I am not an OBDII "expert" either, but I am a former lead OBDII calibrator. I worked at Ford Motor Company on their natural gas truck / van program.

     Ed was correct in claiming the downstream O2 sensor (DSO2S) is used to check the condition of the cat(s). I calibrated and tested the catalyst monitor, so I can give a bit of insight. I believe I went through this explanation here on the Digest about a year ago, but here we go again...

     The DSO2S will show an air / fuel swing at a much slower frequency than that of the upstream O2 sensor (USO2S). This frequency (of the DSO2S) is monitored to ensure it maintains its slow cadence. If it begins to change as rapidly as the USO2S, it is assumed that the cat is not operating properly.

      To ensure that the DSO2S is operational, validity checks are made (similar to that of the USO2S). I tested the DSO2S for high/low voltage, lean/rich shift, intermittent operation, and circuit continuity.

     What you must understand is that we used specific "entry" conditions for these tests which sometimes included engine temp, time within closed loop operation, engine speed, throttle position and vehicle speed. Once these conditions were reached, only then would a timer begin (sometimes 30 seconds long). After which the tests would be run. If the test is then failed, the PCM must observe this failure in two "drive cycles" or "trips" within 40 "trips" of each other. Only then will you see a MIL.

     Phew, if this is all seems a bit complicated - it is, but with good reason. Although it was my job to illuminate the MIL when an emissions component failed, I had two concerns: 1) Under what operating conditions do I know the sensor is reliable; 2) Will the failure be detected during a Federal Test Procedure (FTP). The FTP is a drive cycle which attempts to simulate an LA drive cycle. They have a steady cruise at approximately 48 mph for about two
minutes. This is where many of my tests were run. I sometimes had trouble simulating the failure diagnosis on the road. The timer would be zeroed if I let off the gas for a split second.

     My point is, someone may be driving with a bad O2 sensor and never know it. This can occur if you only race (nothing is tested at WOT), or if you only do city driving.

Sorry for the long-winded explanation, but there is no simple rationale when speaking of the OBDII system.

Matt Blue
Former OBDII guru (not)
'90 & '91 Tsi AWD