Zirconia O2 sensor alternate modes of operation

[Jennifer and Brock Fraser]

>From: "Zack" <zubenubi at inetport.com>
>Date: Sat, 21 Feb 1998 08:28:12 -0600
>Subject: Zirconia O2 sensor alternate modes of operation
>
>All,
>
>I've been reading some SAE papers about O2 sensors, and a couple of them
>mention biasing a zirconia-type sensor with a low voltage and measuring the
>output current, rather than measuring the open circuit voltage as is
>usually done.
>	The response of the sensor to A/F ratio in this regime seems much more
>linear, and thus more suitable to measuring A/F than the usual open circuit
>method (which works extremely well for maintaining lambda = 1 but is not
>very good for measuring lambda at points above or below).
>	My question is this, none of these articles make any specific mention of
>what type of sensor is being used, though they're mostly written by people
>working for research labs of large corporations, so I'm assuming they're
>probably using one-off probes that they've designed with an eye towards
>this method of operation. 
>	So... my question is, for anyone who might now, can any garden variety
>heated zirconia sensor be used in this manner??  Will it be damaged by the
>current (small but not negligible) that flows through the zirconia
>electrolyte when used like this?  What would the usable measurement range
>be?


Your question relates to one that I posted a few days ago, but nobody
responded to.  The type of sensors that you are reading about are very
different than the garden-variety (one, two, three, or four wire) ones
typically used on automotive applications.  Wide-range sensors of the
current-pumping type are very expensive (as compared to standard sensors)
and are currently only used on applications using alternative fuels or
lean-burn techniques - or other specialized applications that can make use
of running in closed loop while outside the stoichiometric region.

The major help is that Honda is now beginning to use it in more and more
applications (more than just the NSX, now) and prices will start to
decline.  I heard that the sensor now (for '98 ULEV California Accords) is
only $170-ish at the dealerships.

My previous question related to "using" these sensors for tuning.
Typically, only Bosch and NGK made "readers" that were compatible with the
sensors.  Unfortunately, it isn't nearly as straightforward as just reading
a voltage with a multi-meter or such, as with typical sensors.  The
technology must be available, but for now - most companies that use the
sensors purchase daughterboards that "read" the signal and then passes the
data to the ECM or AFR meter or whatever else is doing the data
proccessing/display.  

The only commercially available solution I have heard of is an interface
box that converts the signal output to some voltage that is linear with
AFR.  This is sold by NGK for upwards of $400-500.

The usable measurement range will go down all the way to 9.5:1 AFR, and up
to 17.1 lean, at least on the systems I have used.

Anybody found a cheap way to use this technology in your garage?

-Brock