Question on O2 sensors (generic q)

[kenw]

OK, so the Camry threw a code and lit up the Check Engine lamp last week: Bad heater, O2 sensor, Bank 1 Sensor 1.

There was nothing visibly wrong with the removed one, it wasn't cracked, damaged, plugged, coated or otherwise visibly defective. It looked like a part that had been in the exhaust stream for 178,000 miles.... kinda brown tinted! Meter verifed, replaced it, no issue there.

30 minutes of work and the dealer wanted $200 for labor alone....but i digress.

Observation: the car is now running like the proverbial batouttaH*** on 87 octane (it prefers 91+). So I'm guessing the 02 sensor (1 of 3 btw) had been marginal for some time, long enough for me to forget how this pup could run. But not bad enough to throw a code?

The heater is a simple resistive circuit: it's either good ("X" ohms range) or not (infinite ohms, open circuit).

My question (finally, they say): does an o2 sensor in fact go bad over time **without** being bad enough to send an ECU code? The only test in the shop manual is the ohms on the heater. What about the actual sensor? How to test it?

tia for any intelligent answers...

 

[eljefino]

Yup, and that's true for most any sensor.

You've got to slip below "acceptable" when new is "great". Otherwise you'd always be tripping codes. And that's the nature of a feedback fuel system-- your o2 sensor is there to account for calibration differences in your MAF, TPS, ECT, the gas you bought, etc., and apply a "bias" (block learn) to your fuel/timing maps.

o2s are great examples of bum sensors though because they switch rich to lean "several" times per second. The computer will only complain about "lazy" old sensors if they switch once every other second or so. And if it quits entirely its reading has to strongly disagree with everything else the 'puter knows about the engine load and what fuel it's contributing.

How much each sensor contributes is a rather tight engineering secret, but obviously in your case the 02 is highly regarded.

 

[kenw]

talking with a knowledgeable friend, his opinion on the "theory of operation" is that the resistor current varies by the rate of oxidation, which then gives an indication of the 02 content.

Evidently therefore, the rate of oxidation is unmeasureable by standard means, so the only practical way to determine failure is when the circuit totally fails.

So the resistor IS the sensor?

Makes sense to me. Chime in if you know different.

 

[brianl703]

Most oxygen sensors are basically cells that generate a voltage when there is no oxygen present in the exhaust.

There is a type of oxygen sensor that is a resistor, but it is not very common and I can't remember what vehicles use(d) it.

The voltage output on the standard oxygen sensor when new peaks at about 900mV and starts dropping as the sensor ages.

You can measure the output of the standard oxygen sensor by using a high-impedance multimeter..digitals will work, unpowered analog meters won't. A meter with a bargraph display is probably best as the digital display won't update fast enough for you to get a good idea of how fast the sensor is switching.

You can also use a scope, if you have access to one.

 

[eljefino]

I thought o2 heaters were just to get the sensor hotter faster, so the car get into more accurate closed-loop sooner.

You should have at least three wires: Power, ground, and signal (and probably signal ground). However, to diagnose the heater element, the computer monitors its current draw too. The code you got was for the heater, not the measured fuel mix.

 

[brianl703]

quote:

---

Originally posted by eljefino:
I thought o2 heaters were just to get the sensor hotter faster, so the car get into more accurate closed-loop sooner.

---

That's correct. Heater failures are pretty rare though. Usually the sensor slows down enough that the computer is complaining about the switch rate first.

 

[Tosh]

Ken, had you noticed a drop in mpg? Do you have emissions testing there and was it affected? Will you soon be replacing the other two O2 sensors yourself?

 

[brianl703]

Since the post-cat O2 sensor (which is the 3rd sensor) does not control the mixture, there's no benefit to replacing it before you get a code for it (which may never happen, given that the post-cat sensor has an easy life).

 

[PandaBear]

All the above are correct. $200 with labor sounds about right for dealer replacement. I replaced my own before it fail, just because my smog result is barely passing (on the max limit, so will fail the next time) and might as well replace at the same time with my other car to get discounts.

From what I heard heated sensor locates much further down the pipe and do not rely on the exhaust temp, thus last longer, doesn't mean they are more accurate, if the heater fail, the sensor will be more lazy and not function as well.

My car feels preppier with 5% better milage after the replacement. Got my generic ones for $40 (heated) and $30 (non-heated).

 

[kenw]

Thanx for the technical info guys. This sensor was the rear bank (v6) pre-cat unit, absolutely the worst as far as accessiblity! (ramps and a free tool rental made it possible). The front bank pre-cat unit should be a piece of cake, it's right in the open with lots of woking room. If needed, I can replace them myself as well.

No noticeable drop in mileage altho I haven't logged it for ages. As mentioned, immediately after replacing perceptible power was up dramatically. And I mean dramatically! (can't wait to see what she'll do with some nice 93 in the tank.) Based on that, i suspect the mileage was suffering. In rough terms, I always got 4 round trips to work on a tank before......

Never got a clue from emissions (state/Houston sniffer) reports, last one in July was virtually identical to the previous year.

$40 for a heated one? youch! Mine priced from $104 to $154 (dealer).

But then the first lasted 178k so the per mile rate is pretty good....

 

[HardbodyLoyalist]

Oxygen sensors use what's known as an "oxide ion conductor," usually zirconia-based, to create an electrochemical cell. The computer monitors the voltage drop across the cathode (oxygen rich side, exposed to the atmosphere) and anode (oxygen lean side, inside the exhaust stream).

I found a fairly decent write-up describing their operation:

http://home.flash.net/~lorint/lorin/fuel/lambda.htm

 

[brianl703]

quote:

---

Originally posted by kenw:
$40 for a heated one? youch! Mine priced from $104 to $154 (dealer).

---

Honda sensors are $250 each, either from the dealer or Autozone/Advance Auto.

http://www.oxygensensors.com has them for $60, they have the OE connector (they are not a universal sensor) and they work fine, so I think someone's cashing in on Honda's reputation for overpriced parts.

 

[ex_MGB]

Noticed the same thing in my '97 Maximum when I replaced the O2 sensors at 110k when I got a code. Very noticeable increase in pull in the 3k to 4k rpm range during acceleration.

O2 sensors should be considered disposable and replaced every 80k miles for optimal vehicle performance.

 

[kenw]

Thanx again for the info and the links. Brianl703, I've bookmarked that site for sure.....

 

[JHZR2]

Yes, O2 sensors can get lazy... this means that they throw out voltages that are within spec, but whereas the voltage is supposed to fluctuate very fast, and it doesnt when they get 'lazy'.

Our 94 previa's sensor threw a CEL at about 178k... seems about the right time to change it. We changed both, and the funny thing was, the CEL was only intermittent for the ~150 miles we drove it with a poorly perfroming sensor, and the fuel economy averages from long before to long after have stayed the same...

I think ours was going bad by another mechanism, yours was getting lazy.

JMH