o2 sensor readings, suggestions?

[UTKarmann_Ghia]

I failed emissions miserably a few days ago. Two years ago, she passed with flying colors. I have a feeling that this is related to my engine surging like crazy a few months ago. I installed a new Bosch o2 sensor, but the connector was broken in half and the wire exposed. At the suggestion of my mechanic, I cut the connector off, peeled the sheilding back and reattached the center wire to the single wire o2 sensor. Here's what I know:

The mechanic disconnected the freq valve behind the engine and checked that it was working. I'm assuming the he felt to see if it was vibrating? I read that in my Bently.

I verified the o2 sensor was outputting voltage while disconnected. The voltage was at .57, then dropped to .46 when I opened the oil tank cap (simulating a lean condition). I also noticed that the voltage increased when I increased the rpms.

But when I reconnected the sensor to the o2 wire harness, the voltage did not vary at all. I read in the Bently that the voltage should fluctuate after a couple of minutes, but it was locked at .54, no matter if I took off the oil cap or revved the engine.

I read some of the other o2 posts so I verified my dome light was working.

There's alot of things to verify and probably hours of troubleshooting via Bently, but I dont have pressure testers, so I cant test some things. It was weird, but the mechanic didnt want to try to fix it. Who turns away work when they have an empty garage?

Any ideas?

[drive-ability]

Post up the readings on the emissions report. List all the values on the print out. Make sure you list the passing limits as well as what your engine produced, forget about the average car readings. Look to see if any other things failed in the visual or any other area. The O2 should move about but it may have been too cold to begin working. You did fail and its likely the system was hot enough and should have been reading voltage. The .54 is the ecm's pre set for an open loop condition and is plugged in via the ECM. Let us know

[UTKarmann_Ghia]

Ok, if you're faint of heart, please look away

Type ---------Mine-------- Limit
HC------------- 5.1--------- 4.0 FAIL (not terrible)
CO----------- 108.4 -------45.0 SUPER FAIL!
CO2----------297.8-------- ????? (not listed)
NOx-----------0.8-----------8.0 PASS

The emissions in Colorado are not that tough from what I understand and my car passed easily 2 years ago. One thing I did do was gut the CAT which I thought was not effective anyways after 120k miles.

Any ideas?

[sammyg2]

runnning waaaay too rich. Check all ignition components first to make sure you are burning the fuel before doing anything else.
Then, try leaning out the mixture slightly (1/8 turn CCW) to see if it responds.
Then.... start looking at the fuel injection parts. My first instinct tells me to check the ingition first, then the catalytic convertor but that may be a knee-jerk reaction. Then find another mechanic.

[ianc]

From what I understand there is an inverse relationship between HC and NOx. Your car seems to be running very rich. Try leaning it out some. This will decrease the HC and CO and may bring the NOx up some, but it looks like you've got a little headroom in that dept...

ianc

[drive-ability]

Originally posted by UTKarmann_Ghia
Ok, if you're faint of heart, please look away

Type ---------Mine-------- Limit
HC------------- 5.1--------- 4.0 FAIL (not terrible)
CO----------- 108.4 -------45.0 SUPER FAIL!
CO2----------297.8-------- ????? (not listed)
NOx-----------0.8-----------8.0 PASS

The emissions in Colorado are not that tough from what I understand and my car passed easily 2 years ago. One thing I did do was gut the CAT which I thought was not effective anyways after 120k miles.

Any ideas?

HC and CO readings as you see are off. Co is a measure of how unbalanced the air to fuel ratio is. The term running "rich" would fit your situation.
The easiest/quickest way is to change the 02 sensor. You could verify a skewed O2 by doing what you did with the volt meter before. As I read you first post you said the 02 would react to intake conditions while disconnected but not when connected to the engine computer harness. I would connect the volt meter to the 02 sensor with it connected to the engine computer harness, basically have them both connected. Run the wire to the dash area so you can see while you drive. The thinest speaker wire will work as we are only working with Mil volts. You may need to drive it to get the engine to go into open close loop operation. When the computer sensor wire is reading .45 or so fixed and not changing its likely in open loop. As yet we don't know why its staying there so driving and heating up the sensor along with other factors may get it to go into closed loop. We want to see the sensor start outputting voltages which correspond with what the engine is producing. Now if it stays fixed at that area of .45 or there abouts driving my gut says theres something else keeping the computer from switching over to use the 02 sensor input to trim the mixture. Since the test readings were so high your 02 may be skewed and most often the sensor will fail with a problem which prevents it from reading higher 02 voltage outputs / thus it will be reading a leaner mixture than what is in really there and on top of that the computer will add more fuel to raise the voltage to where the computer map feels it should be. In short full "rich" which will produce a sulfur smell poor gas mileage black plugs and high CO and relatively low HC. Take those readings and or replace the 02 sensor and take them again. Post the results. John

[UTKarmann_Ghia]

Thanks John, I'll get her warmed up and run a wire to measure the voltage in closed loop mode. I hate to ask this, but is it possible that the lambda computer is holding the voltage in some way? What kindof variance would I expect from the o2 sensor in open loop?

[drive-ability]

Originally posted by UTKarmann_Ghia
Thanks John, I'll get her warmed up and run a wire to measure the voltage in closed loop mode. I hate to ask this, but is it possible that the lambda computer is holding the voltage in some way? What kindof variance would I expect from the o2 sensor in open loop?

In open loop the sensor voltage may stay fixed, the point is it wont vary as much. It would in close loop when working as designed toggle from .20 to .80 just cruising on a flat road at a steady throttle input. When you hit the throttle to pass someone it should read .90 or so, a rich mixture as well as when decelerating with your foot off the gas it will read something like .10 or .20 lean. All this surely depends on the up stream fuel injection system, sensors, injectors, fuel pressure etc. The O2 sensor is only there to trim mixtures while cruising its not involved at hard throttle inputs. The F/I maps kick in then. Remember the smog test is done at a set rpm speed range so it would be good to see the readings there as well. John

[SC-targa]

Originally posted by UTKarmann_Ghia
[One thing I did do was gut the CAT

Any ideas? [/B]

Why make it hard on yourself? Install another CAT.

Jerry Kroeger

[UTKarmann_Ghia]

The smog test here in Colorado takes the car through all 5 gears on a dyno and varies the rpms quite a bit and they actually give you a graph of what the levels are at different mph/rpms.

Jerry, you're assuming that the problem is the cat. I've heard of many 911's passing emissions with a gutted cat and I dont want to just throw money at this and hope something sticks. I'd like to understand the issue a bit and make some educated purchases. Is there something in the readings that assure you that the problem is the lack of cat? I'm a bit concerned that I'm not seeing the variance of voltage levels at the o2 sensor in closed loop mode.

[SC-targa]

Years ago, when I was still in college and these cars were new, I worked in an EPA certification lab. I saw many, many cars fail the test, even with a cat and God knows what else on them.

The car was engineered to meet those emissions specs with a cat. when you take out the primary emissions reduction element, you now are in uncharted territory and you have to develop your own emissions control system.

If you want to develop your own system, without a fuel pressure tester or a CO meter, you should at least check your distributor cap, rotor, plug wires and plugs. Make sure that your ignition timing is set at 5 degrees BTDC.

As someone else said earlier, your low NOx reading indicates a low peak combustion temp. That and the high HC and CO seem to indicate a rich mixture, or poor combustion. Lean it out a bit and see if it passes on a retest.

You don't need, or even want to be at the mixture setting that the O2 sensor is designed for, because that was planned around a 3 way cat system, which is a bit picky.

Good Luck,

Jerry Kroeger

[UTKarmann_Ghia]

Ok, I've taken some readings from the o2 sensor while it was disconnected, cruising down the road. When I let off the gas, the o2 sensor puts out roughly .1x for a bit then fluctuates around 2 or so. When I'm crusing about 45mph with steady throttle in 4th gear (~2500rpms), the voltage is pretty much rock solid at .83 and stays there.

I have a brand new 02 sensor, but I found that my o2 sensor connector was broken and could have been shorting (not sure). Could that have zapped my o2 sensor or my computer? It appears that for some reason, my sensor output is correct but the computer is not adjusting the mixture. I've had a mechanic disconnect the connector right behind the intake and verify that some plate was vibrating, so he thought that part of the system was good.

Any ideas???

[drive-ability]

The CO is high, so the first thing I would do is look for a sub-system thats adding extra fuel. First check the fuel pressure regulator, inspect the vacuum line attached. Remove it and look close at the regular fitting. Then see if any fuel is in the vacuum line end. Run the car with it disconnected and look at the regulator fitting. being carefully to see if any fuel is leaking out of the regulator fitting. Tap on the regulator while the engine is running see if any fuel comes out. If so theres your problem. I also would look at the evap system canister purge system. Maybe its drawing fuel there. Then make sure the oil isn't saturated with fuel. Smell the oil, check its viscosity etc. The O2 numbers you stated seem in a normal range. anytime your on the gas even a little the mixture should go above .8 volt. that's just the way the systems work. Thats throttle enrichment and normal. It sounds like the driver doing the test could be a factor. The idea a cat might help is not as likely with a high CO level. If its as bad as the readings I would think you would smell a real rich mixture at the tail pipe, but I bet you don't. I would test it again making sure the above things are OK and you have a clean air filter. You may have just got some odd readings. I have seen this happen where one test is much different than the other. Test everything above and the timing. Good luck and just keep your cool.




Type ---------Mine-------- Limit
HC------------- 5.1--------- 4.0 FAIL (not terrible)
CO----------- 108.4 -------45.0 SUPER FAIL!
CO2----------297.8-------- ????? (not listed)
NOx-----------0.8-----------8.0 PASS

[UTKarmann_Ghia]

Well, I've tested the snot out of theo2 sensor. I even installed an older one I had and got the exact, same results. She's at 0.84 when I'm steady on the throttle, but never drops until I completely let off the gas. Is there a way to test the voltage at the computer?

[nhromyak]

And of course, everyone knows an O2 sensor doesn't put out a specific voltage for any length of duration - right?

An O2 sensor will do "hysteresis" (sp?) and fluctuate very rapidly (something like 2 or 8 times a second) across a "normal" voltage of .45 volts.

In other words, if you check the voltage, the O2 sensor maybe putting out that voltage, but for a computer to "read" it correctly, the O2 sensor must be doing "hysteresis".

As the O2 sensors get older, the slow down hysteresis and eventually stop fluctuating, giving you poor gas mileage, and in most cases, poor performance.

Are you using a Digital VOM?

[UTKarmann_Ghia]

I am using a digital volt meter, I'm not sure what you mean by VOM and I've used a new and old o2 sensor with the same results. I'm kindof stumped on what to do next other than to verify that the o2 voltage is received at the computer. Any ideas?

[ianc]

VOM = Volt Ohm Meter

[UTKarmann_Ghia]

Is there a diagram somewhere I can reference that will show me the o2 sensor input to the computer and the output to the frequency valve?