Flunked emissions inspection 3 times in a month now with a '94 Chevy Suburban 5.7 L with throttle body injection and 1 heated zirconia lambda sensor in front of the cat; OBD-I. The first time it was high on HC at idle (320 vs 220 limit) and high on both HC (720 vs 220 limit) and CO (1.69% vs 1.20% limit) at high idle (2000 rpm or so). Replacing an open spark plug wire and the worn but clean spark plugs, as well as a squirrely ignition module (tach would leap about erratically while engine misfired at cruise) seemed to cure the high idle emissions at the 2nd emissions test (CO dropped to 0.02%). But nothing changed at idle; still flunked with similar numbers (306/220 HC). (There had been no OBD-I codes set the entire time, even with the horribly misfiring intermittent ignition over the past weeks.) So I figured the 100k mile old cheap aftermarket cat was fouled from all the misfiring. (I'd also read an opinion that this generation of TBI fueling systems are so inefficient at idle that one needs a cat to pass emissions even if everything else is working fine.) So I put in a new cat and lambda sensor. Driving it the first day after new cat and sensor, it set the malfunction indicator light with a lean O2 sensor code (44). The MIL extinguished itself on the next drive cycle, but I haven't yet cleared the stored code. On the 3rd emissions inspection, flunked again at idle (344/220 HC). So I hooked up my new 4423 Picoscope and monitored the brand new lambda sensor's voltage at idle. It seemed horribly erratic in its timing, with no pattern whatsoever. At cold startup it reads low, and over a minute or so creeps up to very rich, and then starts its erratic cycling when the engine goes into closed loop. The "o2 sensor erratic" data file attached shows this erratic cycling. It paints a broad band (I think) because I hadn't yet learned to use the filter. Then I set up and recorded lamda heater current (with an old current clamp I had lying around), lambda voltage, throttle position, and manifold absolute pressure. Ran it at idle, 2500 rpm, and gave it WOT throttle blips. The lambda voltage seems to behave as I'd expect at higher rpms, but sits rich for long times (11 seconds at one point) and also lean for long times at idle. (I can't seem to attach this file in addition to the first file in this posting. When I try, it deletes the first one, leaving no attachment, and clears the posted text as well. ??) I suspect that some other components (hopefully not the ECU!) are causing the ECU to erratically jerk the fuel (or air) around. My OBD-II scan tool won't plug into this OBD-I vehicle so I plan to get ahold of an OBD-I tool this week to read short and long term fuel trims. Any comments or suggestions as to what I should investigate next?
- 20140221-0003 o2 sensor erratic at idle.psdata
- (3.04 MiB) Downloaded 149 times