If the fuel pressure is not dropping over time, we should be able to rule out a flooded condition, but remember, the inside of the engine is still warm at that point and doesn't need extra fuel to restart. If we go on the assumption you smell unburned fuel that was inside the intake manifold and is going out the tail pipe, that fuel would still have had to leak from an injector. Once the engine cools down more, that raw fuel WILL make it start real easy. You might have the same problem starting a cold engine, but it's just not as apparent.
"they put it on the computer and nothing". Does "nothing" mean there were no diagnostic fault codes or the equipment wouldn't communicate with the vehicle? "Nothing" is hard to interpret.
Allow me to derail my train of thought for a moment. Lets find out where you're smelling that raw fuel. Take a whiff under the hood before you start the engine, both warm and cold. If you smell fuel there, look on each injector to see if one is wet. There's rubber o-rings between each injector and the fuel rail. Those have been known to shrink in cold weather and leak. You SHOULD see the fuel pressure dropping, but it doesn't take much of a leak to make a mess long before the pressure drops.
If you only smell the raw fuel at the tail pipe, and we rule out a leaking injector because the fuel pressure holds steady, that only leaves a sensor issue. At this point I should explain what it takes to set a diagnostic fault code related to a sensor circuit. The throttle position sensor circuit is the easiest to explain. That sensor is fed with 5.0 volts, and for this sad story there's 0 volts on the ground wire. The movable contact inside goes from low voltage at idle to high voltage at wide-open-throttle, but there's mechanical stops that prevent it from going lower than 0.5 volts or higher than 4.5 volts. The only way the signal voltage can go all the way to 0 volts is if there's a break in the 5.0 volt feed circuit, (inside the sensor OR in the wire going to it). 5.0 volts on the signal wire will trigger a diagnostic fault code. Similarly, the only way there can be 5.0 volts on the signal wire is if there's a break in the ground circuit, (inside the sensor or that wire). 5.0 volts on the signal wire will also trigger a fault code but it will be a different one. The important point is the signal voltage has to go outside the 0.5 to 4.5 volt range to set a fault code.
When you get to a sensor with electronic circuitry inside, the same rules apply, but it is possible to get an incorrect reading that is still within the acceptable limits. Chrysler is the only manufacturer that doesn't need a mass air flow sensor to make the engine run right. On all other car brands, the mass air flow sensor has the biggest say in how much fuel the Engine Computer commands to go into the engine. Most mechanical failures result in not all of the air being measured, so you don't get enough fuel. Mechanical issues include a leak in the tube connecting the mass air flow sensor to the throttle body, and bug juice or other dirt on the sensor's element. Electrical issues can result in the sensor telling the computer there's more air going in than what actually is, and the computer will respond by adding more fuel, sometimes to the point the mixture is so rich you can smell it at the tail pipe. This can be a case where the signal coming from the sensor is a valid signal that the computer will believe, but it's the wrong value. As Ford always does everything differently than all other manufacturers, their mass air flow sensors don't develop a signal voltage. They develop a series of square-wave pulses, and the frequency varies in step with variations in air flow. There is a real lot of electronic circuitry inside mass air flow sensors, and the environment those sensors live in is not conducive to their reliability. The point of all this is to say there can be a sensor problem even though there is no diagnostic fault code set related to it.
The next one to look at is the MAP sensor. That is the only one Chrysler uses for its main fuel metering calculations. All the other sensors just fine-tune those calculations. Before the engine is cranked, the MAP reading represents barometric pressure. On other car brands the MAP sensor is used in a number of different ways. On some it continually updates the barometric pressure readings. On some, GMs in particular, it provides a backup strategy if the mass air flow sensor fails. The engine is supposed to run poorly, but it still runs, but actually, it runs pretty well.
I don't know how involved the MAP sensor is on Ford products, but they put it there for a reason. MAP sensors measure intake manifold vacuum. They don't actually do this, but they are so sensitive they could measure engine speed with them by counting the miniscule pulses of vacuum variation each time a piston takes in a gulp of air. Since they do respond to such extremely tiny changes in vacuum, you can imagine how they would react to a vacuum leak. In fact, that shows up sooner on Chrysler products as a running problem, and usually results in a real inexpensive repair. A vacuum leak, especially on older cars that used a hose to connect the sensor to the intake manifold, would result in a lower vacuum reading. You normally have lower vacuum under acceleration or heavy load, so the Engine Computer responds by adding more fuel. Fix the vacuum leak; fix the over-fueling problem. This is another example where the sensor is generating the wrong signal voltage, but it is a valid one the computer believes. There are other things the computer looks at to determine when it needs to set a fault code, but that's not part of this wondrous story.
The first thing I would do is inspect the fresh air tube between the mass air flow sensor and throttle body to be sure there's no cracks, loose hose clamps, or other leaks where unmeasured air can sneak in. Next is to check for vacuum leaks in the hoses and intake manifold gaskets. Pinch the hoses off while the engine is running. On carbureted engines a vacuum leak causes a high idle speed without corresponding power. On fuel injected engines the idle speed is controlled by the Engine Computer, and it will make up for all but the worst vacuum leaks so you can't go by changing speed when you pinch hoses. Instead, look for other indications that something changed, even momentarily, when you pinch a hose. Most commonly idle speed will drop down too low, then pick right back up as the computer takes over for the leak you blocked.
When the engine is still cold, you can spray water around the intake manifold gaskets and watch if it gets sucked in anywhere. Another often-overlooked clue to a vacuum leak is poor response from the cruise control. If it's normal on your vehicle for speed to drop one or two miles per hour when you come to a long hill, but now it drops considerably more, then comes back with a vengeance and over shoots the target speed when you get to the top of the hill, suspect a vacuum leak. The servo pulling on the throttle cable is run by engine vacuum, and when that vacuum is low, the valves open up more and more in an attempt to get the speed back up, then when that speed is achieved at the top of the hill, it takes longer than normal to dump that vacuum and relax the throttle cable.
We never approve of throwing random parts at a problem as a means of diagnosing it. That's the most expensive and least effective way to do it, ... However, when it comes to choosing between the cost of parts when doing it yourself, and the cost of a mechanic, you might consider obtaining a used mass air flow sensor from a salvage yard to see if that solves the problem. Mechanics have the advantage of looking at sensor readings on the scanner, along with reading fault codes. Experienced engine performance specialists can usually catch things the rest of us overlook.
Don't overlook something as silly as an intermittent spark-related misfire. That will send unburned fuel down the tail pipe. I have that issue right now at start-up with my vehicle. It runs on four or five cylinders for about a minute, then it's fine for the rest of the day. I'm pretty sure it's due to carbon tracking inside the distributor cap. I guess with 180,000 miles on the cap and rotor, I should expect a failure once in a while!
Another not-so-common thing you might consider is worn bushings in the distributor. That can cause an erratic camshaft position sensor signal, but usually that will occur all the time.
Friday, November 7th, 2014 AT 11:08 PM