Date:         Fri, 6 Jun 1997 10:53:54 -0700
Sender:       Vanagon Mailing List <vanagon@vanagon.com>
From:         mholser@Adobe.COM (Malcolm Holser)
Subject:      Re: O2 sensor question

Well, the original question was whether the O2 sensor was accomodating
altitude changes, and basically, and indirectly it is, but this is not
the intention of an O2 sensor.

When they go open-loop, and no longer affect the mixture is not all that
simple -- they go open loop for a variety of reasons.  In open-loop mode
the imputs from the O2 sensor are not used at all, and I was saying that
this occurs more often than not at high altitudes (which you don't find,
IMHO, in North Carolina!).  However, after being called on this, I cannot
find the reference I recalled about an rpm limit causing open-loop operations.
I still believe there is one for Motronic systems, but VW's use something
that only resembles a Motronic -- and the data maps are reputed to be simpler.
It is certainly *very* possible I'm wrong.

The O2 sensor is used for closed-loop operation, the closed loop being that
the FI computer sees a lean mixture indication from the sensor, and enriches
the mixture (by increasing the open time of the injector).  The mixture becomes
rich, as evidenced by an increasing voltage on the sensor, and the computer
will lean it.  It then sees a lean mixture, and we are into a loop, back
where this paragraph started.

The system goes closed-loop during "normal" driving conditions.   It breaks
out of the loop, and uses other sensors at many times.  Yes, start-up and
warmup are two of the open-loop, non-"normal" conditions.  So is full throttle,
and this is what the full-throttle switch does.  So is full closed throttle.
So is throttle *movement* -- and for about 1 second after the throttle moves
(this is read by the change in flap position in the air-flow box).  There are
"maps" in the computer memory that are read to determine how long to open
the injectors for various conditions -- various readings on the many sensors.
One of the primary inputs to these maps is the engine rpm, and the maps
are designed to give the desired mixture for these conditions.  Much of the
range is "normal" driving, and the closed-loop operations are done, but
when the sensors indicate that you want to be in a mapped range (or when the
O2 sensor fails) then open-loop operations are done, and the amount of fuel
squirted in is based on the data maps.  A richer mixture is desired for
higher rpms than you get from closed-loop operations.  This was what led me to
believe that there was an rpm limit above which the system goes open-loop.
(and that I *thought* I read this somewhere that I can't find now!).

Regardless, I'm at full throttle quite often at high altitudes, which get to 
10,000 feet near me -- I need to be to get the box over the hills here.  Full
throttle is open loop always.

Someday, maybe I'll actually dive into the map data.  Trying to
do this as black-box analysis is real tough.

malcolm

> The O2 sensor is in closed loop while in startup and warmup, then it goes 
> to open loop after the sensor is at operating temperature.
> 
> Erik Peterson
> Vehicle Research Institute
> 
Backwards, and the primary reason for open-loop at warmup for systems with
heated sensors is not to wait for the sensor to get to temperature -- they
do this *real* fast. 

> On Fri, 6 Jun 1997, kenneth d lewis wrote:
> 
> > Malcom:
> > I beg to differ. I have a dash mounted LED (O2 sensor) monitor. I have
> > NEVER seen the system go into open loop once it reaches operating
> > temperature. In fact the higher the RPMs the faster the lean/rich cycles
> > occur.

Could be I'm wrong.  But it really should go open-loop during full-throttle
at the very minimum, and should be open loop at idle (when the throttle is
fully closed) at least on a Digifant system.  It should go open-loop as
well for a short time during partial-throttle acceleration, but this may be
hard to see.  The fact that the O2 sensor is fluctuating is merely showing
that the mixture is right, not that it is necessarily in closed-loop mode.
I would expect that you might see a pattern to the cycles in closed-loop
mode, but would have thought that they would be too fast to really see.  
Even on a full open-loop system, like the non-Calif Digijet's you can use
an O2 sensor to set the proper mixture by looking for these fluctuations.
That is when you are right.  A stable voltage reading on an O2 sensor only
indicates that the mixture is too lean or too rich.  A stable reading of
.45 volts means you have a bad sensor.  The computer puts .45 voltage on
the sensor itself, and it it reads this it knows the sensor is bad, since
a good sensor won't ever *stay* at this voltage.