Engine / Trani / Electrical?

Tiny
MIKROBKIRK
  • MEMBER
  • 1993 DODGE CARAVAN
  • 6 CYL
  • FWD
  • AUTOMATIC
  • 249,000 MILES
Engine runs perfect it seems, except sometimes when SLIGHTLY pushing on accelerator to go up a hill or slight incline, the engine and transmission will make a load bang or **** and the gauge cluster needles bang down like everything has lost electrical power for a split second. It only does it when you first just touch the accelerator going up a incline. Does not always do it, but 75% of the time. Does not seem to do it on level ground or downhill. When it happens the transmission also bangs real loud like it has also lost power. I think the trani is going out (been thinking that for 2 years) because it is hanging once in awhile between 2nd and 3rd up shift - but mainly has trouble DOWNSHIFTING with a bang between 2nd to 1st, especially if I pull up to a stop a little fast and apply the brakes. If I coast real slow to a stop it seems to be able to keep up. I have also noticed my fuel economy has been going downhill. The onboard computer tells me I only getting about 12 mpg avg. Seems like it is really bad when accelerating. When keeping it at an EVEN speed on the hwy. It is still getting about 22mpg. On constant speed on level ground. When I hit the accelerator it says I go down to 6 or 7 mpg. I have always done ALL maintainenance on this vehicle regularly. I just put new struts / tires / plugs&wires / trani fluid and filter change last spring. The trani was replaced by a dealer many miles ago (@75K). Will a diagnostic scanner be able to isolate or tell me what the problem is? I just hate to let this van go, it has been so good to our family.
Do you
have the same problem?
Yes
No
Tuesday, November 30th, 2010 AT 12:26 AM

3 Replies

Tiny
CARADIODOC
  • EXPERT
The first part of your story sounds like you might have wires rubbing and shorting to ground intermittently when the engine rocks. Check / wiggle / move the harness that runs along the left inner fender and goes under the battery tray. Look for shiny spots where the paint has rubbed off. If you find that, you will likely find some bare wires too.

Caradiodoc
Was this
answer
helpful?
Yes
No
Tuesday, November 30th, 2010 AT 12:39 AM
Tiny
MIKROBKIRK
  • MEMBER
Caradiodoc - Thank You for your reply. I have checked the harness as you suggested. I did not see anything. Cleaned the connectors and pins with LPS Electrical spray cleaner. Also unhooked the harness that goes to the ECM and cleaned the plug and pins. I can give you some more details - The engine runs fine down the highway and the onboard computer says it is getting descent gas mileage while cruising (21 to 23)down the highway at 65. When I press the accelerator the computer tells me mileage goes down to 4, 6, 8 mpg. My AVERAGE mpg has dropped to 12 mpg. I thought maybe the onboard computer was wrong, but keep a log on last 2 fill ups and it is correct. Must be something with the fuel delivery. Could it be a throttle positioning sensor - MAF sensor - or some other sensor? Fuel Rail pressure seems okay. Replaced fuel filter and new fuel pump 2 years ago. Wouldn't a clogged or worn injector's be noticeable at idle or cause a rough running engine or exhaust smell? Thank You for your time.
Was this
answer
helpful?
Yes
No
Thursday, December 2nd, 2010 AT 3:11 AM
Tiny
CARADIODOC
  • EXPERT
Throttle position sensor will just cause a stumble upon acceleration. He has very little affect of fuel delivery. Ain't no mass air flow sensors on Chrysler vehicles. In the words of one very high-level Carquest trainer, "Chrysler is the only company that has been able to make engines run right on just the MAP sensor". That sensor has the biggest say in how much fuel enters the engine. It is fed with 5.0 volts, and it's output signal must remain between 0.5 and 4.5 volts. Anything outside those limits will set a diagnostic fault code and turn on the Check Engine light. What that means is the signal voltage could be wrong, but as long as it is within that range, the Engine Computer will believe it and run the injectors according to the numbers from that sensor.

Here are some other things to consider. Even if the MAP sensor were to report the wrong value, lets say from a crack in the vacuum hose going to it, the result would be a mixture that is too rich. The oxygen sensor would see that and the computer would respond by commanding the injectors to be pulsed on for a shorter period of time. Where the potential problem comes in is the computer can only add or subtract roughly 10 percent of the fuel from the pre-programmed values for any set of conditions, but the amount of fuel varies much more than that based on the MAP readings. Based on the O2 sensor readings, the computer might not be able to overcome a bad MAP reading. The way you would find that is to view live data on a scanner during a test drive. There are two sets of numbers called short term fuel trims and long term fuel trims. The computer starts off with numbers programmed in at the factory. For any combination of MAP, TPS, ambient air temperature, coolant temperature, engine speed, direction of throttle change, speed of throttle change, etc, there is a set of "lookup tables" that contain the calculations of how much fuel to spray in at that instant. As soon as you start driving, the computer watches the oxygen sensor to see how closely the average mixture is staying to perfect. If it sees, for example, that for a certain set of conditions, it is always running rich by 2 percent, it will subtract 2 percent fuel, and it will memorize that number and use it the next time that set of conditions is encountered. That memorized number is one of many called short-term fuel trims. They constantly update while driving.

Over time, when the computer sees that it is always making the same corrections, it moves some of those numbers into the long-term fuel trims. In my story, that would be minus two percent. From then on, as soon as you start to drive, the computer uses those new long-term fuel trim numbers instead of the original numbers programmed in at the factory. If nothing changes in the way the engine is operating, the computer will hold that minus 2 percent in memory, and there would be no need to make additional corrections so the new short-term fuel trim would be close to zero.

Basically what that means is you can use those numbers to see what the computer is seeing. It's normal to see fuel trims of plus or minus 2 or 3 percent, but as they start going higher, you can suspect something unusual is causing it.

Fuel pressure is not monitored. Chrysler pressure regulators almost never cause a problem, but if the return hose going back to the tank was crushed or restricted, it could raise fuel pressure. The computer calculates how long to hold the injectors open based on the pressure it expects to be in the system. If the actual pressure is higher, more fuel will spray into the engine. In a roundabout way, the computer will pick that up from the O2 sensor and reduce the injector on-time. That would show up as a negative fuel trim number.

Here's where things get mixed up. You must remember that oxygen sensors only measure unburned oxygen. They do not respond to unburned fuel. There was a lot of trouble with the original spark plug wires on the 3.3L engines in the early '90s. If you had one that was not firing intermittently, the unburned gas and air would leave that cylinder and head down the exhaust system. If you stood by the tail pipe you might smell that unburned fuel, but the oxygen sensor will only detect the excessive unburned oxygen. The computer will think all of the cylinders are running much too lean so it will increase the on-time of all six injectors. Now five cylinders will be running way too rich, and one will still be dumping unburned oxygen, (and fuel) into the exhaust, so the computer will still think everyone is too lean. There is no amount of fuel it can add to make itself happy because as long as there is that misfire, there will always be that unburned oxygen in the exhaust. When this happens, the fuel trim numbers will be really high, perhaps more than 8 percent, and fuel mileage will be way down.

Here's another one that is not common and can be real hard to figure out. A tiny leak in the exhaust system ahead of the catalytic converter can cause wrong readings from the O2 sensor. In between the pulses of exhaust gas, the momentum creates a small pulse of vacuum in the system. Outside air can be drawn in through a leak and be detected by the O2 sensor. Here again, no matter how much fuel the computer adds, it will continue to see a lean condition.

Oxygen sensors work by comparing the exhaust gas to atmospheric air through a tiny opening where the wires come out of it. If someone is real generous with rustproofing material and they spray some onto the sensor, it may not be able to work properly if the wires are coated.

Besides fuel trim numbers, you can also watch the O2 sensor switching from rich to lean while driving. That normally occurs a couple of times per second and it will start doing so in as little as one or two minutes after starting a cold engine. If it takes a long time, say ten or fifteen minutes at highway speed before it starts switching, there could be a problem with the wires going to its heater circuit. The sensors must reach 600 degrees before they begin to operate. Rather than wait for the exhaust to get it that hot, they put heaters in them to hurry that up. At idle or slow speeds, the exhaust could easily drop below 600 degrees by the time it gets to the sensor. The computer is programmed at the factory to run a little too rich in that condition. When he mixture is lean, you will notice it as a stumble or hesitation, but you will not notice poor performance when it is just a little rich.

If you use an ohm meter to measure the heater circuit, I can't even remember what normal is, but what you would look for is 0 ohms, (not likely), open circuit, (infinite resistance), or something in between. I'm thinking somewhere between ten and one hundred ohms. It won't be 0 ohms, (shorted) because it is on the same circuit as the injectors, fuel pump, and ignition coils. A shorted heater would pop a fuse and the engine wouldn't run.

Some other things to look for when watching the switching of the O2 sensor is whether it hangs in the rich or lean state longer. If it is mostly saying the exhaust is rich, that implies the computer is aware of the condition and does not have enough control to correct it. A leaky injector would be a good suspect. Fuel trim numbers will be high negative. If it stays lean more than rich, that suggests the computer is adding fuel in an attempt to overcome some problem and isn't having any luck. Fuel trims will be high positive.

If the fuel trims are all close to zero and the O2 sensor is switching rapidly, the fuel system is working correctly but the engine is working harder for some reason. It could be the parachute you're dragging but a more likely cause would be dragging brakes. If you park on a slight incline, shift to neutral, and release the brake pedal slowly, the van should creep ahead on its own. On a flat surface you should be able to push the van by hand. There hasn't been much trouble with brake calipers sticking. You're more likely to find the caliper mounting surfaces are rusty or are impacted with mud so they can't slide and release freely.

Keep in mind too that ethanol has much fewer BTUs than gasoline and using gas with ten percent ethanol reduces fuel mileage by about 10 percent. Cold weather affects mileage too. I ran measured gallons through my '80 Volare many times and found I was getting consistently 28.3 mpg with a 4400 pound car with steel bumpers. As soon as it got cold out, the mileage dropped to just over 19 mpg on the highway and warmed up. Liquid gas will not burn. It must be in vapor form, and since a smaller percentage of fuel vaporizes in cold weather, you have to start with more liquid. That means unburned fuel shooting right out of the tail pipe. The really high pressures in fuel injected systems makes the fuel vaporize much more efficiently, but the cold air still hinders that a little.

The high manifold vacuum during coasting would cause an excessively rich condition from the extra pull on the molecules of fuel through the injectors. To counteract that increased force, fuel pressure is lowered. That way, the net difference between vacuum and pressure remains constant. Pressure is dropped by relaxing the spring-loaded diaphragm in the pressure regulator. That's why there is a vacuum hose attached to it. Check that hose for cracks or looseness. If you connect a pressure gauge to the fuel rail, you will see pressure rise when you disconnect that hose.

Low fuel pressure can cause incorrect readings on the overhead trip computer. Low pressure will result in a lean condition. The computer will respond by holding the injectors open longer. That might get the amount of fuel entering the engine back up to where it should be, but he trip computer performs its calculations based on road speed and injector pulse on-time. That computer doesn't know that fuel delivery is low and is being compensated for. It just knows that the injector on-time has increased so it assumes more fuel is going into the engine. While those computers are usually pretty accurate, if you come up with something different when you do the math yourself, believe your numbers first.

Caradiodoc
Was this
answer
helpful?
Yes
No
Thursday, December 2nd, 2010 AT 5:19 AM

Please login or register to post a reply.

Recommended Guides