both) were
pinging or detonating. I believe the detonation wave (a high
amplitude, 20,000 feet per second pressure wave) in one rotor housing
would be picked up through the metal of the engine in both sensors,
without significant difference in amplitude.
That is a good question. But with late engines there are bosses
on each rotor housing adjacent to the combustion chamber. Surely
there would be an amplitude difference.
- We are talking about audio discrimination of WHICH rotor is
pinging/detonating - something that has probably never been done. If
the loud detonation wave is picked up nearly equally by both sensors,
then I don't need 2. Given that, I don't need the extra complexity in
the knock sensor circuit.
The cost of these is so small that it would not pay to make a PC
board with only one circuit. It is as cheap to make three as it is
to make one. If you are hand wiring that is another matter.
The cost of the sensors themselves is also another matter.
I throw these thoughts on the table for feedback from others with
additional experience, and for Paul Lamar, to perhaps hold off a day or
two before freezing the design for the knock sensor audio circuit board.
David Carter
The AirCraft Rotary Engine NewsLetter. Powered by Linux.
http://home.earthlink.net/~rotaryeng/ http://www.linux.org