Date: Tue, 6 Mar 2001 13:57:58 -0500
Reply-To: David Beierl <dbeierl@ATTGLOBAL.NET>
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: David Beierl <dbeierl@ATTGLOBAL.NET>
Subject: Re: O2 sensors effect on power
In-Reply-To: <3AA51CBD.B8FDD14E@islandnet.com>
Content-Type: text/plain; charset="us-ascii"; format=flowed
At 12:22 PM 3/6/2001, Mark Keller wrote:
>The O2 sensor is only used on engines with catalytic convertors. The
>pre cat Bosch fuel injection system is quite capable of giving accurate
>fuel mixture in operational service. The O2 sensor's main function, IMO
>is to safeguard a too rich mixture and thus a cat meltdown.
Actually its major function is to allow precise enough control of mixture
so that a three-way catalytic converter will operate properly and produce
minimum emissions. The three-way cat takes oxygen from oxides of nitrogen,
reducing them to nitrogen; and gives that oxygen to the hydrocarbons and
carbon monoxide, oxidizing them to carbon dioxide and water. In order to
have the available oxygen from the NOx match the needed oxygen for the HC
and CO, the mixture must be almost exactly at the theoretically correct
amount (the "stoichiometric mixture," the mixture where the oxygen in the
air mass is exactly enough to oxidize the fuel charge). This amount,
designated as lambda = 1, is a ratio of almost exactly 14.7 air to one
fuel. This is the *required* mixture for the cat to operate correctly. If
the mixture is the slightest bit lean the NOx rises dramatically; if it is
rich the CO and HC output rise on a steady slope.
The mixture for maximum power is 12.6:1, or lambda = 0.86; the mixture for
maximum economy is 15.4:1, or lambda = 1.05. Both of these values are
outside of the practical measuring range of the standard O2 sensor in an
automotive environment -- in fact in those conditions the sensor really
only provides crossover data for the FI system. The grounding and
shielding requirements to get numerical mixture info would be extreme --
and are unnecessary, since the FI system is perfectly satisfied to know
only where the crossover is.
>IMO the reason a the vanagon horsepower rating drops from 95 to 90 is
>that the O2 sensor loop or rich lean cycling means that the engine is
>running lean half of the time.
No, it's running lean all the time if you want max power, and rich all the
time if you want max economy. The control loop of the FI never gets even
close to either of those points.
>The rich side of the loop is at .75
>volts, while max power is at .84 volts of O2 sensor output.
This is unmeasurable under the practical conditions of the system. The
output of the sensor for these purposes may be considered to be a step
function; the useful information consists of the change from high to low
voltage, not the actual value of the voltage. O2 sensors exist which
provide practical measuring ability, but they are very expensive and
completely unnecessary to the operation of the system (I understand that
Honda has an engine now that needs and uses a sensor that gives numerical
info for some range on the lean side of stoichiometric).
>It is impossible for a Digifant O2 sensored fuel injection system to
>provided more power than a non-senosred digijet, because the digifant
>will never set optimum power fuel mixture, while the digijet system can
>be set to deliver optimum fuel mixture all the time.
True except for one little detail: at full throttle both systems ignore the
lambda sensor and provide a programmed enrichment. The amount of the
enrichment could be reprogrammed in either one...
david
David Beierl - Providence, RI
http://pws.prserv.net/synergy/Vanagon/
'84 Westy "Dutiful Passage"
'85 GL "Poor Relation"