Paul wrote:

Seeing this knock sensor on the Turbo II brought to mind
about what I was going to do about it. Seems to me one way
of approaching an automatic ignition retard is use a separate
PIC computer that passed through the ignition pulse if there
were no knock. As soon as knock was detected it would delay the
pulse by an amount depending on the RPM thereby retarding the ignition.

The analog circuit could be as simple as a comparator with a sensitivity
pot. You set the pot until an LED on the output of the comparator
just goes out under normal operation. If the voltage from the knock
sensor exceeded that threshold the dedicated PIC knock processor
would then delay the ignition pulse. Another output from the PIC
could open the waste gate to simultaneous reduce the boost.
This way Tracy would not have to reprogram his EFI computer
to add knock capabilities. There may be something like this on the
market already.

There is an issue however and that is; do you want a computer to
automatically  reduce the power just because the engine happens to knock
while executing a critical takeoff maneuver? Incidentally a highly likely time
for it to happen.

Another possibility is have the knock computer inject anti detonation fluid.
This option makes the most sense to me.

Paul,
It seems to me that one proper solution is reduce the intake charge temperature.
Since presumably the intercooler is already working at maximum capacity (for the
configuration at the moment), then the alternate methods are to reduce boost or to
cool the charge using "anti-detonation" fluid (ethanol?  water?) or fuel.

The drawbacks of reducing boost are (1) reduced power, perhaps dramatically which
is a "feature" of an automatic system that everyone agrees should be avoided, (2)
reliance on a mechanical linkage in an extreme heat area (wastegate), and (3)
exhaust plumbing design complexities.

If one were to use your comparator circuit's output as an input to Tracy's EFI,
and have the EFI enrichen the mixture at the onset of detonation, you haven't
added any more points of failure that aren't already present in any other
automatic system (knock sensor, knock signal conditioning, EFI computer).  Using a
wastegate or "anti-detonation" fluid adds more ways for the system to fail
(wastegate, wastegate actuator, anti-detonation fluid pump, anti-detonation fluid
reservoir empty, ...).

Other than configuring your installation to ensure that detonation cannot occur at
"max rich" settings, what have I overlooked/misunderstood?

Russell Kent

Nothing. No. You did a good job. I just thought of another input to the control 
system and that is the intake manifold temperature as you mentioned above. 
There is a strong correlation between intake manifold temp and knock. If that 
goes above a certain limit the mixture could be enriched or the ATF could
be injected if Tracy was not involved :-). One other thing. Enriching the mixture
is likely to increase the TIT on a rotary. Not a good thing.

Paul Lamar
 
The Aircraft Rotary Engine Newsletter.              Powered by Linux.

http://home.earthlink.net/~rotaryeng/            http://www.linux.org