Downstream o2 - Fuel trim or no?

[5-90]

Damn. Have to get even more basic I think.
If I force the ecm into limp mode by disconnecting an 02 sensor entirely, I still get the same white smoke/vapour at startup. No change whatsoever.
So whatever the problem is, it's not at all related to any electronics.

Will try to track down a vacuum gauge to test cat today. It's seeming like the best option now.
[

VISIBLE TAILPIPE EMISSIONS KEY:
- Black Smoke. Carbon particulates, from very incomplete fuel combustion. Find out why you've gone so rich.
- White "smoke." This is actually steam - literally. You've got coolant getting into the combustion chamber somehow. Perform a compression test to find the potential source, then a leakdown test on suspect cylinders.
- Blue "smoke." Steam again - this time, from engine oil getting into combustion. Suspect oil control rings or valve guide seals (if you get this mainly on startup or acceleration, replace valve guide seals. If you get this constantly, it's likely your oil control rings have worn or lost tension. Prepare for a re-ring.)

Nota Bene - The only smoke with a potential electronic cause is the first, the black. This is the one where fuel trim is thrown out of whack. However, it's still possible for this to have a mechanical cause - typically, the pintles in the fuel injectors are failing to seal for some reason. Either/or. Clean the injectors thoroughly before sourcing any parts, and (if possible,) examine the switching waveform for the HEGO sensor (spec life for HEGO in pre-OBD runs about 100kmiles, in OBD-I about 75-80kmiles, and in OBD-II about 60kmiles.)

Any white or blue steam exiting the tailpipe will have a mechanical cause.

(Another trick to find the source for white or blue steam is to examine your spark plugs. Cylinders that are fouling with oil will usually have wet-looking black deposits on the electrode insulator, while cylinders fouling with water will be noticeably cleaner than the rest - because they're literally being steam-cleaned. If a cylinder is fouling excessively with fuel, deposits will be dry, sooty, and possibly flaky.)

 

[Devilfrog]

when I let the XJ sit for a few days, then start it I get lots of white steam out the back .... no coolant losses, and it's mainly since adding a Bosch UHEGO a while back. Once it's done the cloud, it's perfectly normal as long as I keep driving it, park it for a few days and it pushes a small cloud again.

My rear DHEGO just popped a code, but the sensor and the cat are 2000 stock parts so it's not unexpected.

Otherwise she runs fine and gets the same fuel economy as always.

 

[Keekleberrys]

No it should swing wildly. Both of them work the same way by reporting lean/rich.

You able to do the vacuum test for a clogged cat? Cruiser's got a writeup around here somewhere. Vacuum leaks too if you haven't already, that's a lot of NOX

I think you are mistaken. In OBD2 vehicles the downstream o2 sensor is used to monitor catalyst efficiency. In a properly functioning system your 14.7:1 stoichiometric ratio is determined by the upstream sensor and then checked against the downstream to see if you're polluting. Emissions is the entire point behind OBD2.

If your downstream is doing the same as your up, it is very likely the catalyst is damaged or no longer working effectively. If your catalyst is working effectively why would the downstream be reporting the same as the up? It wouldn't... because the 3way cat is converting HC,CO, and NOx with reduction/oxidation reactions that consume O2.

 

[holysirsalad]

I think you are mistaken. In OBD2 vehicles the downstream o2 sensor is used to monitor catalyst efficiency. In a properly functioning system your 14.7:1 stoichiometric ratio is determined by the upstream sensor and then checked against the downstream to see if you're polluting. Emissions is the entire point behind OBD2.

If your downstream is doing the same as your up, it is very likely the catalyst is damaged or no longer working effectively. If your catalyst is working effectively why would the downstream be reporting the same as the up? It wouldn't... because the 3way cat is converting HC,CO, and NOx with reduction/oxidation reactions that consume O2.

You've misunderstood what I wrote, but also you're incorrect in your generalization.

The both O2 sensors function in identical manners by reporting a "rich" or "lean" signal in relation to a fixed oxygen density ratio. The use of this input for the upstream O2 sensor is obvious, too little oxygen is rich, too much oxygen is lean. The pattern at which the upstream O2S reports rich/lean is called the switching rate. The downstream sensor's input should mimic the upstream's switching rate after the PCM considers time delay as the gas flows through the pipe and a pre-calculated effect of the catalytic converter. If the downstream O2S fails to fall into an acceptable switching rate, for example permanently reads lean or is slow to react, then a code is thrown to cause further investigation.

As I stated in post #7, starting at least for the 2000 model year Chrysler used the downstream sensors in calculating the fuel trim as well. I personally believe this is for tightened EPA rules on how emissions system failures are handled, as noted by the phrasing "...to control emissions", similar to how an EPA Tier 4 diesel will go into limp mode if you run out of DEF.

The below except is from the 2000 XJ FSM section 14 "Fuel System - Fuel Injection" pages 30-31:

OXYGEN SENSOR—PCM INPUT

DESCRIPTION
The Oxygen Sensors (O2S) are attached to, and protrude into the vehicle exhaust system. Depending on the emission package, the vehicle may contain either 2 or 4 sensors. On non-California emissions packages, 2 sensors are used: upstream (referred to as 1/1) and downstream (referred to as 1/2). On California emissions packages, 4 sensors are used: 2 upstream (referred to as 1/1 and 2/1) and 2 down-stream (referred to as 1/2 and 2/2).

OPERATION
An O2 sensor is a galvanic battery that provides the PCM with a voltage signal (0-1 volt) inversely proportional to the amount of oxygen in the exhaust. In other words, if the oxygen content is low, the voltage output is high; if the oxygen content is high the output voltage is low. The PCM uses this information to adjust injector pulse-width to achieve the 14.7–to–1 air/fuel ratio necessary for proper engine operation and to control emissions. An O2 sensor must have a source of oxygen from outside of the exhaust stream for comparison. Current O2 sensors receive their fresh oxygen (outside air) supply through the wire harness. This is why it is important to never solder an O2 sensor connector, or pack the connector with grease.

Four wires (circuits) are used on each O2 sensor: a 12–volt feed circuit for the sensor heating element; a ground circuit for the heater element; a low-noise sensor return circuit to the PCM, and an input circuit from the sensor back to the PCM to detect sensor operation.

Oxygen Sensor Heaters/Heater Relays: On a certain non-California emission package, the heaters on both sensors are fed battery voltage from the ASD relay which is controlled by the PCM. Refer to ASD relay for more information. On another non-California emission package, the heaters on both sensors are fed battery voltage from the two O2S heater relays. The O2S relays are also controlled by the PCM. On the California emission package, the heaters on all 4 sensors are fed battery voltage from the two O2S Heater Relays.

The O2 sensor uses a Positive Thermal Co-efficient (PTC) heater element. As temperature increases, resistance increases. At ambient temperatures around 70°F, the resistance of the heating element is approximately 6 ohms. As the sensor’s temperature increases, resistance in the heater element increases. This allows the heater to maintain the optimum operating temperature of approximately 930°-1100°F (500°-600° C). Although the sensors operate the same, there are physical differences, due to the environment that they operate in, that keep them from being interchangeable.

Maintaining correct sensor temperature at all times allows the system to enter into closed loop operation sooner. Also, it allows the system to remain in closed loop operation during periods of extended idle.

In Closed Loop operation, the PCM monitors certain O2 sensor input(s) along with other inputs, and adjusts the injector pulse width accordingly. During Open Loop operation, the PCM ignores the O2 sensor input. The PCM adjusts injector pulse width based on preprogrammed (fixed) values and inputs from other sensors.

Upstream Sensor (Non-California Emissions): The upstream O2S sensor (1/1 sensor) is located in the exhaust downpipe before the catalytic convertor. It provides an input voltage to the PCM. The input tells the PCM the oxygen content of the exhaust gas. The PCM uses this information to fine tune fuel delivery to maintain the correct oxygen content at the downstream oxygen sensor. The PCM will change the air/fuel ratio until the upstream sensor inputs a voltage that the PCM has determined will make the downstream sensor output (oxygen content) correct.

The upstream oxygen sensor also provides an input to determine catalyst efficiency.

Downstream Sensor (Non-California Emissions): The downstream heated oxygen sensor (1/2 sensor) is located near the outlet end of the catalytic convertor. The downstream sensor is also used to determine the correct air fuel ratio. As the oxygen content changes at the downstream the PCM calculates how much air fuel ratio change is required. The PCM then looks at the upstream oxygen sensor voltage and changes fuel delivery until the upstream sensor voltage changes enough to correct the downstream sensor voltage (oxygen content).

The downstream oxygen sensor also provides an input to determine catalyst efficiency.

Upstream Sensors (California Emissions): Two upstream sensors are used (1/1 and 2/1). The 1/1 sensor is the first sensor to receive exhaust gases from the #1 cylinder. Both of the upstream O2S sensors are located in the exhaust manifold just before the mini-catalytic convertors. They provide an input voltage to the PCM. The input tells the PCM the oxygen content of the exhaust gas. The PCM uses this information to fine tune fuel delivery to maintain the correct oxygen content at the downstream oxygen sensors. The PCM will change the air/fuel ratio until the upstream sensors input a voltage that the PCM has determined will make the downstream sensors output (oxygen content) correct.

The upstream oxygen sensors also provide an input to determine mini-catalyst efficiency.

Downstream Sensors (California Emissions): Two downstream sensors are used (1/2 and 2/2). The downstream sensors are located in the exhaust down-pipes just after the mini-catalytic convertors. The downstream is also used to determine the correct air fuel ratio. As the oxygen content changes at the downstream the PCM calculates how much air/fuel ratio change is required. The PCM then looks at the upstream oxygen sensor voltage and changes fuel delivery until the upstream sensor voltage changes enough to correct the downstream sensor voltage (oxygen content).

The downstream oxygen sensors also provide an input to determine mini-catalyst efficiency.

 

[Keekleberrys]

You've misunderstood what I wrote, but also you're incorrect in your generalization.

The both O2 sensors function in identical manners by reporting a "rich" or "lean" signal in relation to a fixed oxygen density ratio. The use of this input for the upstream O2 sensor is obvious, too little oxygen is rich, too much oxygen is lean. The pattern at which the upstream O2S reports rich/lean is called the switching rate. The downstream sensor's input should mimic the upstream's switching rate after the PCM considers time delay as the gas flows through the pipe and a pre-calculated effect of the catalytic converter. If the downstream O2S fails to fall into an acceptable switching rate, for example permanently reads lean or is slow to react, then a code is thrown to cause further investigation.

As I stated in post #7, starting at least for the 2000 model year Chrysler used the downstream sensors in calculating the fuel trim as well. I personally believe this is for tightened EPA rules on how emissions system failures are handled, as noted by the phrasing "...to control emissions", similar to how an EPA Tier 4 diesel will go into limp mode if you run out of DEF.

The below except is from the 2000 XJ FSM section 14 "Fuel System - Fuel Injection" pages 30-31:

I wasn't really disputing the fact that the PCM may use information from the downstream o2 sensor, so much as I was pointing out the fact that your understanding of how oxygen sensors work together is wrong.

Yes the upstream switches quite rapidly to maintain the proper stoichiometric ratio, but the downstream voltage signal should be very steady due the action of the catalyst.

All an oxygen sensor does is send an analog voltage signal in relation to the amount of oxygen in the exhaust stream. The majority of which is consumed by the three-way catalytic converter in order to reduce excess emissions of hydrocarbons, carbon monoxide, and oxides of nitrogen. Oxidation and reduction reactions consume a majority of O2 in the exhaust stream, so the amount of oxygen in the exhaust stream by the time it reaches the second sensor should be fairly constant.

If you are seeing the same level of activity on an oscilloscope for the downstream that you are seeing on the up, then your catalytic converter is toast, because it's waveform should be almost constant or changing extremely slowly. This is basic engine performance. Check it out on an oscilloscope some time.


The oxygen density of the exhaust stream is NOT fixed from one sensor to the other, or there would be no point in having two. The primary reason for the second sensor is to verify effectiveness of the catalyst, that's OBD 2. Regardless of however else chrysler may have used the seconday sensor data as input to the PCM.

 

[holysirsalad]

Yes the upstream switches quite rapidly to maintain the proper stoichiometric ratio, but the downstream voltage signal should be very steady due the action of the catalyst.

Thanks, that part I didn't understand.

I found a lovely illustration of the two. Blue is an upstream and red downstream with no cat, green is upstream and brown is downstream of a working cat.

 

[Sidewinder CC]

Yes the upstream switches quite rapidly to maintain the proper stoichiometric ratio, but the downstream voltage signal should be very steady due the action of the catalyst.

Could you explain what you mean by "downstream voltage signal should be very steady" ?

The downstream O2S is supposed to switch like the upstream O2S. If the downstream O2S is 'steady' something's dead.

I don't see this when viewing my O2S activity with my scanner. My downstream O2S switches around the same as the upstream sensor, about every 2 seconds.

'99 XJ.

 

[Keekleberrys]

Could you explain what you mean by "downstream voltage signal should be very steady" ?

The downstream O2S is supposed to switch like the upstream O2S. If the downstream O2S is 'steady' something's dead.

I don't see this when viewing my O2S activity with my scanner. My downstream O2S switches around the same as the upstream sensor, about every 2 seconds.

'99 XJ.

Serial data responds too slowly to observe o2 sensor data. You have to look at it oscilloscope.

 

[Sidewinder CC]

Serial data responds too slowly to observe o2 sensor data. You have to look at it oscilloscope.

Even if I had an Oscilloscope I'm too stupid to understand what it's telling me. The only thing a simple person like me has is a scanner that shows live O2S voltage data, and a FSM that tells me what to look for with a pending, or actual, Cat failure.

 

[blistovmhz]

This is killing me

I believe that what would MAKE SENSE, would be to expect the downstream O2 readings to be mostly level around 0.45V. This makes sense to me because the 4.0L engine is designed to run a tad lean so that there is enough oxygen left in the exhaust stream to enable catylisation in the cat. If you run too rich, you won't have enough o2 for the cat to work properly, and too lean you run the risk of engine damage, with no benefit to the cat process. Seems to be there should be JUST enough o2 for the cat to do it's job, which means there should ideally, be very little to no O2 left after the cat. Any rapid and wide fluctuation in the DS O2 sensor would indicate a failure of either the ECM to send the correct amount of O2 to the cat, or a failure of the cat to consume the O2 and turn it into H2O and CO2.

That said, that's just what makes intuitive sense to me but perhaps computers weren't quite there in 1998? No matter what, it seems like a wide fluctuation in O2 post Cat, means your fuel trim isn't being set efficiently.

And of course, as no two people agree on the same operation of the emissions systems, I'm still no closer to knowing whats wrong with mah Jeep Fuuuuu.

Also keep in mind that I don't see the white smoke BEFORE the Cat. If I separate the downpipe from the collector, I don't see any smoke coming out the header/collector. Maybe because the exhaust gases aren't the right temperature to make the white smoke, but I think this bit of data is probably key to understanding the problem. Are the gasses just heating up due to catalysation, and turning white, or are they still too hot at the collector, and they turn white when it cools down in the tailpipe? Either way, something odd is happening at the Cat.


Is there ANY way to know definitively if I'm running too rich? The cat obscures a slight rich condition by performing it's intended function. Does someone else have a good running 4.0L and know what their Fuel Trim is?

 

[Sidewinder CC]

This is killing me
Is there ANY way to know definitively if I'm running too rich? The cat obscures a slight rich condition by performing it's intended function. Does someone else have a good running 4.0L and know what their Fuel Trim is?

Let me check my O'Scope... No wait, I don't have one.

All I can tell you is what I see on my '99, which is behaving itself at the moment.

After start up, and at first going into closed loop, the STFT shows about -3.0% to -3.6%, meaning it's pulling fuel off. I expected that.

After warm up to operating temp and driving it, the STFT comes down to -1.0% to 0.0% at idle and 0.0% at 2,000 RPM (standing still) and the same while under load (driving). LTFT stays at around -3.0% until it is dumped after driving for awhile.

If you see above -10% LTFT you have a potential problem. If you get to -30% LTFT, Say Goodnight Gracie.

 

[blistovmhz]

Just started my Jeep. It's cold and damp out.
FT = -9
STFT = -8
LTFT = +7.8
Fuel consumption at idle = 0.98G/h (gallon per hour)

Once it warms up, it'll show: (By warm up, I mean the post-cat temp is 450F).
FT = -2 to +2
STFT = -2 to +2
LTFT = +1 to +3
Fuel consumption at idle = 0.64G/h


What's confusing me is that my FT/STFT are always negative or closer to zero, while the LTFT is always around +3. I don't understand how this can be true if the STFT is never +3.

Anyhow, here's some pictures/video.

This is about 40 seconds after startup.

The smoke it's making.

Link to video:
http://www.youtube.com/watch?v=DKNq0SHuOGs

 

[CobraMarty]

Once warm, drive it around and watch the 'Fuel Status' and see how often and what conditions it goes into open loop.

 

[blistovmhz]

Goes into Open loop at WOT, and when decellerating with no throttle, with RPM's above 1200. Otherwise it always stays in closed loop.
The video posted above is closed loop as well. Again, in terms of performance, fuel economy, and smoke production, nothing changes if i'm in open or closed loop. I force it open, and i still get the same shitty 16mpg, same smoke, same loss of power. Let it go into closed loop and identical response.

 

[blistovmhz]

Just figured something else out.
When the engine is warm and not making any smoke, If I pull the #6 injector wire or spark lead, i get smoke again. Doesn't do this with any other cylinder.
So if it doesn't get spark and/or doesn't get fuel, it makes smoke.
WTF?

 

[Digger87xj]

Could you explain what you mean by "downstream voltage signal should be very steady" ?

The downstream O2S is supposed to switch like the upstream O2S. If the downstream O2S is 'steady' something's dead.

I don't see this when viewing my O2S activity with my scanner. My downstream O2S switches around the same as the upstream sensor, about every 2 seconds.

'99 XJ.

If you can view a graph of the O2s on a scan tool output, you would see the upstream with a spiked pattern; Very quick and extreme switching from lean to rich (0v to.9v) as the ECM adjusts fuel injector pulsewidth. Looking at the downstream you should see even, smooth swings that look fairly flat. Usually some where between .3v to .6v. If the down stream waveform becomes steeper and spiked like that of the upstream, the CAT is loosing it's ability to clean the exhaust.

The downstream 02 is used to "fine tune" fuel mixture to keep the CAT inside this target window of .3v to .6v If the ECM sees the D/S voltage trending too high, i.e. staying more towards .6v, it will lean the mixture to set the trend more neutral, and vice-verse.

The ECM will set a CEL for cat efficiency when the average stays too high, and it cannot compensate by adjusting mixture, or the waveform becomes too spiked and starts reacting quickly to fuel mixture changes.


All that being said, a downstream O2 will not, typically, affect driveability in the least, including fuel efficiency.

 

[Digger87xj]

Have you considered the possibility of a MAP sensor failing?

 

[blistovmhz]

Checked map against a vacuum gauge. Looks bang on.
Again, the problem isn't just about the fuel economy. It's mostly about where that smoke is coming from, and the fact that I get that smoke whether it's in open or closed loop. Seems like it has to be a vacuum leak, cat failure, exhaust leak, or ... some other mechanical failure.
No sensor data is going to matter when we're in open loop mode.

 

[Matthew Currie]

According to the 99 FSM, the upstream sensor will be seen to switch frequently, and when the catalyst is working, the downstream sensor will switch less often, and when the catalyst is working and storing excess oxygen, the downstream sensor should show lower overall oxygen amounts than the upstream. When an unspecified ratio of downstream to upstream switches is reached, it sets the MIL. When no catalyst action at all is occurring, the two sensors will read the same, though there will be a time lag between the two. In the case of the 99, there is no mention of the downstream sensor influencing any running parameters. Its function is only to monitor catalyst efficiency.


A question for blistovmhz: Have you checked the individual injectors for leakage and the like? I lose track of what's been done here, but if pulling an injector line causes a problem, I'd check for dribbling and other injector faults.

 

[5-90]

Let me check my O'Scope... No wait, I don't have one.

All I can tell you is what I see on my '99, which is behaving itself at the moment.

After start up, and at first going into closed loop, the STFT shows about -3.0% to -3.6%, meaning it's pulling fuel off. I expected that.

After warm up to operating temp and driving it, the STFT comes down to -1.0% to 0.0% at idle and 0.0% at 2,000 RPM (standing still) and the same while under load (driving). LTFT stays at around -3.0% until it is dumped after driving for awhile.

If you see above -10% LTFT you have a potential problem. If you get to -30% LTFT, Say Goodnight Gracie.

Don't need one.

With OBD-II, an ELM327 adapter (Bluetooth) and Torque Pro (five bucks) will give you the same sort of display, per HEGO sensor, and you can watch them all at once. Direct HEGO signals, LTFT, STFT - you can put them all up on the screen and see what's going on.

MUCH cheaper than an O-scope, and you can watch it while driving (or with someone else driving, or...)

Oh - and you can datalog, so you can examine it later. Nice plus (can't do that with an O-scope.)