[Supras] Wideband O2 sensor

[berniek at technicaldevelop.com]

Sean:

    Thanks for the information.  However, I spoke to a tech guy at 
Innovate about 6 months ago and was told that to leave the sensor 
in-place permanently would destroy it in 6 months at sustained highway 
speeds each day if mounted in the turbo elbow.  So I cautiously 
approached going up in boost using the A/F thermal conductivity meters 
at first, starting at 6 PSI at first (stock boost) and ending at 15-17 
PSI.  The Dawes device was gradually relied upon only after it 
demonstrated continued repeatability with respect to either the 
Marquette or Sun thermal conductivity A/F meters.  Granted, it is not 
the best choice since it works over a very narrow range at the very high 
end of the stock sensor output.  So I will keep a cautious eye on it and 
will back it up with checks with the exhaust gas analyzers as 
necessary.  Yes, I would rather have a wideband, but I'd also like to 
have something permanent and repeatable to assure that the mixture under 
boost is sufficiently rich. 


    There is another potential problem with using the second output of 
the LC-1 in place of the stock sensor, if the LC-1 is located just 
upstream of the cat, and that is the lag time introduced by the 4-5 feet 
of downpipe.  Have you ever noticed a type of "fast chatter" under 
normal driving, which makes the car want to buck slightly?  That is due 
to the fact that the stock sensor goes through the stoichiometric range 
very quickly, which is why its normal for its voltage output to 
oscillate during normal closed loop operation.  The closed feedback loop 
really cannot be made stable due to this, despite the fact that it is 
called "closed loop".  This happens mostly just as the sensor heats up, 
and then reduces in effect.  The effect is made worse when using larger 
than stock injectors due to a larger increment of fuel resulting from 
the ECU changing the pulse width in accordance with what it expects from 
stock injectors.  In control loop lingo, the so-called "loop gain" has 
been increased by using larger injectors. 


    Using the second analog output of the LC-1 so far from the engine 
would only make the instability problem worse yet due to introduction of 
a lag.   The stock sensor is very fast, as can be verified by the 
"propane torch" test.  If I were to use a wideband at the cat, I'd still 
use the stock sensor to feed the ECU due to its proximity to the 
engine.  Further, the inspection Nazis here in NJ look under the car 
with mirrors to verify the integrity of the exhaust system.  It will be 
bad enough having only one cat if they are knowledgeable enough to know 
that the car should have a precat also.  If they see something strange, 
like a bung and sensor just before the cat, the "tamper evident" box on 
the form is likely to be checked, one thing I want to avoid like the 
plague. 


    Just a moment ago I went to the Innovate website and reviewed one of 
the videos about the LC-1.  The point was made that the maximum 
temperature to which the sensor should be exposed is 1300F, and the body 
should not be exposed to more than 900F.  Further points were made that 
the best place to install the sensor is just before the cat, and the 
worst place was at the turbo outlet.  Innovate even sells a heat sink 
for the sensor body, but installing it in the stock location on the 
turbo elbow is still a very hot place under boost conditions. 


    How long have you had your LC-1?  I believe you indicated that it is 
in the stock location.  How often do you find it necessary to do a 
calibration?  I'm asking not to be a wise guy, but because of what I 
understand from the manufacturer vs. the way in which it is located. 


       Bernie