Oxygen Sensor Woes
 

I'm having a very tough time getting my car through emissions and I think it is O2 sensor related. Here's what I'm running:

ZZ4 w/TPI injection and 1.6 rockers
Accel DFI Gen 6 system
Long tube headers
O2 sensor (standard 3-wire narrowband) approx. 1.5 inches behind collector flange in passenger side header
Mallory Hyfire 6AL

And the problem: I went to get emission tested last weekend and failed (twice) in all areas - idle and loaded. I discovered after some computer logging that my O2 sensor is not reading at idle - it shows ~0.02 volts and this continues at low loads as well. I did have the computer setup to run closed loop at idle and the computer is adding fuel because it thinks it is lean. The sensor is also very slow to respond to load changes. When it does read, the highest reading I get is maybe 0.300 volts. I have occasionally seen it read on the rich side, but only briefly. I really don't trust the lean reading because just walking behind the car while idling will make your eyes water - it is very rich and friends who have driven behind me complain that it stinks.

At the time of the emissions test, I did not have the Mallory box hooked up. I tried last night to hook up a 4-wire narrowband I borrowed from a friend, but I got the same types of readings - nothing at idle and lean once loaded. The Mallory box allowed me to add some timing at idle and also pull 25% of the injector pulse, so it definitely ran better and I think I'm getting closer. The idle quality certainly improved, but I am now in open loop at idle to prevent the computer from adding fuel again.

So my big question is - what next? I think I need to get a wideband that I can monitor and also feed the narrowband signal into the computer. I have worked with the widebands a great deal in my past jobs, but not with my own car or with others, so I'm concerned that I'll run into the same issues that I have with the narrowband. However, I do strongly believe that a wideband should be able to keep itself hot enough to read (and be accurate) at idle. It just seems odd that the narrowbands - both heated - are not reading. Is it possible I have a giant exhaust leak that is tricking the sensor?

Open to all opinions and thanks in advance!
Ryan

Is this something you've been driving for a while?

Yeah, about 1500 miles. It's only been running like this for a few hundred.

Ryan

So the calibration is good. Have you verified fuel pressure?

No, but I was going to include that this weekend when I get another chance to work on it. I have a feeling it's just fine.

I think I'm going to check out the O2 wiring and make sure the heater circuit is working properly before I go much further.

Ryan

Well let us know what you find. You call your ECU a gen 6. Is it a 6.0 ECU? Rectangular in shape and brushed aluminum in color? Calmap software? I worked at DFI about 6 years when the Meaney bros. were there.

Yep, that's it! I'm also using the DOS-based version of the Calmap software. I think the computer part number is 74022 - just the plain Jane version. If I had lots of cash around, I'd probably move up to the Gen 7 or the Holley Commander 950.

Well now I know who to ask when/if I have problems :)

Ryan