Paul Lamar Wrote:

I have maintained for years the way to control a cylinder or
a rotor is to use EGT for feed back. Tracy Crook says he can't
do it because the EGT thermocouples don't respond fast enough.
I have found smaller thermocouples that have much faster response
than your typical aircraft engine probe.
Paul Lamar

There are a lot of challenges here.  Here are some off the top of my head:

1.  EGT sensors have inherent accuracy limits.  2-5% comes to mind.
2.  Slow response as mentioned above.
3.  The EGT that gets sensed depends to a degree upon the flow characteristics that the probe is immersed in.  Things like mounting location, exhaust system design, RPM and MAP all affect this.
4.  The EGT that an engine will generate is also dependent upon a variety of factors, such as induction air temp, RPM, MAP, ignition timing, mixture, even engine wear.

It may be that there are too many variables affecting absolute EGT readings for it to be used to directly control mixture.  This is where an oxygen sensor shines (if only it could be used with leaded gas).  However, the point at which EGT peaks does correspond to a given fuel-air mixture.  This, of course, is the theory behind manual mixture control - control by reference to an EGT peak, not by reference to an absolute EGT temperature.  The peak could be found electronically by controlling mixture change such that the EGT rate of change is slower than the probes' response time.

Seems like a workable control scheme would be to schedule fuel injection based upon an RPM/MAP curve during throttle transitions, but to have some method for adjusting mixture to peak EGT (+ of - some given amount) during constant conditions.  EGT peaks found during constant conditions could then be fed back to modify the RPM/MAP curve.

The potential for EFI, as opposed to mechanical fuel injection,
applied to an air cooled  piston aircraft engine is great.
With the proper thermocouples each cylinder could be controlled
individually. This would be particularly important in a two cycle
engine to prevent piston seizures. So far, to my knowledge,
nobody has done it yet.

I wouldn't limit it to air-cooled piston engines.  Applies just as well other engine types (even rotaries).

You are in a position to do it Rick. Why don't you give it
a try for the experimental market?

Time.  (sigh)
Rick

The AirCraft Rotary Engine NewsLetter.              Powered by Linux.
ACRE NL web site.
http://home.earthlink.net/~rotaryeng/            http://www.linux.org