Problem and a challenge

I like to think through a problem before I start taking things apart, rather than just replacing and seeing if that fixed it (like the stealerships). But, this one has me a bit baffled. This is a problem that has been going on for some time, and now getting worse, but still not failing, which makes it tougher...

If anyone has any ideas, please post. I'll list all the gory details below and will be accurate. I think each is an important symptom. It's an 89 XJ, 4.0L, auto, 4wd, PS, air.

Several months ago I got gas after days of rain (didn't drive in the rain though). About 70 miles into that tank (3/4 full or so) it started stumbling intermittently, so bad that it could stall. I could raise the idle, or accelerate hard, and it would run fine.

I thought it was bad gas.

So I added all sorts of additives, ran several tanks through, then it was fine.

A few weeks later I put in a half tank of the same brand gas, but from a different station (it closed down for some reason (ws a BP) since then BTW), and about 60 miles into it, same thing. I drained the tank right away (through the pump again) and put different gas in. All was fine.

Mysterious, but maybe bad gas brand...

A few weeks later, on the usual brand of gas, it started stumbling again. I put in some Marvel oil in the tank and it ran fine within a few minutes. For several weeks, and over 2000 miles, all was fine if I added the oil. I didn't add the oil one time and it had a little bit of a problem. Recently, it started to have an occassional problem with the oil too.

Maybe the bad gas ruined something...

I then started to think an injector was bad... It would try to cut out - a hard miss - when it would do it. I even changed the plugs, gapped and all. The old ones weren't bad though. No change. No plug fouling.

The Marvel oil seemed to take care of things...

Now, 2 days ago, I am down to 3 gal and stop, put in the oil, and fill up. Immediately upon starting, it stumbles. A rolling and cutting out kinda of stumbling. It was before any new gas or oil could've reached the fuel rail...

Now I'm not sure...

- all the electrical I can check is fine.
- plugs are good.
- seems to act up after a fill up (twice right after), or in the same driving period after a fillup.
- it's intermittent
- pump is noisy, but is always on and constant, although sometimes it's quiter
- Marvel oil in tank helped many times, but it's getting worse.
- it'll tend to right itself after a while if oil is in the tank (until now)
- no oil and it's very bad
- I jumpered the pump resistor to give it more oomph while stumbling and no change.
- I can disconnect the resistor while running and the engine will run for a couple seconds then stumble to a stall. Not an immediate cutoff, so elec connection to pump wouldn't do this.
- Always starts fine, even after a stall
- heavy accel or load is fine, it's idle and constant speed that it stumbles
- when it's real bad, even accel has a little stumble, but is still better than idle
- drained filter from inlet side and the gas looked fine.
- flow at the rail seems fine (I use the pump to drain the tank by jumpering it.
- banged on knock sensor and no effect
- disconnected EGR and no change
- when it runs fine, it's great and gets 21.5 mpg highway
- draining tank thorugh pump seemed to fix it twice
- Marvel oil in tank definitely fixed it all times but the last one.
- filter was new before this
- have run 'dry gas' and Berrymans through in case of water

I'd like not to be stalled and at the mercy of some shop if it goes - probably on a trip. Any ideas on what to test or what it could be? Some of the symptoms seem to contradict others. I've got a trip next week and would like to fix this. Detailed tests appreciated too. Maybe I'm not seeing a test that would help get at this intermittent problem.

Thanks in advance.
-Skip

I don't really know what your problem is, it could be numerous things. Maybe even many things that all add up, been there done that.
A couple of things to look at, if the fuel is running much moisture, you can often see it on the plugs as a rust color. But sometimes this just happens when the humidity is really high.
I was having a serious miss in my Dodge that came and went. I added periodic doses of carb. cleaner and other additives, just in case I had moisture in the tank. Climbed in the truck one morning when it was minus twenty out and heard a rumble form the rear. Jumped up and down on the rear bumper a few times, sounded like a I was stirring a pitcher of ice tea and ice cubes. Pulled the tank and got about 4 or 5 big hand fulls of ice out of the tank. Gas additive will suck up some moisture, but it takes awhile and repeats to get it all, if there is a bunch in there.
Try bouncing up and down on your bumper with the motor idling for awhile and see what happens.
The fuel is constantly circulating in a loop and the injectors are only using a portion of it.
If you feel you have to pump the fuel out again, jump up and down on the rear bumper, while it's pumping out, you may get more of the water. I've got an old Arrowhead glass water jug (five gallon), if you let the fuel set overnight you can see the water in the bottom and siphon the fuel off the top and use it again.
Different fuel companies often get there fuel from the same sources, often the only difference is the additives.
Some years ago, watered fuel was a big problem here, I installed a diesel fuel filter and housing on my pick up truck. The diesel filter has a drain XXXX in the bottom and collects the moisture, which you can drain off periodically.

in this order, i would:
test the fuel pressure,
suck up a can of sea foam or boogie juice of choice via a vacuum line, inspect my fuel lines,
do a fuel pump volume test,
check out the fule pump relay
change the fuel filter again.

You said idle and constant speed. My idle would hunt, up and down a few hundred, in cycles, idle was rough and the motor would buck occasionally at 2000 RPM or so. Turned out to be the O2 sensor (mostly).
Later, I still had an idle that would cycle some, I cleaned all the connectors on the bottom left of the motor (among others). I doubt they had ever been cleaned. Smoothed things out quit a bit.
I cleaned all of the grounds, seriously making sure of every connection. Even soldered the ground rings that weren't soldered from the factory and added a battery ground strap to the front clip and another ground strap from the dip stick holder to the firewall. Found about ten horses I didn't know I had.
I've never had a MAP fail, but have disconnected it to see what happens, it really messes up the fuel ratio and will stall the motor at idle and even at speed if you release the pedal. But seems to want to run really rich, generally.
If your getting 21.5 miles to the gallon, you may have a lean running motor. Take it out on the interstate, run it for five miles at a constant 45-55 or so, pull into a rest stop and pull a center plug while it's still warm and look at the color, if it is chalk white, you could be running lean, it should be a shade of light tan (bone) or better yet a light white/grey. XJ's run a pretty tight fuel air ratio, so sometimes it's hard to tell if they are running lean but 21.5 MPG is something I've never seen in a 4.0 Another sign of a lean running motor, is it will rarely run up to redline in high gear (often starts crapping out as low as 3500 RPM's).

I've thought of the O2 sensor, but its connector is a little hard to get to when it's hot and acting up - I should be able to pull it and it will go into open loop and be fine. However, it has still done this in open loop too (just started), so with that info (there's so much info I've gathered on this I forget it all until prompted...) and the correlation with fillups, I haven't followed the O2 sensor route yet.

The 21.5 is an all-time best at constant 65mph, 100% highway (1 exit and 1 entrance ramp only), flat road and no hard accel. Mostly it's 18-20 if I go easy with some city driving. Even less city mpg. The plugs are all white-grey and look good. I just tried to get the most mpg last trip and that's what I got over a few tanks. BTW, it ran great during that time.

I will try the bumper shake, and forgot one other _possible_ symptom. It was late, and I wasn't sure at the time, but I was going down a road, engine fine, but it had been acting up earlier. There were construction spots that were only 10-20 ft long, but bumpy, so I went kinda slow and took my foot off the gas (engine idle, maybe 20 mph) before each one. I noticed that on each bump, the tach would drop as if it was stumbling. Not big or a stall, but the bigger ones made it do that. It was immediate though, not something where the water would have time to get to the rail.

Like the immediate response I got after filling up....

BTW, my idle motor is disconnected, and my idle set as one would an older car, with a little more to keep the oil pressure up (about 700-800 in D). It's always been rock solid except during these acting up times, which affect cruising too).

Not to change the topic - I still need all the input I can get, but if water gets in your filter, say in the fibers, does it 'block' gas flow even if you dump it out? I've heard of something like that (wet filter paper blocks gas) but don't know the details...

All the above suggestions are good, but while you're at it, don't neglect the ignition. It is possible that a marginal ignition could cause problems that are temporarily helped by other measures that improve combustion a little, such as fuel additives. I had a problem like this with my Chevy truck. I was sure it was fuel related, and a can of injector cleaner would make it well for a tankful, etc. It always started well. I finally found the problem: the carbon button in the distributor cap was gone!

NEW UPDATE:

For 2 days it has stumbled continuously while running. Fuel flow was good, and previously measured was in spec.

So today, I decided I was going to drive it no matter what. Started it and it was a lower shaky idle. I went to the back corner side and rocked it violently back and forth. It got real bad and almost stalled.

Stopped and it went to a rough idle again. Shook again and it started again - as I was shaking it and what should be sooner than the fuel would've gone from tank to rail. Stopped and it stopped again.

Did it a third time and nothing changed, but now the idle was steady. Just drove it 10 miles and not a single burp!

The wife got in, gave me the eye, and asked "are you going to do that every time we go someplace?" :laugh3:

WELL, now I'm back to a week ago. Before, it needed oil to run smoothly, but it did run OK (with oil, otherwise continuously bad). Can a tank of bad gas ruin the injectors or pump in a way that oil is needed to make it run right? I could see something like Berryman's making up for dilution or getting better performance, but oil actually lowers the fuel value while increasing lubrication. I doubt it's counteracting any current water in the tank.

Another symptom added to the list...

skipc said:

NEW UPDATE:

For 2 days it has stumbled continuously while running. Fuel flow was good, and previously measured was in spec.

So today, I decided I was going to drive it no matter what. Started it and it was a lower shaky idle. I went to the back corner side and rocked it violently back and forth. It got real bad and almost stalled.

Stopped and it went to a rough idle again. Shook again and it started again - as I was shaking it and what should be sooner than the fuel would've gone from tank to rail. Stopped and it stopped again.

Did it a third time and nothing changed, but now the idle was steady. Just drove it 10 miles and not a single burp!

The wife got in, gave me the eye, and asked "are you going to do that every time we go someplace?" :laugh3:

WELL, now I'm back to a week ago. Before, it needed oil to run smoothly, but it did run OK (with oil, otherwise continuously bad). Can a tank of bad gas ruin the injectors or pump in a way that oil is needed to make it run right? I could see something like Berryman's making up for dilution or getting better performance, but oil actually lowers the fuel value while increasing lubrication. I doubt it's counteracting any current water in the tank.

Another symptom added to the list...

Click to expand...

Maybe the oil is a red herring. It seems especially suspicious that the problem seems to have been connected with a specific source of fuel. If it changes when you rock it, I'd guess it's either a problem picking up fuel or a wiring problem that happens when you rock. You could be having trouble with the fuel pickup in the tank, or there might still be water in the tank. Have you drained the tank dry? I've seen contamination bad enough that it took more than one can of drygas to soak it up. Maybe additives help you run with the water that's in the tank, but haven't actually gotten rid of it. I had an old Colt Vista that exhibited all the symptoms of water in the tank. I tried everything. It cleared up on the eighth can of drygas.

I'd also get underneath and check out all the wires and hoses. Look for dangling wires to the fuel pump, corroded connectors, etc., as well as the classic CPS grounding on the hot exhaust manifold. My old 87 had a habit of cutting out every time I went over a railroad track. I noticed that the sender/pump harness was dangling so that when I hit a big bump the axle could hit the harness. Tying up all the wires to the tank cured it. I never bothered to find out where the problem originated, because it never happened again in a hundred thousand miles or so.

Back in my AMX days I got a bad tank of gas. The dealership owned by a friend's father had just gotten one of this siphon pump/tanks, so we went down one evening and pumped all the gas out of the tank. AMX tank was 20 gallons, same as an XJ. We pulled over 3 quarts of water out of my gasoline. Lemme tell you, it would have taken a loooooong time and a LOT of DryGas to work that much water out of the system.

Update:

I went on a trip, and it started consistently stalling at idle immediately upon starting after a fillup. After that, some days it runs like new, others it stumbles at idle and cruising speeds.

However, something else went bad - it started having bad accel and a top speed limitation - classic pump or filter symptom. It would only do this after running a while. I pulled the pump and the screen was completely clogged. I replaced the screen (pump pressure and flow fine), all rubber hoses, and filter. That solved the accel problem and it's back to the old problem again.

BTW, the 89 XJ fuel pumps come in 2 types: Bosch (less common) and 'the other one'. I have the Bosch, and it won't take the other one without modding the bracket. I'd prefer the Bosch anyway. Just be aware to pull it before buying - I wasted some bucks on the wrong type (too late to take it back). NAPA can get the Bosch one. Other places just look at you funny.

Anyway, checked the EGR valve once when it was rough and it was fine. I can move it and make the same symptoms, but it didn't seem to be causing it at the time.

Rough terrain seems to aggravate it, and getting out to open the hood and check it seems to fix it. Hard accel is no problem. I got it to stumble for miles at 25 mph by just holding it at speed and getting it to just stumble the whole time. It wasn't good for the wife's nerves, but you gotta do what you gotta do... Next day, I couldn't get it to do it.

-S

I know you checked the egr valve,, but just as a test, you might consider crimping off the vaccum line to it, and see if it helps. The egr when it wears some can intermittently hang open. It can take a bit of sleuthing to find it because it will snap shut again just as unpredictably.

Have you changed the fuel filter?

Here is a shot in the dark, maybe the tank is contaminated with something (rust), everytime you fill up it stirs up the bottom, putting this stuff into suspension. Maybe this clogs the filter or the injectors temporarily or until you puy MMO in. Drop the tank and clean it out maybe. This seems really odd so I thought and odd solution might work.

Good luck
katarn

Yes, the rubber lines and filter were changed, as were everything in the tank but the pump itself. The screen-sock was very clogged, but I couldn't get the whole tank 'cause it hung up on the hitch mounting I have.

Anyway, it was beginning to 'feel' like the EGR after all the fuel components were replaced. I took a screwdriver one night, while it was stumbling, and pulled it open and let it snap shut several times (the EGR that is). Since then, it's run great, even after a fillup without MMO. I don't want to spoil things and pre-announce success, but it seems that the EGR was part of the problem.

If it comes back, I'll focus on playing with that and getting more of a cause-effect relationship going. It also seems like the pump is much quieter with a new screen too, even though pressure and volume were in spec up until the very end.

Thanks much for all the hints. Maybe the MMO did something to lube the sticking part of the EGR valve? Maybe it let the remaining filter move fuel easier. Maybe it was just a red herring. This was a tough one cause it kept hiding away. Hopefully it is done and if it comes back it'll be directly fixable by a few slams of the EGR. If so, it'll confirm all.

THANKS AGAIN!

If it gets mysterious again, I'll repost.

Skip

I had an EGR, that had a weak spring or a diaphragm. Whatever it was, was causing it to have too little pressure, it didn't seem to want to stay closed at lower RPM's like it should. Might want to pull off the EGR and have a look at the piston and seat for build up, a little hard to get to, but cleaning the junk out isn't that big of a deal. I've used a little dab of ceramic putty (muffler repair paste) instead of a gasket with good results, I've also made my own gasket with 3-5 mil copper foil (which is cheap and works for many things). I've got some special high temp. grease especially made to lube the EGR parts. A faulty EGR is often only noticeable at low RPM's above 1500 or so you hardly notice if it's hanging open. I also had really leaky transducer, I guess something internally sprung a leak. I noticed it as a vacuum leak, took me forever to find it.
Rough idle, an idle that changes up and down in cycles or a hard miss at around 2000 RPM's are all signs of a questionable O2 sensor. a bad O2 sensor will act just like fuel starvation from a weak pump or restricted plumbing.

what about a plugged cat?

Check ALL electrical connections and grounds for corrosion, and dirt.
Check Alternator output, voltage and current.
Check Fuel Pump current draw, as the pump gets older the rollers and bushings wear.
Low engine speed combinded with a weak alternator and old fuel pump will cause wierd stalling problems

Well, in the interest of making this a reference for others, here is what happened:

It turned out to be an EGR problem. Taking the shaft between the valve and diaphragm and moving it rapidly in and out several times has cured it. Even with BP gas. I suspect a small bit 'o crud kept it open at times. It didn't feel sticky, and thousands of miles later I think I can declare it "cured".

The fuel starvation problem was classic clogged fuel filter sock in the tank, probably caused, or hastened, by my running it almost dry and filling up with a small amount (many times) in order to flush what I though was continual bad gas. It stirred up years of dust/dirt in the tank, in higher than normal concentrations, and they were filtered out and clogged it faster than would've normally happened.

The 2 problems were technically independent.

Other notes:
The fuel pump seems to be much quieter with the sock replaced, although flow before and after was in spec.

It is curious that the 2 stations that caused my problems were shut down shortly thereafter (I didn't do it . One temporarily when many stations got water infiltration, and the other permanently.

Hope this helps others.

Skip

I'm thinking the vent on the tank is clogged. Next time it starts, stop but leave it running. Go back and loosen the gas cap enought to break the seal and leave it that way. Hop in and drive it a while. If the problem goes away, you most likely have a clogged vent and the negative pressure in the tank is hampering the fuel pressure. Its a shot in the dark, but a cheap test to run.

And by the way, NONE of us around here have ever had our wives give us crap about how a vehicle runs. :laugh3:

It's all fixed...

Vent wasn't a problem, but thanks.