I had an idea the other day. The RX8 computer keeps the engine
        at stoichiometric mixture ratio at all times according to the
   OBD II
        data
        I have captured so far on my RX8. Even under full throttle.
        Stoichiometric
        is 14.7:1 to one as I recall so that is not the best for best
   power.
        Best power happens at 12.5:1. Cheap way around this is to
        fool the stock computer into thinking it is getting more air
   than it
        really is.
        Consequently it will put out more fuel richening the mixture.

        The idea is you program a $50 PICAXE computer to take data in from
        the mass airflow
        sensor and the throttle position sensor and modify the mass air
   flow
        data going
        to the stock computer when the throttle is wide open. This device
        could be installed by disconnecting the mass air flow sensor
   and the
        throttle position sensor and plugging them into the PICAXE. Two
   cables
        coming out of the PICAXE could then be plugged back into the
        sensors. This way
        the standard RX8 engine configuration can be easily and quickly
        restored by doing the
        install in reverse.

        On the other hand when the throttle is partially open during
   highway
        cruise
        you could lean the mixture to get better MPG. The rotary will
   run at
        mixture ratios
        of up to 17:1 or higher at steady speed low load freeway travel. I
        don't know exactly
        what this will do to the catalysts so do this at your own risk.

        --
        Paul Lamar ...No rotor no motor.

   Hi Paul,

        Sorry, but at full throttle/high rpm - or any time you are off
   closed loop - the mixture is not 14.7.  Stock typical ratio at high rpm
   is ~11:1 to 13:1, sometimes less.  Even in closed loop, the target
   mixture is mapped and varies up and down from stoic.  We tried hard and
   could not improve the mix in feedback from stock.  Above that, we did
   make some improvement (~5% max), and that is what we deliver in the
   reflash service we now offer.

   Jim Mederer

   Ops you are right. I forgot I took this acceleration run in 2004. I
   was wrong.
   Here is a screen capture. Yep Lambda goes rich.  I was looking at a
   lot of
   steady speed data.

   BTW note how much higher temperature the intake system is than ambient
   air temp. Looks like it needs some insulation.

   Well then only half of my half ass idea will work. The lean part. I
   have never seen
   Lambda go much over 1.02 on the RX8 which is about 15:1. I have seen it
   go well over 15:1 on the Honda Odyssey. I think you could lean
   it on the freeway to 17:1 or 18:1.

   Most of this data is for a piston engine
   but we have found the rotary will run much leaner. A lot of our guys
   are cruising
   their engines at 17:1 or higher.  My guess is.... it is
   naturally stratified charge due to the rotating combustion chamber.

   Of course it will never be as good as a turbo compound rotary.

   The advantage of the PICAXE over the reflash is it can be
   reinstalled cheaply and quickly.

   Paul Lamar ...No rotor no motor.


Stoich has to be nearly the highest EGT, and hence the hardest on the
cat(?)..  I guess richer allows the cat to clean up the HC and yet still
stay cool, while lean may cause excess NOx which the cat can't do
anything about..

Then again, reading up on NOx production, it appears that Honda has a
method for dealing with Lean Burn emissions - as maybe you have seen Paul:

http://www.insightcentral.net/encyclopedia/encatalytic.html


Another interesting article (though somewhat biased):

http://en.wikipedia.org/wiki/Lean_burn
<http://en.wikipedia.org/wiki/Lean_burn>


Matt-

Yes I noticed that Honda stuff on nox. That should help
the rotary if implemented.