ox sensor on sequential efi,high overlap cam.... a bit lo ng

[Greg Hermann]

>The WWII FI birds (which often had some pretty wild camming to get the power
>required) and supercharging, EGT was the indicator of choice for mixture
>control but no one worried about it til you got up and pretty well
>stabilized in cruise (or for the multiengine planes with a flight engineer,
>had the engine up in the power range).   The wild cam basically says the
>designer (and the user) don't really care about low end quality and are
>willing to sacrifice it for upper rpm maximization of HP!
>
>Lets try to little mental game playing here.  First assume it is possible to
>control mixture with high overlap and low rpm.
>If the engine is really setup well and the cam isn't too wild, you can
>usually get them to idle fairly well.  The secret to getting efi to really
>control A/F under these circumstances would be to at least get the engine to
>stumble and blow raw fuel into the exhaust fairly consistently.  If the
>anomalies approaches any type of periodicity, then you can predict the next
>event with a definable level of certainty (and write code to correct for
>it).
>
>A computer with enough overhead to measure engine speed (and therefore
>detect surges and power drop offs) every 5-6 degrees like the EFI332 project
>does would be necessary.  But it would also have to look at O2, EGT and MAP
>or MAF (to detect regurge backing up in the intake) and the system would
>have to have a separate algorithm to handle this (out of map) situation.
>Even at that, to get any kind of consistency would probably require multiple
>O2 sensors (EGT isn't going to be very useful as most of the fuel isn't
>getting burned in this situation!) and an averaging algorithm to come up
>with an "average" A/F and compare this to A/F at specific points in the
>exhaust header (to detect rich slugs in the exhaust).  This information
>could then be used to measure the amount of correction necessary in the next
>event and when the correction needed to be made prior to the next event.
>
>Rick
>
Which brings me to the point of saying: MAYBE we should take a hard look at
adding a whole lot more I/O capability to the 332 project--for starters 8
Thermocouple inputs for EGT--

Regards, Greg