We could have named this one “Misty” - there was a poisonous cloud coming from the pipe.

Work Order
1993 Jeep Wrangler
2.5L
5 speed manual shift
87,460 miles

Starts hard, runs rough

Preliminary Inspection: Under the hood

 At first glance, it was a neat-looking little ride, and judging from the books I saw in the passenger seat, it belonged to a college student.  From what I could tell from my preliminary, he did most of his own work.  He had a switch panel mounted in front of the passenger’s seat with a colorful array of wires which led out under the hood to some relays mounted on the inner fender.  From there, more wires snaked their way to the lights he had mounted at every corner of the vehicle.  I’m always tempted to straighten a mess like that out, even if the customer doesn’t have it written up, but I resisted the temptation and focused on the problem.  When I turned the key, the engine ground for a few puffs, and then the overrunning clutch on the starter drive gave way and the starter motor whined in protest.  I turned the key again.  Yah yah yah whine.  Yah yah yah whine.  I knew the guy had been starting this thing because he had driven it in for service.  One possible cause of a hard start in a port fuel injected engine is a stuck closed Idle Air Control.  The spark plugs will foul quickly if the IAC doesn’t allow some air to pass the throttle plate while the injectors are delivering the extra fuel needed to start a cold engine. I smelled gasoline and wondered if I had a problem of that sort.
  I kept fiddling with it, pressing the accelerator to the floor to shut the fuel injectors down (this is called “clear flood” mode in the books; the injectors won’t kick in until the engine finally starts and surpasses approximately 400 rpm, which means it has finally started), and finally managed to get it to breathe some fire.  It ran like a three-legged dog.

Wrangler History

 Jeep fanatics will know that the Wrangler hasn’t always been called Wrangler.  First it was a gleam in the eye of a design engineer named Karl Probst at the Bantam car company, and the prototype was called the “BANTAM RECONNAISSANCE CAR” (BRC). It passed every test the Army could subject it to. A quiet firestorm of controversy concerning the origin of this distinctly American vehicle has been going on ever since. When the smoke cleared in the early years, Willys wound up building it as the “GP” or General Purpose vehicle.  There are some grizzled old veterans of World War II who will say that the Jeep won that war.  The franchise has changed hands a few times.  Chrysler dumped the old round headlights in 1987 and gave the primary title of Wrangler to this veteran soldier.  While Jeep’s 2.5L four-cylinder engine received throttle body fuel injection a few years before Chrysler bought out AMC, the old 258 cid carburetor-equipped straight six hung on until the Wrangler was outfitted with Cherokee’s powerful 4.0L, complete with Chrysler multi-point fuel injection in 1991.  At the same time, the 2.5 liter 4 cylinder was retooled to receive its own little set of four port fuel injectors.

Sputtering and Misting

 As I limped the Wrangler into the shop, I had to drive down an aisle between two rows of busy technicians.  Since re-starting it was so much trouble, I left it running.  I’d have to work fast. If I let it belch out too much mist it would ruin the taste of everybody’s coffee.  I hooked up the scope pronto and saw a pattern that indicated that the problem had to be on cylinder #1.  I shut the engine down long enough to hook the fuel pressure gauge to the rail.  I already had a pretty good idea what the problem was.  Pressing the accelerator to “clear flood” mode once more, I managed to get past the repeated kick-out-and-whine cycles of the starter drive and fire the engine up again after a half-dozen attempts.  The fuel pressure was bouncing on about 32 pounds with the engine running.  But when I killed the engine, the pressure dropped rapidly.  On a healthy fuel system, the fuel pressure should hold steady, with very little pressure loss for several minutes after the fuel pump stops running.  If the pressure leaks back through the fuel pump check valve or past the fuel pressure regulator back into the tank, the fuel boils in the rail, filling the fuel rail with air bubbles, causing hard hot-engine starting.  If the fuel leaks through the fuel rail or through a ruptured fuel pressure regulator diaphragm into the intake, it obviously floods the engine, which also causes hard starting (more often a no-start). Furthermore, the gasoline trickles past the piston rings and dilutes the crankcase oil. 
There is one another possible cause for this type of pressure loss. If the fuel is leaking out of the system into the environment, it’s generally pretty easy to spot, since fuel will be dripping from the area around the leak.
 I knew when I saw the fuel pressure dropping that I had partially pinpointed my problem.

Scoping it out

 Connecting the oscilloscope, I had a quick look at the pattern, and saw a firing spike on number 1 cylinder that looked a lot different from the rest.  Ordinarily, and high firing spike indicates high resistance on the corresponding spark plug circuit.  Usually, when a spark plug is fouled with oil or gasoline, the spike will be shorter on that cylinder, but this injector was blowing so much fuel in there that the spike on number 1 actually indicated higher resistance than the rest.  At any rate, I knew which cylinder to focus on in my investigation.

Out with the spark plugs

 I screwed all four spark plugs out and had a look. Number 1 spark plug (far right plug in photo) was nice and wet.  Switching on the key to briefly run the pump (it only runs for a couple of seconds at key on), I shined my flashlight into the spark plug hole into the number 1 cylinder and saw a puddle of gas gathering on top of the piston.  Right away, I siphoned the gas out of the cylinder with a vacuum bottle (homemade tool of mine).  I have on occasion disabled the ignition system, covered the holes with shop rags, and spun the engine with the spark plugs removed to blow the gas out of the cylinders in situations like this, but it’s not a very good idea to do it that way.  One guy I used to work with used that procedure on a Jeep Cherokee 4.0 liter without killing the ignition system.  The spark ignited the fuel, and the Jeep looked like it had an anti-aircraft gun under the hood for a second or two as big balls of flame belched up from the holes and into the air.  Needless to say, everybody was measurably impressed at his pyrotechnic display.

Watching the leak happen

Removing the fuel rail, I got my drain pan ready and had my helper switch the key on.  The number one nozzle was spraying a steady stream into the drain pan, even with the wire connector unplugged, so I knew there were no electrical shorts on the injector signal line or driver problems in the PCM.
 Every so often, an injector will drip a little and flood a hot engine.  This problem will manifest itself as an intermittent hard start, usually after a hot soak.  It’s interesting how one injector can drip a few drops of fuel into the intake and flood the whole engine, but I’ve seen it many times over the past dozen years or so.  The vapor apparently permeates the hot intake and immediately wets the spark plugs when the intake air starts flowing.  The intermittent factor comes into play because the same intake valve isn’t always open when the engine shuts down.  When you’re looking for this problem, let the engine get good and warm, then shut it down.  After about ten minutes, start the engine and let it run for about five seconds or so, then shut it down again.  Remove all the spark plugs and you’ll find the ceramic on one spark plug soiled with hydrocarbons. That will be your suspect cylinder.  Next pull the fuel rail, being careful to inspect the injector tips.  You should see several dirty injector tips on every injector except the bad one, which will have its tip washed nice and clean by the gasoline. Bingo.  This is the best way I’ve found to locate a single injector with a very slight leak.

DTC’s

 I had already condemned the injector, but I generally like to have a look in the scan tool window to see what I can find in there while I’m at it.  This time, I found an “O2 STAYS AT CENTER” code.  We see this a lot on Jeeps, and it may or may not have been the result of the intense hydrocarbon fouling the Oxygen Sensor experienced because of the bad injector.  Sometimes the Jeep Oxygen Sensors will intermittently go sour and hang on a voltage slightly above the 0.5-volt mark. The PCM will think this mildly rich Ox sensor signal is legitimate, because most of the time the O2 sensor behaves normally.  When the PCM tries to use the Ox sensor signal as a feedback when  adjusting the fuel trim, and engine performance begins a sudden nosedive. When the PCM sees engine performance deteriorating, it yanks the fuel trim back to zero, then drops it into the negative again to try and elicit some response from the stubborn Oxygen Sensor. This syndrome can cause a hair-raising buck-jerk scenario on Chrysler products that will repeat itself now and again on the highway. To complicate matters further, there won’t be a single DTC to guide the technician,  and the scan tool window might even show a normally operating Oxygen Sensor at the time the investigation is under way.  A road test snapshot recording of the fuel trim (if you have the tool to do it) will show a nasty magnified saw-tooth pattern with massive swings between negative and positive numbers while the problem is present.
 I had a life-sized picture of a comeback on this one if I didn’t at least warn the guy that his Ox sensor needed replacing along with the injector.

Fixed

 He okayed the $90 injector and a starter drive, but he didn’t want the O2 sensor.  I guess he was little short on cash.  It turned out that the starter drive we ordered from Chrysler wouldn’t fit the starter, and we couldn’t locate a starter drive locally. It appears that most shops don’t repair starters and alternators any more.  Since there wasn’t a starter drive to be had locally and the one we ordered from Chrysler wouldn’t fit, we installed new starter to save time (if not money) and get him on the road.  In addition to all this, we changed the oil filter and the engine oil, which was about 30 percent gasoline. The young fellow seemed satisfied when we were finished.

Richard McCuistian is an ASE Master technician, and earned a level III Master Engine Technician certification from Ford in 1996.  He can be reached by e-mail at This e-mail address is being protected from spam bots, you need JavaScript enabled to view it