> Perry,
    >
    > Do you have a turbocharged engine or NA?
    >
    > Doug Fir

   It's stock 1986 NA.
   In the early days (1999) I was worried about the EFI not being a
   redundant system, but it's reliability is now proven (it was "proven"
   anyway in car use). Redundancy is only needed for failure prone systems,
   such as magnetos.
   The only redundancy I have is a dual battery system, independently
   selectable but I always run "both". Two sealed r-g lead-acid batteries.
   I've never had a battery failure either, I replaced the originals after
   5 years, running on the second set now.

   Perry

   I also might add, Perry is a very talented professional electronic
   engineer so EFI holds no mysteries for him.

   Now on the other hand (correct me if I am wrong Perry) he still has
   not got it to lean the engine manually :)

   I still think you need to try a pot in the mass airflow sensor
   circuit Perry.
   --
   Paul Lamar ...No rotor no motor.



   I've designed avionics since 1980, mostly CRT displays and now LCD HUDs.
   Some digital, some software, but mostly analog and switching power
   supplies. EFI is not mysterious, however I have not dug into the
   internals of the Mazda computer. I just use it as is. I doubt there is
   any way to get schematics and source code from Mazda.

   Sometimes I commute to work by airplane and park at a guy's house at
   UAO. He is impressed that I just hop in, turn the switch, and takeoff,
   as if it were a car.

   For leaning, what I did was build a variable gain amplifier. The airflow
   sensor signal is intercepted and routed to a switch. With the switch in
   one position it sends the AFS signal to the ECU normally. In the other
   position it routes it to the variable gain amplifier and that output
   goes ot the ECU. For mixture I monitor the voltage on the O2 sensor. It
   does work, I used it on the round-trip to Kansas summer 2006. Being
   conservative, I made the adjustment range too narrow. I need to take it
   back out of the cockpit and make a slight adjustment to fix that.

   Perry

   Perry,

   have you ever determined if the Mazda ECM (computer) has a limp
   home mode? This can be a real attention getter if power or rpm is
   suddenly reduced due to a sensor over-run . Any other quirks that
   you have noticed or modified? Also, have you had any problems with
   the O/2 sensor using leaded fuel?

   Thanks, Joel



There is a recent Contact magazine article where some Subaru guys had
trouble with limp-home mode. The only sensor I've ever "lost" was intake
air temp due to some dirty contacts, it had no noticible effect on
engine operation. I have two LEDs in the cockpit, wired to the ECU fault
code discrete outputs. They flash a two digit code if a sensor fails.
The only sensor failure that guarantees the engine won't run at all is
the crank angle sensor. Failure of the other sensors could cause reduced
performance, the most critical after CAS would be the airflow sensor.
Throttle position sensor is only used at very low throttle settings.

The O2 sensor is not necessary for operation. Mine is disconnected from
the ECU and routed to a cockpit voltmeter. The ECU only runs closed loop
below 3000 RPM, but still runs fine open loop down to idle.
I'm not recommending the stock ECU for aviation use, just pointing out
that it does work and works very well. Since the RWS system is redundant
and designed for aviation use, it would still be the preferred setup. Up
until the recent runup in gas prices I didn't care about the rich fuel
burn because with mogas the hourly fuel cost is still much cheaper than
with avgas.

Perry


The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.