The intake period is 270 degrees of e-shaft rotation as are all other
  strokes. In a p-port there is a depression in the rotor face that
  connects the intake port with the exhaust port. Surprisingly the
  pressure in the intake is higher than the pressure in the exhaust.
  Therefor air flow, if any, goes from intake to exhaust in theory.
  However air has mass, friction, adhesion and inertia. So what
  goes where and how much is arguable.

  Here is a Racing Beat 3 rotor P-port turbo making 900 HP as I recall.
  Some of the end housing are aluminum.
  Note the size of the turbo charger :) One would think the injectors,
  being on the plenum side of the long intake runners, would lose
  fuel to the exhaust. Jim Medere told me that was not the case.
  In fact the BSFC was slightly better with the injectors at the start
  of the runners. I can't wait to see what happens when Dave Lemon
  builds a welded p-port with a slide throttle RX8 engine. Should be
  around 600 HP or more instead of only 420 HP :)

  As I recall the 3rd gen RX7 stock body land speed record at 268 MPH was
  set with a 3 rotor, 3 turbo side port engine making only about 750 HP.
  It had an ice water inter-cooler as I recall.

  BTW a picture taken from my C182 made the centerfold of Pop Sci
  mag the year of the flood at Bonneville. That was the first year the
  recording holding R.B. 3rd gen 3 rotor RX7 showed up.

  http://www.youtube.com/user/PaulLamar2#p/u/11/JUf_aWVVl7o

  With our man on here Csaba  Csere narrating no less.
  I got my 2 cents in at the end.

  Paul Lamar


  I have to wonder about the concern everyone has been showing about
  overlap.
  There is considerable overlap in any piston engine. The exhaust and
  intake
  valves are both open on even a low output piston engine. There is even
  more
  overlap in a high output engine. I haven't timed any cams lately but
iirc
  the cams have about 30 degrees of overlap. Due to the velocity and
  mass of
  the intake charge some will often flow out the exhaust. Since the
rotary
  doesn't have valves in the way timing is very important, but a lot
of the
  overlap events are similar.
  Bill Jepson


 My concern was in the volume of air that passed from the intake through
 the chamber and out the exhaust, and the amount of fuel in that flow. It
 appeared to me that if the injector timing could be manipulated so that
 at cruise power the air passing through to the exhaust carried the least
 possible amount of fuel, better efficiency would result.
 It may be that the volume lost is inconsequential, or that the timing
 cannot improve the loss - but since lean cruise fuel flow is less than
 half max power flow and therefore injectors are on less than half the
 time, it seems reasonable that some injection timings will be better
 than others.
 One reason direct injection engines can get better fuel mileage is that
 they lose no fuel out the exhaust during valve overlap.  It is also one
 reason that port injection engines can get better mileage than those
 with throttle -body injection.
 I ask the group's forgiveness for my intrusions - I know I have a gift
 for causing unintentional irritation.

 Victor Roberts


 You have a valid point there Victor. When the fuel is injected
relative to
 top dead center could be important.

 So far we have no control over that as Tracy does not publish the source
 code of his system. We think he injects fuel once a rev. I'll find out
 as soon as I get my engine instrumented with a data aq system.

 Paul Lamar


Easy to check the behavior of Tracy's system - scope one of the
injectors.

If he injects once per rev (which I think he must), and you want to
change
the timing of the injection, just put in a delay to the system trigger
Tracy's computer.  It's kind of a glue-on kludge solution, and the best
delay will vary with RPM, but it should still work fine.


Regards,

Matt-


It is not possible to kludge the injector timing on Tracy's computer, as
the injector and ignition timing are calculated from the same input. The
injector timing is whatever Tracy decided.

Victor Roberts