Tech to Tech
The Last Word On O2 Sensors ... For Now
Will I ever get off the subject of oxygen sensors? Probably not,
but this will be the last one for a while. The whole computer-controlled
system is designed around the needs of the catalytic converter
and the O2 sensor is responsible for making the final trim on
the fuel mixture to achieve the correct result. O2 sensors are
critical, and what’s more, sensor technology has changed.
The last two columns covered low O2 readings and high O2 readings respectively. First a correction. My July column stated that the only reason for a rich O2 reading was if the engine was rich. Not true and I knew better. Some types of contamination can cause a false rich shift, causing the engine to run lean. I’ll try to be more careful, but don’t hesitate to fax me your opinion if you’re so inclined. I know my friends in Arizona don’t hesitate to criticize my mistakes at every opportunity! (I should add the traditional e-mail <G> for "grin" here, since I’m just kidding.)
Now let’s discuss sweeping O2 signals that may be false. When looking at a nice, normal, sweeping pattern on a scope, the natural inclination is to assume everything is fine. After all, isn’t that what it was intended to do? Well, yes, but it still must be compared with the exhaust gases.
At my shop, we once spent three days and a carburetor overhaul trying to fix a 2.0 percent CO failure on a Land Cruiser. The O2 sensor signal was sweeping perfectly the whole time! I should have thought this one out better, but it was about 10 years ago and we were all a lot greener on feedback systems. So what happened? Well most good running engines with a 14.7/1 AFR will produce about a 1 percent to 1.5 percent O2 reading in the exhaust stream. So the O2 sensor is calibrated to measure less than that as rich and more than that as lean. OK, this Land Cruiser had blown an air pump in the past and the carbon-fiber vane fibers had gone downstream. A small chunk had lodged in the upstream valve of the diverter valve. So even though the vacuum signals to the diverter valve were correct, some air was always being fed upstream. The O2 sensor saw the air and drove the mixture richer. The rich mixture had almost no O2 in the exhaust and the false air was just enough to have the O2 sensor in a happy sweep.
We marked that one up to experience. Let me define "experience" for you. Experience: What you get when you didn’t get what you wanted! For me, this brings up one of the positive things about being involved in the Automotive Service Association (ASA). ASA members often share these technical (or business) "experiences" with one another so we all don’t have to learn the hard way.
Now back to the mighty O2 sensor. Keep in mind that the O2 sensor voltage can shift higher or lower depending on the type of contaminate. Some common contaminates are silicone (from antifreeze or engine sealers or products used to clean and shine), lead from leaded fuel, carbon fouling the vents and oil ash. Cracked O2 ceramic sensor thimbles and fouled reference holes can cause negative readings.
These contaminants are pretty easy to find when the problem has become so acute that the sensor is stuck high or low. However, prior to failing high or low, the sensor has been shifting down or up while still sweeping. Also, O2 sensors spending more time low than high or vice versa can indicate problems. This is why a scope is so important and using the complete rich/full lean test is so important. Remember that when driven full lean, we want to see under .2 volts and when driven full rich, we want to see over .8 volts. Too high a voltage or too low (negative) indicates a sensor drifting out of calibration.
So far all the writing I have done on O2 sensors has been primarily about zerconia dioxide O2 sensors. These were the standard for many years. However, in the late 1980s the car manufacturers had to throw us another curve, so some started using titania dioxide sensors. Remember the rule of thumb that low O2 sensor voltage equals lean (L=L)? Well that’s not always true with titania dioxide sensors.
Why the switch? I posed this question to all the brilliant minds on the "For Techs Only" section of the CARS forum on CompuServe. Here’s what the cyber-techs said. First, titania dioxide O2 sensors are cheaper to produce, which means we are going to see plenty of these babies! Second, they heat up and read faster, which is good news for emissions. Third, they need no reference air, this is why Jeep went to them early. (Mud can’t foul the reference air hole because there isn’t one.)
OK, so how is this sensor different? Well it doesn’t produce its own voltage. It acts as a variable resistor to modify a reference voltage supplied by the computer. Now here is where it gets ugly. The sensor and circuit can be designed to work at any voltage and even the reverse of what we are used to (low voltage can be a rich mixture and high can be lean).
Boy, was I surprised when I saw my first Jeep sweeping between 0 and 5 volts! I was sure something had to be wrong, so off I went to the trusty manuals! My CD-ROM-based manuals gave a generic description of a regular O2 sensor and signal. Hmm. Was this right or wrong? Time to consult the sacred factory manual! It said the O2 signal changed relative to fuel mixture. Wow, what an in-depth description! I suppose the factory guys didn’t figure we might look at the signal with a voltmeter and want to learn information on the parameters, much less use a scope!
So where do you go from here? Well, your choices are an expert, a technical hot line or on-line. I logged onto CompuServe and asked for information. The next day I had a response from a Jeep tech who I knew was a top tech. He confirmed that this Jeep used a 5-volt sweep and that the sensor read backward. Case closed. Get on-line!
So who is using what type of sensor? This is knowledge gained from experience, and I believe what I heard on-line is correct. Jeep and Eagle used 5-volt, backward-reading O2 sensors from 1989 to 1992. Everybody else using titania sensors read in the normal direction, except that some (Hyundais) may read as high as 1.5 volts. Looks like comparing to the tailpipe is becoming more and more important!
When installing a new sensor, check that the vents in the thimble of the sensor match the old sensor. Some have clockwise slots and some are counterclockwise. Also, the number of slots may vary. The designers of these sensors are very careful to test many different configurations on each engine to find the one that responds the best. Different engines have different heat and swirl patterns in the exhaust, so the O2 slot pattern can be very important. I have heard reports of installing an O2 sensor with a different vent pattern causing dramatic slowing of the sweep rate.
Changes in O2 sensor technology must be dealt with. Improvements in technology will constantly change the way we work and the rules we work by. That’s what makes our jobs so interesting and frustrating at the same time.
If you have a comment or question for Mark Warren, the best way to contact him is by sending a fax or e-mail directly to his office. Warren’s fax number is (520) 790-9173, and his e-mail address is 76727.3631@compuserve.com (no period at the end).
TechTips || Stat Corner || News Briefs || Directions
News Briefs Extra || Taking The Hill || More Info About AutoInc.
Chairman's Message || Tech To Tech
Guest Editorial || Diversifying Your Collision Shop || National VOC Rule Sparks Congressional Interest
AutoInc. Magazine ®, Vol. XLIV No. 8, August 1996