O2 sensor response times

Garfield Willis garwillis at msn.com
Mon Mar 27 23:32:49 GMT 2000


On Mon, 27 Mar 2000 16:29:44 -0600, Gary Peyton <gpeyton at sws.uiuc.edu>
wrote:

>I had read somewhere that the response time of the platinized zirconia
>sensors was 2-3 seconds, and that was one reason why they werent very good
>for close engine management during acceleration (along with the way the
>L-jetronic system used that information).

Ahh, this allows me to clear up a possible mistake I made a few posts
ago when Mike "tah" M. asked about a "platinum O2 sensor". AFAIK, the
platinum is just used to provide the metallic connection between the
ZrO2 material and wiring to the sensor. It's probly vacuum-deposited
onto the outsides of the layers of ZrO2 that make up a sensor, giving a
metallic area that can be wired to. Don't quote me, I'm not a physical
chemist; there might indeed be other electro-chemical reasons as well
why platinum is the material of choice, but it appears from diagrams on
how most all ZrO2 sensors are built, that the platinum IS the
metallic/electrical connection to the ZrO2 substrate. In the real world,
you're likely to see anything from 2Hz to 0.5Hz stoich-crossing
frequency. 3 seconds is getting on the slow side.

>Maybe I'm suffering under a misconception here, since the www.tech2tech.net
>discussion says that "A good O2 sensor will usually module several cycles
>per second.". Are they referring to a simple one-wire zirconia sensor?

Yeah, sorta. What's in view there is the overall cycle-time of the
control loop in closed-loop mode. From the time the controller sees the
measured mixture is lean of stoich, till it increases the injector
pulse-widths, till the time the mixture passes from lean thru stoich to
rich of stoich, and the sensor now says "we're rich", and the cycle
starts over again. Ordinary switch-type O2 sensors/controllers all work
this way, be they one, two, or heated 4-wire (there are some wideband
4-wire sensors, but I'm of course leaving them out of the switch-type
category). Obviously one of the parameters in the phase-lag of this
feedback loop is how far from the combustion chamber in *time*, the O2
sensor actually sits. Transport delay of the gases to the sensor, has to
be added to the sensor and controller delays.

Also, one of the key aging/contamination phenom of all these sensors is
that they become sluggish in their response to changes in exh gas. This
will slow down cycle-time of the controller also, of course, because it
will take longer for the sensor to SEE the mixture change that the
injectors are providing, the slower the sensor is to respond. He's just
giving a ruleOthumb for these stoich-crossing type controllers. BUT even
in controllers based on wideband sensors, which ARE capable of holding
the mixture right at stoich (or any other commanded AFR) without any
significant excursions lean or rich, AFR swings MAY be programmed into
the controller *anyway*, to provide the cat with needed lean & rich
periods, depending on what type of cat is used in the design of the exh
system. This is explained on pg. 656-657 of Heywood's "Internal
Combustion Engine Fundamentals". [BTW, especially if you're a science
type as your signature suggests, you should have this book or something
like it that's as good. There is a wealth of both real-world practical
insights AND good in-depth understanding on many DIY topics in Heywood's
book.]

>If a sensor like the LAF O2-pump
>type sensor employs 2 of these, how can its reponse time be any better?

Well, the description of the pump-type sensors as "being made of 2
ordinary switch-type sensors" is no doubt an "operational analogy" for
the sake of explanation, not necessarily a statement of actual fact of
physical construction and performance/behavior, per se. My refuge as a
person who hasn't a clue about stuff such as "solid electrolytic nernst
reactions" or "gas ion diffusion constants in porous ceramics", is to
say, "well, they just are faster". Heh.

There's your daily dose of/on exhaust gas.

Gar


----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org




More information about the Diy_efi mailing list