Question about O2 Sensor and Bad Temp Reading

I've spent the better part of two hours now reading previous posts and I'm a little confused so I thought I'd pose my scenario and see what you guys think:

98 XJ 2wd
AW4
8.25 with 3.55 gear ratio
bone stock with no extra gear on factory tire size for the Limited model
155k

I have two concerns:

1. My mileage has taken a drop recently. I'm getting between 13-15mpg. 6 months ago I was getting 16-18mpg. Idle quality is fantastic and I have no MIL. The jeep has plenty of power and shows no outward signs of distress. I did a tuneup about 1000 miles ago (plugs, wires, cap, rotor and replaced a piece of hosing that runs along the valve cover (PVC?). I run 10w-30 mobile one synthetic oil. I'm toying with replacing one or both of the oxygen sensors. I have every reason to believe they're original with 155k on them; however idle quality is great with no surging or symptoms of poor running. When I start the Jeep from a cold, the 4.0 is noisy (exhaust leak?) for about 15 seconds then quiets down. It's done this for 3 years and it getting no worse. It is not valve noise and I have solid oil pressure at all times (45 cold and 30 hot). Could i have a crack in the exhaust manifold that's getting worse or maybe just a bit of piston slap? I don't want to replace the O2's if they're fine only to find out I need a header. If I replaced the O2, should it always be both?

2. My temperature gauge gives me inconsistant readings. The Jeep will warm up normally and the gauge will come up then fall out or give a half reading. It never overheats and I have solid heat in the cab. I'm thinking I need to clean and refresh my battery connections and the CTS connection. Am I right in assuming this is a gauge reading issue, and not an actual temp drop in the engine? Radiator, T-stat are all fresh and radiator hoses get warm as t-stat opens. Could this be impacting my mileage drop off?

Thanks for your help everyone.

1. There are a dozen reasons you could be loosing that 3-4 mpg. First start with that temp issue. Is your torque converter locking up as it should be? Do you have a brake dragging? Are you using different fuel? Do you have a clean fuel filter? How are your injectors looking/sounding? How is the lube in your 8.25? Tires at proper PSI? Wheel bearings in good shape, have you ever repacked/replaced the front bearings? Have you ever replaced you catalytic converter? How does your exhaust manifold look, is it cracked? Throttle body and intake gaskets in good shape? How is your EGR valve? Is it clean is the tube clean and free of cracks? Even though you CEL is not on have you checked for any codes?
As you can see there are many things you can look at and check/test before throwing good money at parts to try to figure this out.
2. Check the wiring for the temp sending unit. Connectors can come loose bad grounds could give a false reading. Get a lazer temp gun and shoot different areas of the head, radiator, thermostat housing, exhaust and intake manifold, and your radiator hoses. This will give you a good indication of any temp changes that could be really occurring.
Sounds like if this is a real temp drop then it is probably what is causing you to see that drop in mpg. But checking out everything else isn't a bad thing. Clean up all those grounds while you are in there too that never hurts. Good Luck let us know.

Note from DJ
I am in total agreement with all suggestion offered by 1985xjlaredo.
Alco take a look at this posting; http://www.naxja.org/forum/showthread.php?t=99498
OP by 98XJSport on 10-8-2006
What sensors come online when the idle switches from the higher warm up idle to the lower normal idle? I have a 98XJ 4.0 auto. I’m just trying to track down the acting up sensor that only happens after it switches idle mode.

Look at all of the good information given please.

You will not find the conclusion until post # 16 by OP 98XJSport on page # 2 dated 10-2-2010.

Please keep us posted as to your progress.

Thank you for the quick and detailed replies with valuable information. I will try to answer your questions with the information I have before moving forward:

1. Yes, my torque convertor is locking up as it should at about 42 mph.
2. I don't have any brakes dragging. I replaced the front pads and rotors about 1500 miles ago and they have performed well with no pulling or dragging. There was no sign of corrosion or rust on the guide pins.
3. I am sourcing fuel from multiple locations. I always run regular and only get gas from Chevron, 76 or Shell stations.
4. I have not checked the fuel filter. I was under the impression that a bad fuel filter would provide driveability or hard start conditions. I have neither, but I'm certainly willing to replace it to rule it out.
5. The injectors sound normal. That is to say they sound like they always have in the three years I've had the XJ. The idle is smooth and consistent and acceleration is smooth with full power.
6. The lube in the 8.25 is one year old. I had the pinion bearings replaced when the rear diff starting growling.
7. To be honest, I know nothing about the condition of the wheel bearings. They don't make noise.
8. I have not replaced the catalytic converter. I bought the XJ with 133k on 09 so I assume the exhaust system is original.
9. The exhaust manifold is definitely original and I believe it has a small crack. When I start the XJ in the morning I can hear a very minor exhaust leak for about 15 seconds. When the header warms up a bit, I believe the leak is plugged. I was aware of this when I bought it, but it's gotten no worse in 3 years. I never hear an exhaust leak while driving or when pushing the motor. I have sprayed WD40 around the intake manifold while the engine is running to rule out an intake leak. The idle remains constant.
10. The throttle body and intake gaskets are clean. I took them apart and cleaned them when I did the tune-up about 1000 miles ago. I was amazed at how spotless the throttle body was.
11. I have not checked the EGR valve.
12. The CEL came on about two years ago and showed codes for NSS. I rebuilt it using steps from this site and the light has not re-appeared since. I have not checked for recent codes.

Again, thank you for the excellant advise. I will check the EGR for sure and look at replacing the fuel filter. I followed the link in the second response and it appears the ultimate solution was in fact a replacement oxygen sensor. I have priced the upstream sensor at around $55. I am willing to try this since it's not that much money and I'm quite sure it's original. Should I replace both the upstream and downstream at the same time? I'm thinking yes.....

Um, I would get out your ohm meter and check CTS at cold to hot and make sure it's good .

Then do the same with the MAT sensor, relative resistance is the same for each. Check the FSM I only have OBDI FSM so I cant tell you what they should be.

Better to check before you start "throwing parts at it".

Oh, and no EGR on that era 4.0l

Sounds like you have covered you bases well. Have you had any luck with the temperature reading issue?

So I've done some more research on here and I want to clarify something. Does my 98 have a CTS and a sending unit for the gauge on the head by the valve cover? Some posts say the 97+ just has the one CTS on the T-stat. Is this correct?

Tomorrow I'm going to pull the connection at the CTS and clean it. If the connection is bad and the gauge is cutting out, does that mean the engine is entering open loop and using more fuel until the connection is re-established?

My plan is to clean inspect the CTS and replace as necessary. I am then willing to replace the upstream O2 as well and run a can of BG44k thru the fuel tank.

How does this sound? I figure if cleaning the connection doesn't work the two sensors will run me around $100 to ensure closed loop running. Since it's got 156k on it I figure the sensors owe me nothing at this point. Thoughts?

Typical O2 sensor service life is 100 - 150,000 miles. The O2 sensor become slower to respond to changes in exhaust gases and you MPG's can decline slightly. A drop of 40-50% in MPG;s usually signals O2 sensor failure. You can test O2 sensors for proper functioning, the procedures have been posted here previously.

Fluctuating temperatures on the dashboard gauge usually signal a sticky or failing thermostat.

Keep in mind, many states require winter blend and summer blend gasolines for smog regulations. Winter blend gas and cooler temperatures can result in as much as a 10% drop in you MPG's, summer vs. winter.

I would normally suspect the thermostat as well, but the way the reading drops off leads me to believe its a connection issue. Plus, the thermostat is 1 and a half years old with about 10k on it. It was replaced at the same time as the radiator. The engine appears to warm up as normal.....idle is 750 spot on and I have warm heat in the cabin after just a couple of minutes of driving. The gauge will sometimes be slow to read period, or will come up and then suddenly drop off or give a half reading. It then always jumps back up to just under 210 or just under halfway on the gauge. If the temperature of the engine was really fluctuating that quickly wouldn't I see a drop off in heat in the cabin?

I still have the same questions: Is there a second sending unit for the gauge, and if the CTS on the T-stat housing is giving a false reading is it puting the ECM in open loop regardless of the fact that the engine is warm?

I do agree the winter blend may be contributing, but last winter I was averaging 16.5 and now I'm lucky to eak out 13 or 14 while feathering the throttle. It just seems like a bone stock 2wd that runs as smoothly as mine should do better.

Are all your tune-up parts fresh or tested good, even the coil ? You noted that you recently have cleaned the throttle body and Idle Air Controller. Used a bottle of SeaFoam or BG44 injector cleaner in the gas tank recently ?

Thermostat are inexpensive, even the good quality ones. New or old, a weak or faulty thermostat is faulty. If you have temperatures fluctuations, a fresh thermostat is cheap, simple, easy, and a logical replacement.

1997-2001 have just one Coolant Temperature Sensor on the thermostat housing. Inspect the CTS wire plug for corrosion and the wires for melted insulation or other damage. A faulty CTS can cycle the engine back to open loop. Test its function or simply replace it as part of your routine Preventative Maintenance.

A cracked manifold lets fresh air into the system and the O2 sensors can read that as lean, increasing the fuel mix richness. You should replace the manifold either way, it will eventually pay for itself with slightly increased fuel efficiency.

The fuel 97-01 fuel filter is only serviced if fuel pressure or flow volume testing indicates a problem or if you are changing the fuel pump assembly.

Replace the radiator cap, it may be bad. Check the over flow bottle for air bubbles by running the engine at about 3000 rpm for a few minutes once it is hot.

As others have said, you have no EGR, only one CTS, so check, test the CTS per the FSM with an ohm meter, hot, and cold. Exh Manifold leak can cause bad gas mileage, related to the O2 sensor getting bad data, and your O2 sensor IS on its last legs (age..., slow response hurts mileage but does not throw a code) , which may be most of your mileage problem. If the engine is running cold, (bad T-stat) that will hurt mileage, but it sounds like that is not your problem, but you temp data posts are a bit confusing, and hard to follow. Get some real temp data, get an IR gauge and use an ohm meter to read the CTS directly, and post the actual ohms and temp data.

The plug on the t-stat housing is the Coolant Temp Sensor. If it was suffering a faulty connection, the engine would most likely not run. Try it yourself...With the engine idling, pull the plug, and the engine will stumble and stall. Seriously, don't try this unless you've got a code scanner, because it WILL throw a code. The CTS does go bad, but fails soft so that it doesn't usually throw a code.

Given the miles on your vehicle, it's time to replace the first, or 'upstream' O2 sensor, and the CTS. I did both last winter on my '96, and MPGs jumped back up to where they belong. The CTS and O2 sensor are the two main culprits in MPG problems, as they are the main devices that need to be satisfied to allow the vehicle to attain closed loop operation.

Thank you again everyone for your input. I made some calls today and the prices for a CTS are:

$18/30/42 for the value brand thru the premium. Based on previous posts I think $42 is money well spent for a quality sensor.

The upstream O2 is $57 for the NTK.

I'm thinking I'll replace both and run a bottle of BG44K through the fuel tank and see what it does.

It may take me a few days to get this done as it's raining and I have work, but I'll report back with the results. Hopefully this will solve both the MPG and weird temp readings.

PS: The radiator cap was replaced a year and a half ago with the radiator. Sometimes the gauge will work correctly and I can see the temp come up to about 210 and then drop as the t-stat opens. I believe it to be working correctly.

My opinion; buy a cheap IR thermometer and a DMM, and see if the CTS is actually the problem. I'd also get the NTK O2 sensor from RockAuto; they were markedly cheaper than the parts stores here, for the same parts.

Whilst theres been a heap of good info and advice offered in the thread ....

FWIW .... Check and clean the temp sensor plug ... and make sure its seated firmly.

Had your described issue of intermittent coolant temps being displayed, last year ... It was going from 195* to 100* and back again.

.... and it was just the connector.

Test the other stuff as well tho ... as its all good preventive maintenance/fault finding.

I thought I would take a minute to update this thread. I was finally able to work on the XJ despite work, snow and endless rain. The solution was incredibly simple and was made possible by the excellent advice from this forum:

First I checked the system to make sure I had coolant to rule out an air leak causing the false reading. The radiator was completely full so I moved on to the temp sensor.

I pulled the connector and inspected the two prongs. They did not appear corroded but I gently ran a file along them and then blew out any dust. I then applied a small dot of dielectric greese to the prongs and re-installed the connector.

The result has been a perfectly working temp gauge over 10 warm up and cool down cycles. I believe I can call this a success as the gauge used to fail every day. I filled up on fuel and will be closely monitoring mpg to see if there's an uptick now that the PCM is getting correct data regarding warm up. A quick and dirty read of the mileage computer showed roughly a 10% increase, however this is by no means scientific.

Thank you everyone for your help. I may still replace the O2 out of preventative maint., but at least I solved the problem and it cost me nothing. Jeeps are great, and the support community for them is second to none.

Figuring the O2 sensor didn't owe me anything and it was payday, I went ahead and changed out the upstream sensor two days ago with an NTK unit. First of all, I was very pleased with the quality of the replacement unit. It came ready to go with a factory OEM connector and applied anti-seize compound capped and ready for installation.

Installation was very simple and I've had no problems with the 50 miles I've driven since then. One thing I found interesting....at first I thought the sensor had been replaced at least once because I found what appeared to be a zip tie pulling the wire away from the pipe, but upon closer inspection, it was a plastic cap on one of the oil pan bolts. Is this original? It was brittle, but still holding the wire. If this is original equipment, I believe the senor to be original with 156k on it. It sure looked old.

Anyway, I'll go thru a few tanks and see if anything changes. I did notice the overhead computer said 87 miles until empty at installation and the next day it said 137 miles. I know that's by no means scientific, but a guy can dream right?

good to know

Yeah, the zip-tie/bolt cap is factory. I've always heard it takes about 2-3 tanks of gas to see the effects of an O2 sensor. Another popular thing is to disconnect the battery when you change a sensor, reason being you're starting with a fresh memory, to let the computer 'learn' from a baseline. While your hands are dirty, a new Coolant Temp Sensor wouldn't hurt, either. It's at the T-stat housing (96+, OBDII). It's the other sensor that must be happy in order to achieve closed loop operation.