Failed smog test in Colorado, CO too high at 2500rpms

[Rob Patterson]

Here are the details:
94 engine complete with wire harness and computer in 87 XJ.
94 Automatic computer with an AX-15.
65mm Throttle body
K&N filter charger kit
Jet Stage 2 chip for 94 automatic computer
87-90 Borla header with EGR port blocked off and DEI header wrap
1 year old 2.5" cat
1 month old Bosch O2 sensor
1 month old 24# FMS injectors.

Jeep ran great from Florida to Colorado while averaging 19mpg at 70mph on 35" tires and 3.55 gears. Yes the speedo is corrected!

When I went for the smog check, it passed the HC with 107 ppm at idle and 204 ppm at 2500rpms.
The CO passed at idle with .69% but failed at 2500rpms with 8.84%.

I have re-ran the test without the Jet stage 2 chip and 2 bottles of 91% isopropyl alcohol in 1/2 tank of gas. I drove the Jeep for about 2 hours and ran it hard to help clean it out. It still failed CO at 2500rpms with 12.2%. So it did not improve when I removed the chip.

I can only think of a couple of things:
1. Bad O2 sensor even though it is new.
2. The EGR plug on the header is leaking; allowing oxygen to enter the exhaust in front of the O2 sensor. This would cause the computer to see a lean condition and richen the mixture.
3. Bad cat

What do you all think?

 

[87xjco]

It's pretty easy to check the O2 sensor voltage output. If I remember right, approximately .9 volts equals the 14.1 stociometric ratio that is optimal. If your O2 sensor is reading a lean mixture, voltage will be closer to .5 volts, a rich mixture would be closer to 1.4volts.

If the O2 sensor readings can fluctuate between .5 or .6 volts (lean mixture can be induced to make it read low, like a vacuum leak) and up to 1.4 volts (by revving engine and loading it with fuel) then the O2 sensor is functioning properly

If at 2500rpm you are reading .7 volts or lower I would suspect your diagnosis #2 is on the right track.

If the O2 sensor reads close to .9 volts at 2500 which is close to an optimal reading, you may have a bad cat.

But if at 2500 rpm's you are reading 1.1 volts or more, I would suspect a rich mixture (opposite of what you mentioned) the computer would try to lean out the mixture. But the computer only has a limited parameter of control. I suspect your large throttle body and larger injectors are more than the engine can efficiently burn and the computer can compensate for.

Similar readings to what would happen if you installed a large carburetor on a small engine.

My .02

Terry

 

[gjxj]

If the O2 was continuously out of range the computer would see it and throw a code (Not quite sure what computer you have there though..) One thing that can happen however is the sensor is contaimated and just reads 'off'. No real way to check for that except to replace it.

 

[xjbubba]

The O2 signal should be constantly changing between ~.2v and ~.9v under conditions other than WOT. ~.5v=optimal fuel/air mixture. In closed loop operation the ECM drives the fuel mixture above and below the optimum mixture level to insure proper CAT operation. The number of times the mixture crosses above and below the .5V level is important--too few crossovers per sample-time indicate a "lazy" O2 sensor, and can result in a failed SMOG check.
With regard to the EGR plug. Are you saying your EGR is disabled? Or are you saying your using a different EGR port, and the one on your header is possibly sucking air? No EGR would definitely cause problems passing a SMOG check. Also, the leaking EGR plug would tell the ECM that the engines running lean, and would compensate by enriching the mixture; read:BAD MILEAGE. Not 19MPG at 70MPH!