I bought the Car Chip which is a small microcomputer

with flash memory that you plug into your OBD II connector
under the dash.

http://www.davisnet.com/drive/products/carchip_products.asp

Here is some of the results of the five parameters
that can be chosen out of three dozen or so to be recorded
for up to 300 hours. I was interested in the fuel mixture
as a function of engine temperature. My theory is the
manufactures are richening the mixture when you start your
car to light off the catalytic converter resulting in excess fuel
consumption during start up. I am still collecting
data as in this case the engine did not start from cold
but was at 150 F.

I suspect this so called "trim" is a percentage above
and below 14.7:1 stoichiometric. For example + 12.5% above
would be 12.86:1 rich.... I think.

Notice the 2004 Honda Odyssey engine spends most of the time on
the lean side even at 65 MPH.

Paul Lamar ...No rotor no motor.

Wow! neat data, just what I need for the cheap EFI programming.
Thanks Paul.

Doc Custer
Retired Software Engineer and sometime Audiologist

More to come shortly.

OK Doc,

Here is the results of a short trip to the gas station.
Note the engine started at 60 F at stoichiometric. It then
went lean almost immediately. As I recall the RX8 engine
was stoichiometric 99 % of the time. Perhaps Honda is on
to something here. I won't know for sure until I use
the same data acq CarChip on the 2004 RX8.

BTW the percentage load is misleading. That is a percentage
of peek power at that modest RPM. For example 50% of 50 HP
at 2000 RPM is 25 HP. At least I think that is what it is
saying.
-- 
Paul Lamar ...No rotor no motor.

I am surprised that it starts at stoichiometric. I would have thought it
would be on the rich side. Shows what I know. :)

I am wondering how to get the initial MAP. Just thinking about it, seems to
me that until the engine fires, the O2 sensor is going to indicate too lean,
i.e., lots of O2 in the exhaust. That means we can't dynamically build the
start MAP, but have to have a computed solution not based on the O2 sensor
output.

I am also guessing that the engine computer has a separate table for start
up until the engine fires.

Right now I am thinking that an aircraft engine with a dribble injection
system is going to have a problem with fuel in the line between the pump and
the injector if you just stop the PWM fuel pump. I think we are going to
have to have a shut down regime use up or otherwise deal with the fuel in
the line. That is standard procedure: pull the mixture control all the way
out and don't kill the mags until the engine dies. So either the mixture
control on the EFI has to have a shut off position or we need a separate
control to tell the uC we are shutting down. Also we need a reliable way to
purge all the gas out of the fuel line between the service fuel pump and the
injector; I hope this doesn't mean a purge pump.

So to summarize, it seems to me that the uC is going to have to have at
least 3 modes: start (open loop), run (closed loop), and shut down (activate
purge mechanism, whatever that turns out to be).

Am I making any sense?

Doc Custer
Retired Software Engineer and sometime Audiologist



-------------------------------------------------------

Let the pilot decide. Set the leaning knob straight up
on start up. Clockwise rich. Counter clockwise lean.
If it don't start the pilot can lean it
or richen it... his choice. That is what pilots do now.
They learn what their airplanes require to start as far
as prime is concerned.

Higher altitude airports require less fuel to start
perhaps, maybe, I think. You definitely need to lean it
to take off. I would not leave it up to the computer. Keep
the pilot in the loop just like what we have been flying for the
past 50 years. Open loop. Later on after wide and extensive
experience let the computer do more of the particularly
tedious jobs. Don't fix it until it is broke :)

In other words keep it simple.