Jord79

Hi, I'm a new owner with a 96 XJ 4.0 auto.

I bought it with a rough idle issue and still trying to iron it out. Rough idle is terrible within 30 seconds of start up (i.e. when the O2 sensors take over control of the fuel trim) The RPMs will slowly fall until engine almost stalls, then suddenly increase again - this pattern repeats until engine temperature increase when idle steady but is still a little lumpy. In addition to this - the engine will buck until warm and will want to stall when coming to a stop. This may or may not be related.

I used an OBD scan app to check for codes. Initially it gave a code for O2 sensor 2 having a high voltage. I replaced the sensor and cleared the code. Code has not come back after 200 miles of driving, but the rough idle still persists.

So using the OBDii app, I have monitored various sensors and found the following...

During the 'rough idle phase' the STFT is reading as rich and leaning the fuel by up to almost -30% before shooting back up to 0% and repeating - this occurs in time with the drop and sudden increase in RPMs which makes sense if it is essentially cutting fuel. LTFT trim is currently reading at +11% though an earlier scan showed the LTFT drop steadily to -27% during the 'warm up' period. I'm hoping this was just an anomaly.

I need to check the functionality of the O2 sensors before I start investigating other problems, which I can do to some degree with the scan tool - but it is a bit of a head scratcher!

First off, the scan app (OBD Car Doctor) is telling me that there is 1 oxygen sensor present. However, there are also some dynamic perimeters with readings that I think relate to the O2 sensors:

First one is 0211 (Bank 1) which gives a reasonably steady reading around 0.6V, which sounds about right for an O2 sensor?

Next is 0212 (Bank 1) and gives a reading which fluctuates between 0.04V and 0.06V - is this one of the O2 sensors?

I also get a reading 0211 (Bank 2) which is shown in percentage and reads between -1.6% and -2.6%. I have no clue about this and if the banks refer to banks of cylinders, then there shouldn't be a Bank to as the car is an I6.

If anyone has any ideas about these figures, I'd be extremely grateful.

Thanks!

nujeepguy

my o2 -2 was bad, read above 1 volt, changed it. both now read and vary under 1 volt like .3 to .7, always changing.

salad

What brand O2 sensors did you use? There are known problems with Bosch and Denso in these - NTK (the OEM) work reliably.

Have you tried disconnecting the sensors and confirming that it runs well in open loop when warmed up? Closed loop can have problems if the coolant temperature is not accurate. Note that 1996 uses a totally separate sending unit for the dash gauge.

Wiring under the XJ for the O2 sensors can sometimes fall down and get chewed. Would be a good idea to check it out.

Upstream O2 readout should fluctuate a LOT. Rear should be steady as the cat smooths things out.

Jeepin'_Aint_EZ

posting to see new post....

Edit:

Salad is right on. I've actually had a driveability problem when my o2 sensor went bad. My XJ would stumble and almost die but catch itself and continue running. I replaced both o2 sensors (I believe they were Standard Brand but it was a while ago) and found that the upstream I installed was faulty right out of the box. I chased my tail for a bit and circled back, threw another new upstream in and it's ran fine ever since. Good times!

nujeepguy

mine runs fine since change o2-2

97grand4.0

I'm sure you've already checked the fuel pressure? 0211 is the upstream sensor. That's the one that trims the fuel. 0212 is after the cat. That one makes sure the cat is working right, for lack of a better way to put it. You just have the front and back sensor both would be Bank 1. No bank 2.
A steady reading of around .6 is not a good test tool. I recommend a graphing obd2 scan tool for about $40 on flea bay. Honestly if that were believable I would say the sensor is no good. The reading should fluctuate between .1 and .9vdc once the sensor is heated up to temp. A stuck sensor will sit around mid volts, but a slow digital or averaging test meter will do the same. Also it is a heated sensor I think, and the heater needs to work, it's job is to heat it up faster so it works sooner.

Jord79

Thanks for the replies thus far.

I don't know what brand the new downstream O2 sensor is (Generic) - so it could be flawed. Do the figures it shows tell that this could be a problem?

Alarm bells were ringing when the scan app said 'Oxygen Sensors 1 present' (followed by the info ECU1 00000011) should it not say '2 present' if they are functioning correctly?

However, it is then baffling that I am seeing voltage readouts for both (high for 0211 na low for 0212)

I would source 2 X new recommended sensors if I could be sure the problem isn't caused by some other fault. Does anyone have the correct part number for the NTK sensors?

The wiring to the rear sensor looks good.

BlueRidgeMark

The downstream sensor (after the cat) is not part of running the engine. You could literally cut the wires and it would not affect the engine. It ONLY tells the computer when the cat is malfunctioning.

Scanners often show readings for sensors that aren't there. Ignore them.

It can't be said too often or too strongly, use an NGK/NTK sensor ONLY for the O2.

Also check the fuse for the heater circuit, as mentioned. Then check for vacuum leaks and exhaust leaks.

By the way, welcome to the forum and thank you for posting a coherent description of your problem and what you have done about it, and for using actual punctuation and English!

Jord79

Thanks again.

Checked NTK website which gives different part numbers for front and rear (relating to connector type and lead length) a visual inspection should determine which is correct.

I will ignore post-cat sensor for now if what you say is correct.

I will check heater circuit fuse (if I can find it!).

I'm ruling out vacuum leaks at the moment as I believe a a vacuum leak leads to a lean condition and my figures indicate a rich condition. I also believe a faulty or stuck EGR valve could cause this.

97grand4.0

Well, I would personally not jump to conclusions without a bit more testing. No offense but I am too skeptical of your test method to spend money on an 02 at this point.

Jord79

The scan tool does graph, I've just not explained the data properly. I'll post pictures of the results tomorrow. The voltage on the upstream does fluctuate slightly between .5 and .7

heater could be the issue as the fluctuating idle occurs when the ecu first goes into closed loop, but eases after the engine has warmed slightly more

97grand4.0

Yeah if its not going to .1 and .9 or darn close when fully warm you got a bad puppy there.

salad

Yeah that's not normal.



Like I said before, the unplugging test is free

BlueRidgeMark

An exhaust leak will give you a rich condition. It can allow fresh air into the exhaust stream and fool the PCM into thinking you are running lean, thus it dumps in fuel.

Jord79

thanks for this, this is really useful!