1997 Chrysler Cirrus lxi cold starting

I have a couple of thoughts. First of all, do you live on a farm and are filling up from a bulk tank that has summer-blended fuel? It is hard to ignite in cold temperatures.

Next, when it doesn't start, pull out a spark plug to see if it is wet with fuel. If it is not, use a "noid light" connected to one of the injector connectors to see if you're getting pulses from the Engine Computer.

Use a mechanical fuel pressure gauge to find out exactly how much pressure you have. Although this is more of a GM problem, pressure that is low by five to ten pounds will cause very hard starting. Fuel pressure and quality are not monitored in any way by the Engine Computer so there will be no related fault codes. If the problem is fuel-related, the engine should start on starting fluid.

Caradiodoc

You're lucky you don't have a GM product. They make their money on expensive electrical repairs. What you describe sure sounds like no or weak spark. For some reason, a lot of aftermarket scanners will not communicate at times with the car. You might try borrowing a Chrysler DRB3. I've never had a problem with mine, and with a plug-in card, it will do emissions-related stuff on any brand of car starting with '96 models.

Under live data, look for the camshaft position sensor and the crankshaft position sensor. They will be displayed as "no" or "present" during cranking. There is some confusion as to their operation. According to the service manual, one determines spark timing and the other one determines injector synchronization. That implies you will be missing fuel OR spark when one of them fails, but in reality, you won't have either if one fails because pulses from both are needed for the Engine Computer to turn on the automatic shutdown (ASD) relay. That relay sends voltage to the injectors, coils, alternator field, O2 sensor heaters, and fuel pump or pump relay. A quick check of the ASD relay is to measure the voltage on the dark green / orange wire to an injector, a coil, or to either of the small terminals on the back of the alternator. You should find battery voltage there for just one second after turning on the ignition switch, then it should come back during engine cranking. You might also hear the fuel pump run for that one second. If the voltage doesn't come back during cranking, suspect the camshaft position sensor or the crankshaft position sensor. Based on the symptom of cold temperature, my first thought would be the crank sensor because its air gap is critical. If the gap isn't set correctly it can cause intermittent stalling when warm and a hard starting problem. All it takes is one good pulse every second during cranking to turn the ASD relay on and keep it on, but the spark won't be consistent.

Along the same line, the Engine Computer keeps the ignition coils grounded until the pulses from the cam and crank sensors tell it to open the circuit to one of them to fire the plug. When you stop cranking, the ASD relay turns off, which for all practical purposes is just like opening the points on the old breaker point systems. That will fire each coil once. That is likely what is happening when you observe it stumbling after a long crank time. That, and the wet plugs, and the ineffective starting fluid all point to a spark issue. It would sure be nice to see what the DRB3 displays for the cam and crank sensors during cranking. You might also look at the waveforms coming from them with an oscilloscope.

You might get another clue by listening under the fuel tank while a helper is cranking the engine. You should hear the pump running steadily. If it turns on and off, suspect the sensors. The same thing would be happening to a test light on the dark green / orange wire to the coils and injectors. Sometimes the test light can be misleading because it can lose contact with the terminal in the connector due to the bouncing of the engine. I prefer to connect it with a clip lead so I can just stand back and watch.

Caradiodoc

Yup. I get exactly what you're saying, but I don't think a loss of compression is the issue. That only leaves fuel, spark, and correct timing. We know timing isn't the problem because it runs sometimes. There's two potential reasons there are no codes. Most commonly the problem is in a non-monitored circuit. That would include all of the fuel supply system and the ignition coil. The Engine Computer only knows it fired the coil. It has no way of knowing if a spark occurred.

The second reason for no codes is the conditions haven't been met to set the code. As an example, when the computer sees pulses from the crank or cam sensor, it expects to see pulses from both of them. If the 8 volt supply wire is broken, neither sensor will work. Most of the time no codes will be set because the computer thinks the engine isn't rotating.

On older cars, no code would be set for the speed sensor if there already was one stored for the MAP sensor. That's because the MAP sensor was needed to detect high manifold vacuum to realize that the car was coasting, therefore it was moving, therefore there was supposed to be a signal from the speed sensor. That's how it differentiated between a bad sensor vs. Standing still at a stop light.

Those are just a couple of examples of why there could be no code. I know you're not using a simple code reader, but the reason for mentioning the Chrysler scanner is there is no aftermarket scanner that will ever do as much as the manufacturer's equipment. The aftermarket stuff has progressed a long way from just reading sensor values and switch states but they are still always a few years behind. With the DRB3 you can also fire the ignition coil and injectors without having to crank the engine.

There's no point in looking at injectors or the fuel supply system since the plugs were wet and starting fluid didn't help. All that leaves is spark, but you say that's fine too. I'm under the assumption this has been an ongoing problem. If that's not correct and the problem just started, draw a little fuel from the test port, throw it on the floor and see what happens when you throw a lit match on it. We had one car come in with the same symptoms you're describing and after spending all day on it, the fuel turned out to be mostly water. Right after that a second car came in with the same problem. The "fuel" put the match out.

Since we're obviously looking for something unusual, here's one you might consider. When the core inside the crankshaft position sensor is cracked, it can develop multiple pulses. Normally it develops 12 pulses per crankshaft revolution corresponding to firing three injectors but extra pulses can trick the Engine Computer into thinking the engine is running faster than it really is and it will try to create the "idle flare-up". That's where engine speed goes to 1500 rpm for a couple of seconds right after startup. You could compare that to pumping the gas pedal during cranking which would lead to flooding. At that point spark won't do any good.

I'm sure your mechanic hates throwing parts at a problem, but this might be the time to try a new crankshaft position sensor. New ones come with a thick paper spacer stuck to the end or a thin plastic rib molded on the end to set the air gap. If you reinstall an old sensor with the rib, you're supposed to cut the remaining part off, then use a new paper spacer.

Caradiodoc